CA2429726A1

CA2429726A1 - Artificial chromosomes, uses thereof and methods for preparing artificial chromosomes

Info

Publication number: CA2429726A1
Application number: CA002429726A
Authority: CA
Inventors: Gyula Hadlaczky; Aladar A. Szalay
Original assignee: Individual
Current assignee: Glaxo Group Ltd; Hungarian Academy of Sciences Biologic Research Center
Priority date: 1996-04-10
Filing date: 1997-04-10
Publication date: 1997-10-30
Also published as: CA2429724A1; CA2429724C

Abstract

Methods for preparing call lines that contain artificial chromosomes, methods for preparation of artificial chromosomes, methods for purification of artificial chromosomes, methods for targeted insertion of heterologous DNA into artificial chromosomes, and methods for delivery of the chromosomes to selected cells and tissues are provided. Also provided are cal! lines for use in the methods; and cell lines and chromosomes produced by the methods. In particular, satellite artificial chromosomes that, except for inserted heterologous DNA, are substantially composed of heterochromatin, are provided. Methods for use of the artificial chromosomes, including for gene therapy, production of gene products and production of transgenic plants and animals are also provided.

Description

ARTtFICIAL CHRC?MOSOMES, USES THi:REOF A~ID~ METH~DS F~R
PREPARING ARTIFICIAL CHROM~S(JIVIES

i FIELD C,~F. THE tNVENT10N .
S - The. present invention relates to methods for preparing cell lines' that contain artificial chromosomes, methods for isolatian of the artificial chromosomes, targeted insertion- of heterologous DNA into the chramosomes, delivery of the chromosomes to selected cells and tissues and methods for isolation and large-scale production of the chromosomes. Also provided are cell. trines for use in the methods, and cell (fines and chromosornes produced by the methods. Further provided are cell-based methods for production of heterologous prciteins, gene therapy methods and rriethods of generating transgenic animals, particularly non-human transgenic animals, that .use artificial .
chromosomes.
BACiCGROi,JND OF ThiE INVENTiO>1t ' Severa8 viral vectors, .non-viral, and physical delivery systems for gene therapy and recombinant expression of heterologous nucleic acids have been.developed fsee, e_:,g_, IVliteni et al. (1 S~3) Trends Biotech.
. 20 1 ? =~ 62-166]- The presently available systems, howe4'er, have numerous limitatiflns, particularly where persistent, stable, or controlled gene expression is re4uired. These limitations includea t1] size limitations because there is a Limit, generally on order of about ten kilobases (k8], at most, to the size of the DNA insert (genie]'that can be accepted by viral vectors, whereas a number of mammalian genes of possible therapeutic importance are well above this limits especially if all control elements are included; (2) the inability to specifically target integration so that random integration occurs which carries a risk of disrupting vital genes or cancer suppressor genes; (3) the expression of randomly integrated therapeutic genes may be affected by the functional compartmentalization in the nucleus and are affected by chromatin-based position 'effectsn (4) the copy number and consequently the expression of a given gene to be integrated into the genome cannot be controlled. Thus, improvements in gene delivery and stable expression systems are needed (see, e'4., Mulligan (1993) Science 26~:926-932].
In addition, safe and effective vectors and gene therapy methods should have numerous fee#ures that are not assured by the presently available systems. For example, a safe vector should not contain ~NA
elements that can promote unviianted changes by recombination or mutation in the host genetic material, should not have the potential to initiate deleterious effects in cells, tissues, or organisms carrying the vector, and should not interfere with genomic functions. In addition, it would be advantageous for the vector to be non-integrative, or designed for site-specific. integration. Also, the copy number of therapeutic gene(s1 carried by the vector should be controlled and stable, the vector should secure the independent and controlled function of the introduced gene(s); and the vector should accept large (up to Mb size) inserts and ensure the functional stability of the insert.
The limitations of existing gene delivery technologies, however, argue for the development of alternative vector systems suitable for transferring large Iup to Mb size or larger]_ genes and gene complexes together with regulatory elements that will provide a safe, controlled, and persistent expression of the therapeutic genetic material.

mQ,m At the present time, none of the available vectors fulfill all these requirements. Most of these characteristics, however, are possessed by chromosomes. Thus, an artificial chromtasome would be an ideal vector for gene therapy, as well as for stable, high-level, controlled production a of gene products that require coordination of .expression of numerous genes or that are encoded by large genes, and other uses. Artificial chromosomes for expression of heterotogous genes in yeast are available, but construction of defined mammalian artificial chromosomes has not been achieved. Such construction has been hindered by the lack of an isolated, functional, mamrnaiian centromere and uncertainty regarding the requisites for its production and stable replication, lJnlike in yeast, there are no selectable genes in close proximity to a mammalian centromere, and the presence of long runs of highly repetitive pericentric heterochromatic ~NA makes the isolation of a mammaiiar~ centromere using presently available methods, such as chromosome ~vaiking, virtually impossible. Other strategies are required for production of mammalian artificial chromosomes, and some have been developers. For example, tl.S. Patent No. 5,288,625 provides a cell line that contains an artificial chromosome, a minichromosome, that is about ~,0 to 30 megabases: Methods provided for isolation of these chromosomes, however, provide preparations of only about 10-2(~~/0 purity. Thus~
development of alternative artificial chromosomes and perfection of isolation and purification methods as wail as development of more versatile chromosomes and further characterization of the minichromosomes is required to realize the potential of this technology.
Therefore, it is an object herein to provide mammalian artificial chromosomes and methods for introduction of foreign DNA into such chromosomes. it is also an object herein to provide meth~ds of isolation and purification of the chromosomes. It is also an object herein to provide methods for introduction of the mammalian artificial chromosome into selected cells, and to provide the resulting cells, as well as transgenic nan-human animals, birds, fish and plants that contain the artificial chromosomes. It is also an object herein to provide methods for gene therapy and expression of gene products using artificial chromosomes. It is a further object herein to provide methods for constructing species-specific artificial chromosomes de novo. Another object herein is to provide methods to generate d-a novo mammalian artificial chromosomes.
1Q SUMMARY ~F THE INVENTION
Mammalian artificial chromosomes [MACs) are provided. Also provided are artificial chromosomes for other higher eukaryotic species, such as insects, birds, fowl and fish, produced using the MACS and methods provided herein. Methods for generating and isolating such chromosomes are provided. Methods using the MAGs to construct artificial chromosomes from other specie, such as iinsect, bird, fowl and fish species are also provided. T he artificial chromosomes are fully functional stable chromosomes. Two types of artificial chromosomes are provided. fine type, herein referred to as SATACs [satellite artificial chromosomes or satellite DNA based artificial chromosomes (the terms are used interchangeably herein)) are stable heterochromatic chromosomes, and the other type are minichromosomes based on amplification of euchromatin.
Artificial chromosomes provide an extra-genomic locus for targeted integration of megabase [Mb~ pair size DNA fragments that contain single or multiple genes, including multiple copies of a single gene operatively linked to one promoter or each copy or several copies linked to separate promoters. Thus, methods using the MACs to introduce the genes into cells, tissues, and animals, as well as species such as birds, fowl, fish ~s.
and plants, are also provided. The artificial chromosomes with integrated heterologous DNA may be used in methods of gene therapy, in methods of production of gene products, particularly products that require expression of multigenic biosynthetic pathways, and also are intended for delivery into the nuclei of gerrnline cells, such as embryo-derived stem cells (ES cells], for production of transgenic (non-hurriany animals, birds, fowl and fish. Transgenic plants, including monocots and dicots, are also contemplated .herein.
Mammalian artificial chromasomes provide extra-genomic specific integration sites for introduction of genes encoding proteins of interest and permi megabase size DNA integration so that, for example, genes encoding an entire metabolic pathway or a very large gene, such as the cystic fibrosis (CF; -- 250 kbl genomic C~NA gene, several genes, such as multiple genes encoding a series of antigens for preparation of a multivalent vaccine, can be stately introduced into a cell. Vectors for targeted introduction of such genes, including the tumor suppressor genes; such as p53, the cystic fibrosis transmembcane regulator cDNA
(CFTRI, and the genes for anti-HIV r'cbo~ymes, such as an anti-HIV gag ribozyme gene, into the artificial chramosomes are also provided.
The chromosomes provided herein are generated by introducing heteroiogous DNA that includes DNA encoding one or multiple selectable markers) into cells, preferably a stable cell line, growing the cells under selective conditions, and identifying from among the resulting clones those That include chromosomes with more than one centromere a~ndlbr . fragments thereof. The amplification that produces the additional centromere or centromeres occurs in cells that contain chromosomes in which the heterologous DNA has integrated near the centromere in the pericentric region of the chromosome. The selected clonal cells are then used to generate artificial chromosomes.

Although non-targeted introduction of DNA, which results in some frequency of integration into appropriate loci, targefad introduction is preferred. Hence, in preferred embodiments, the DNA with the Selectable marker that is introduced into cells to initiate generation of artificial chromosomes includes sequences that target it to the.an amplifiable region, such as the pericentric region, heterochromatin, and particularly rDNA of the. chromosome. For example, vectors, such as pTEMPUD and pHASPUD [provided herein], which include such DNA
specific for mouse satellite DNA and human satellite DNA, respectively, 1~ are provided. The plasmid pHASPUD is a derivative of pTEMPUD that contains human satellite DNA sequences that specifically target human chromosomes. Preferred targeting sequences include mammalian ribosomal RNA (rRNA) gene sequences (referred to herein as rDNA) which target the heterologous DNA to integrate into the rDNA region of 'l5 those chromosomes that contain rDNA. For example, nectars, such as pTERPUD, which include mouse rDNA, are provided. Upon integration into existing chromosomes in the cells, these vectors ca~i induce the amplification that results in generation of additional centromeres.
Artificial chromosomes are generated by culturing the cells.witi, 20 the multicentric, typically dicentric, chromosomes under conditions whereby the chromosome breaks to form a minichromosome and formerly dieentric chromosome. Among the MACS provided herein are the SATACs, which are primarily made up of repeating units of short satellite DNA and are nearly fully ,heterochrornatic, so that without 25 insertion of heterofogous or foreign DNA, the chromosomes preferably contain no genetic information or contain only non-protein-encoding gene sequences such as rDNA sequences. They can thus be used as "safe"
vectors far delivery of DNA to mammalian hosts because they do not contain any potentially harmful genes. The SATACs are generated, not from the minichromosome fragment as, for example; in U.S. Patent No.
6,288,625, but from the fragment of the formerly dlicentric chromosome.
In addition, methods for generating euchromatic minichromosomes and the use thereof are also provided herein. Methods for generating one type of MAC, the minichromosome, previously described in U.S.
Patent No. 5,288,625, and the use thereof for expression of heteroiogous DNA are provided. In a particular method provided herein for generating a MAC, such as a miriichromosome, heterologous DNA
that includes mammalian rDNA and one or more selectable marker genes is introduced into cells which are then grown under selective conditions.
Resulting cells that contain chromosomes with mvr~e than one centromere are selected and cultured under conditions whereby the chromosome breaks to form a minichromosome dr~d a formerly rraulticentric .4tYpically dicentric) chromosome from which the minichromosome was released.
Cell lines containing the minichromosome and the use thereof for cell fusion are also provided, fn one embodiment, a cell line containing the mammalian minichromosome is used as recipient cells for donor D'NA
encoding a selected gene or multiple genes. To facilitate integration of the donor DNA into the minichromosome, the recipient cell fine preferably contains the mini;chromosome but does not also contain the formerly dicentric chromosome. This may be accomplished by methods disclosed herein such as cell fusion and selection of cells that contain a minichromosome and no formerly dicentric chromosome. The donor DNA
is linked to a second selectable marker and is targeted to and integrated into the minichromosome. The resulting chromosome is transferred by calf fusion into an appropriate recipient calf line, such as a Chinese hamster cell line [CH~~o After large-scale production of the cells carrying the engineered chromosome, the chrornosome is isolated. In particular, metaphase chromosomes are obtained, such as by addition of colchicine, _g_ and they are purified from the cell lysate. These chromosames are used for cloning, seguencing and far delivery of heterologous DNA into cells.
Also provided are SATACs of various sizes that are wormed by repeated culturing under selective conditions and subcloning of cells that contain chromosomes produced from the formerly d~centric chromosomes. The exemplified SATACs are based on repeating DNA
units that are about 15 Mb [two --~.5 Mb blocks]. The repeating DNA
unit of SATACs formed from other species and other chromosomes may vary, but typically would be on the order of about 7 to about 20 Mb.
70 The repeating DNA units are referred to herein as megareplicons, which in the exemplified SATACs contain tandem blocks of satellite DNA
flanked by non-satellite DNA, including heterologous DNA and non-satelfite DNA. Amplification praduces an array of chromosome segments [each called an amplicon] that contain two inverted megareplicons '15 bordered by heterologous ("foreign'°) DNA: pepeated cell fusion, growth on selective medium and/or SrdtJ (5-bramodeoxyuridinie] treatment or other treatment with other genorr~e destabilizing reagent or agent, such as ionizing radihtion, including x-rays, and subctoning results in cell fines that carry stable heterochromatic or partially h~terachro~r~,atic 20 chromosomes, including a 150=200-Mb '°sausage" chromosome, a a00-1000 Mb gigachromosome, a stable X50-400 Mb rnegachromosome and various smelter stable chromosomes derived therefrom. ~Chese chromosomes are based on these repeating units and can include heterologous DNA that is expressed.
25 Thus, methods for producing MACS of both types 4i.e., SATACS
and minichromosomes) are provided_ These methods are, applicable 'to the production of artificial chromosomes containing cents°omeres derived from any higher eukaryotic veil, including mammals, birds, fowl, fish, insects and plants.

_10_ The resulting chromosomes can be purified by methods provided herein to provide vectors for introduction of heterologous DNA into selected cells for production of the gene product(s) encoded by the heterologous DNA, for production of transgenic (non-human) animals, birds, fowl, fish and plants or for gene therapy.
In addition, methods and vectors.for fragmenting the minichromosomes and SATACs are provided. Such ev~ethods and vectors can be used for in vivo generation of smaller stable artificial chromosomes. Vectors for chromosome fragmentation are used to produce an artificial chromosome that contains a megareplicon, a centromere and two telomeres and will be between about 7.5 Mb and about 60 Mb, preferably between about 10 Mb-15 Mb and ,30-50 Mb.
As exemplified herein, the preferred range is between about ~.5 Mb and 50 Mb. Such artificial chromosomes may also be produced by other 'k 5 methods.
lso4ation of the 15 Mb l:or 30 Mb amplicon containing two 15 Mb inverted repeats] or a 30 Mb or higher m~ItimEr. such as 50 Mb, thereof should provide a stable chromosomal vector that can be manipulated is~
vitro. Methods for reducing the siz~ of the MACS to generate smaller stable self-replicating artificial chromosomes are also provided.
Also provided herein, are methods for producing mammalian artificial chromosomes, including those provided herein, in vitro. and the resulting chromosomes. The methods involve in ~ritrQ assembly of the structural and functional elements to provide a stable artificial chramosome. Such elements include a centromeree two telomeres, at least one origin of replication and filler heterochromatin, e~a., satellite DNA. A selectable marker fior subsequent selection is also generally included. These specific DNA elements may be obtained from.the artificial chromosomes provided herein such as those that have .been generated by the introduction of heterologous DNA into cells and the subsequent amplification that leads to the artificial chromosori,e, particularly the SATACs. Centromere sequences for use in the in vitro construction of artificial chromosomes may also be obtained by , employing the centromere cloning methods provided herein. In preferred embodiments, the sequences providing the origin of replication, in particular, the megareplicator, are derived from rDiVA. These sequences preferably include the rDNA origin of replication and amplification promoting sequences. .
i0 Methods and vectors for targeting heterologous DNA.into the artificial chromosomes are also provided as are methods and vectors for fragmenting the chromosomes to produce smaller but stable and self-replicating artificial chromosomes.
'1-he chromosomes are introduced into cells to produce stable transformed cell lines or cells, depending upon the source of the.cellsa Introduction is effected by any suitable method including, but not limited to eiectro~oration, direct uptake, such as by calcium phosphate precipitation, uptake of isolated chromosomes by iipofection, by microcell fusion, by lipid-mediated carrier systems or other suitable method. The - resulting cells can be used for production of proteins in the cells. The chromosomes can be isolated and used for gene delivery. Methods for isolation of the chromosomes based on the DNA content of the chromosomes, which differs .in MACs versus the authentic chromosomes, are provided. >!llso provided are methods that rely on content, particularly density, and sine of the MACa..

These artificial chronnosomas can be used in gene therapy, gene product production systems, production of humanized genetically transformed animal organs, production of transgenic .plants and animals _ {non-human), including mammals, birds, fowl, fish, invertebrates, vertebrates, reptiles and insects, any organism or device that vrould employ chromosomat elements as information storage vehicles, and also for analysis and study of centromere function, for the production of artificial chromosome vectors that can be constructed in v' ro, and for the preparation of species-specific artificial chromosomes. The artificial chromosomes can be introduced into cells using microinjection, cel!
fusion, microeell fusion, electroporation; nuclear transfer, electrofusion, projectile bombardment; nuclear transfer, calcium phosphate.
precipitation, lipid-mediated transfer systems and other such methods.
Cells particularly suited for use enrith the artificial chromosomes include, but are not limited to plant cells, particularly tomato, ~arabidopsis, and others, insect cells, including silk warm cells, insect larvae, fish, reptiles, amphibians, arachnids, mammalian cells, avian cells, embryonic stem cells, haematopoietic sfiem cells, embryos and cells for use in rr~ethods of genetic therapy, such as lymphocytes that are used in methods of adop-tive immunotherapy and nerve dr neural cells. Thus methods of pro-ducing gene products and transgenic (non-human) animals and p[ants are provided. Also provided are the resulting transgenic animals and plants.
Exemplary cell lines that contain these chromosomes are also .
provided.
Methods for preparing artificial chrornosornes for particular species and for cloning centromeres are also provided. For example, two exemplary methods provided for ger9erating artificial chromosomes for use in different species aye as follov~rs. First, the methods herein may be applied to different species. Second, means for generating species--°I 3-specific artificial chromosomes and for cloning centromeres are provided.
In particular, a method for cloning a centromere from an animal or plant is provided by preparing a library of DNA fragments that contain the genome of the plant or animal and introducing ,each of the fragments into.a mamrna(ian satellite artificial chromosome [SATAC) that contains a centromere from a species, generally a mammal, different from the selected plant or animal, generally a non-mammal, and a selectable marker. The selected plant or animal 'ss one in which the mammalian species centromere does not function. Each of the SATACs is introduced into the cells, which are grown under selective conditions, arid cells with SATACs are identified. Such SATACS should contain a centromere encoded by the DIVA from .the library or should contain the necessary elements for stable replication in the selected species.
Also provided are libraries in which the relatively large fragments of DNA are contained on artificial chromosomes.
Transgenic [non-human) animals, 'invertebrates and vertebrates, plants and insects, fish, reptiles, amphibians, arachnids, birds, fowl, and mammals are also provided. Of particular interest are transgenic (non human) animals and plants that express genes that confer resistance or reduce susceptibility to disease. For example, the transgene may encode a protein that is toxic to a pathogen, such as a ~rirus, bacterium or pest, but that is not toxic to the trarzsgenic host. Furthermore, since multiple genes can be introduced on a MAC, a series of genes encoding an antigen can be introduced, which upon expression wiH serve to immunize [ire a manner similar to a multivalent vac:cinel 'the host anima( against the diseases for which exposure to the antigens provide immunity or some protection.

~l ~r°
Also of interest are transgenic (non-human) animals that serve as models of certain diseases and disorders for use in studying the iiisease and developing therapeutic treatments and cures thereof, Such animal models of disease express genes atypically carrying a disease-associated . ~ mutation], which are introduced into the animal on a MAC: and which induce the disease or disorder in the anirrtal. Similarly, MACs carrying genes encoding antisense RNA may be introduced into animal cells to generate conditional "knock-out" transgenic (non-human! animals. In such animals, expression of the antisense RNA results in decreased or complete elimination of the products of genes corresponding to the antisense RNA. Of further interest are transgenic mammals that harbor MAC-carried genes encoding therapeutic proteins. that are expressed in the animal's milk. Transgenic (non-hurx~an~ animals for use in xenotransplantation, which express MAC=carried genes that serve to 't 5 humanize the animal's organs, are also of interest. Genes that might be used in humanizing animal organs include those encoding human surface antigens.
Methods for cloning centfomeres, such as mammalian centromeres, are also provided. In particular, in one embodiment, a ~0 library composed of fragments of SATACs are cloned into YACs [yeast artificial chromosomes] that include a detectable marker, such as DNA
encoding tyrosinase; and then introduced into mammalian cells, such as albino mouse embryos. Mice produced from embryos containing such YACs that include a centromere that functions in rnammats will express 25 the detectable marker. Thus, if mice are produced from albino mouse embryos into which' a functional mammalian centromere was introduced, the mice will be pigmented or have regions of pign~tentation:

_'I 5_ A method for producing repeated tandem arrays of ~NA is provided. This method, exemplified herein using teiomeric DNA, is applicable to any repeat sequence, and in particular, low complexity repeats. The method provided herein for synthesis of arrays o~F tandem DNA repeats are based in a series of extension steps in which successive daublings of a sequence of repeats results in an exponential expansion of the array of tandem repeats. An embodiment of the method of synthesizing DNA fragri~ents containing tandem repeats may generally be described as follows. Two oligonucieotides are used as starting materials. 0ligonucleotide 1 is of length k of repeated sequence (the flanks of which are not relevant) and contains a relatively short stretch (60-90 nucleotides) of the. repeated sequence, flanked with appropriately chosen restriction sites:
5'-S1»»»»>»»»>»»>»»»S2 1.5 where S1 is restriction site 1 cleaved by E1, S2 is a second restriction site cleaved by E2 > represents a sirripie repeat unit, and '_' denotes a short (8-10) nucleotide flanking sequence complementary to oligonucfeotide 2:

3'- S3-5' where S3 is a third restriction site for enzyme E3 and which is present in the vector to be used during the .construction. Tl~~e method involves the fot4owing steps: (1 ) oiigonucleotides 1 and 2 are annealed; (2) the annealed ofigonucfeotides are filled-in to produce a double-stranded (ds) sequence; (3) the double-stranded DNA is cleaved -with restriction enzymes E1 and E3 and subsequently iigated into a vector (era., pUCl9 or a yeast vector) that has been cleaved with the same enzymes E1 and ES: (4! the insert is isolated from a first portion of the plasmid by digesting with restriction enzymes E1 and E3, and a second portion of the plasmid is cut with enzymes E2 (treated to remove the 3'-overhang) and E3, and the large fragment (plasmid DNA plus the insert) is isotated~
(5~ the two DNA fragments (the S1-S3 insert fragment and the vector plus insert) are ligated; and (6) steps 4. and 5 are repeated as many timi;s as needed to achieve the desired repeat sequence size. In each extension cycle, the repeat sequence size doubles, i.e., if m is the number of extension cycles, the size of the repeat sequence will be k x 2'~ nucleotides.
DESCRIPTION OF ThIE DRAVViNGS
Figure 1 is a schematic drawing depicting formation of the 90 MMCneo [the minichromosome~ chromosome. A'-G represents the successive events consistent with observed data that would Lead to the #ormation arid stabilization of the minichromosome,.
Figure 2 shows a schematic summary of the manner in which the observed new chromosorvies would form, and tha relationships among °l5 the different de nova formed chromosomes. !n particular; this figure shows a scherv~atic drawing of the de ncwo chromosome fofmatlon initiated in the centromeric region of mouse chromosome 7. (A) A single E-type amplification in the centromeric region of chromosome 7 generates a neo-centromere linked to the integrated "foreign" DNA, and 20 forms a dicentric chromosome. Multiple E-type arnpiification forms the ~i neo-chromosome, which separates from the remainder of mouse chromosome 7 through a specific breakage between the centromeres of the dicentric chromosome and which was stabilized in a mouse-hamster hybrid cell line; (B) Specific breakage between the centromeres of a 25 dicentric chromosome 7 generates a chromosome fragment with the neo-centromere, and a chromosome 7 with traces of heterologous DNA at the end; (Cp Inverted duplication of the fragment bearing the neo-ce.ntromere results in the formation of a stable ned-minichromosome; (D) Integration of exogenous ONA into the heterologous DNA region of the formerly dicentric chromosome 7 initiates H-type amplification, and the formation of a heterochromatic arm. By capturing a euchromatic terminal segment, this new chromosome arm is stabilized in the form of the "sausage" , chromosome; (E) BrdU f5-bromodeoxyuridine) treatment andlor drug selection induce further H-type amplification, rwhich results in the formation of an unstable gigachromosome: (F) Repeated BrdU treatments andlor drug selection induce further H-type amplification including a centromere duplication, which leads to the formation of another heterochromatic chromosome arm. It is split off from the chromosome 7 by chromosome breakage, and by acquiring a terminal segment, the stable megachromosome is formed:
Figure 3 is a schematic diagram of the repiicon structure and a scheme by which a megachromosome could be produced.
Figure 4 sets forth the relationships among some of the exemplary cell lines described herein.
Figure 5 is a diagram of the plasmid pTEMPU~.
DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS
Definitions Unless defined otherwise, all technical and scientific terms used herein have the same meaning as is commonly understood by one of skill in the art to which this invention belongs. All patents and publications referred to herein are incorporated by reference.
As used herein, a mammalian artificial chrorr~osome [MACl. is a piece of DNA that can stably replicate and segregate alongside endogenous chromosomes. It has the capacity to accommodate and express heteroiogous genes inserted therein. It is referred to as a mammalian artificial chromosome because it includes an active mammalian centromere(s). Plant artificial chromosomes, insect artificial chromosomes and avian artificial chromosomes refer to chromosomes that include plant and insect centromeres, respectively. A human artificial chromosome [HACK refers to chromosomes that include human centromeres, SUGACs refer to insect artificial chromosomes, and AI/d~:Cs refer to avian artificiaE chromosomes. Among the tVIACs provided herein are SATACs, minichromosomes, and in vitro synthesized artificial chromosomes. Methods for construction of each type are provided herein.
As used herein, in vitro synthesized artificial chromosomes are artificial chromosomes that is produced by joining the essential components f at least the centromere, and origins of replicationy in vitro.
As used herein; endogenous chromosomes refer to genomic chromosomes as found in the cell prior .to generation or introduction o~F
a MAC.
As used herein, stable maintenance of chromosames occurs when at least about 85%, preferably SCl%, more preferably 95%, of the cells retain the chromosome. Stability is measured in the presence of a selective agent. Preferably these chromosomes are also maintained in the absence of a selective agent. Stable chromosomes also retain their structure during cell culturing, suffering neither intrachromosomal nor interchromosomal rearrangements.
As used herein, growth under selective conditions means growth of a cell under conditions that require expression of a selectable marker for survival.
As used herein, an agent that destabilizes a~ chromosome is any agent known by those of skill in the art to enhance amplification events, mutations. Such agentsP which include Brdll, are well known to those of skill in the art.

"~! 9~
As used herein, de r~ovo with reference to a centromere, refers to generation of an excess centromere as a result of incorporation of a heterologous DNA fragment using the methods herein.
As used herein, euchromatin and heterochromatin have their recognized meanings, euchromatin refers to chromatin that stains diffusely and that typicalBy contains genes, and heterochromatin refers to chromatin that remains unusually condensed and that has been thought .
to be transcriptionaily inactive, Highly repetitive DNA sequences [satellite DNA]. at feast with respect to .mammalian cells, are usually 1 ~ faceted in regions of the heterochromatin surrounding the centromere [pericentric heterochromatinl. Constitutive heterochromatin refers to heterochromatin that contains the highly repetitive ~NA which is constitutively candensed and genetically inactive.
As used herein, BrdU refers to 5-bromodeoxyuridine, which during replication is inserted in place of thymidine. BrdU is used as a mutagen;; it also inhibits condensation of metaphase chromosomes during cell division.
As- used herein, a dicentric chromosome is a chromosome that contains two centromeres. A rnulticentric chromosome contains more than two centromeres.
As used herein. a formerly dicentric chromosome is a chromosome that is produced when a dicentric chromosome fragments and acquires new telomeres so that two chromosomes, each having one of the cent~oniares, are produced. 1~ach of the fragments are replicable chromosomes. If one of the chromosomes undergoes amplification of euchromatic DNA to produce a fully functional chromosome that contains the newly introduced heterofogous DNA:and primarily [at least more than 5~°l0~ euehromatin, it is a minichromosome. The remaining chromosome is a formerly dicentric chromosome. If one of the chromosomes _'~(~_ undergoes amplification, whereby heterochromatin [satellite DNA1 is amplified and a euchromatic portion [or arm] remains, it is referred to as a sausage chromosome. A chromosome that is substantially all heterochromatin, except for portions of heteroiogous DNA, is called a SATAC. Such chromosomes [SATACs] can be produced from sausage chromosomes by culturing the cell containing the sausage chromosome under conditions, such as BrdU treatment and/or grawth under selective conditi~ns, that destabilize the chromosome sc~ that a satellite artificial chromosomes ISATAC1 is produced. For purposes hereirg, it is 1~ understood that.SATACs may not necessarity be produced in multiple steps, but may appear after the initial introduction of.the heterologous DNA and growth under selectivb conditions, or they may appear after several cycles of growth under selective conditions and- E3rdU treatment.
As used herein, a SATAC refers to a chromosome that is 95 substantially all heterochromatin, except for portions of heteroiogous.
DNA. Typically, SATACs are satellite DNA based artificial chromosomes, but the term enompasses any chromosome made by the methods herein that contains more heterochromatin thaev euchromatin.
As used herein, ampliflable, where used in reference to a 2~ ~ chromosome, particularly flee method of generating SATACs provided herein, refers to a region of a chromosome that is prone to amplification.
Amptifcation typically occurs during replication and ~ther cellular events involving recombination. Such regions are typically regions of the chromosome that include tandem repeats, such as satellite DNA, rDNA
25 and other such sequences.
As used herein, amplification, with reference to DNA, is a process in which segments of DNA are dr~plicated to yield two or multiple copies of identical or nearly identical DNA segments that are typically joined as substantially tandem or successive repeats or inverted repeats.

As used herein an amplicon is a repeated DNA amplification unit that contains a set of inverted repeats of the megareplicon. A
megareplicon represents a higher order replication unit. For example, with reference to the SATACs, the megareplicon contains a set of a tandem DNA blocks each containing satellite DNA flanked by non-satellite DNA. Contained within the megareplicon is a primary replication site, referred to as the megareplicator, which may be involved in organizing and facilitating replicbtion of the pericenfiric heterachromatin and possibly the centromeres. Within the megareplicon ti-sere may be smaller [e.g., 50-300 kb in some mammalian cells] ;secondary replicons.
In the exemplified SATACS, the megarepiicon is defined by two tandem --7.5 Mb DNA blocks (see, e.a., Fig. 3l. Within each artificial chromosame (AC] or among a population thereof; each amplicon has the same gross structure but may.contain sequence variations: Such vaciatio.ns will arise as a result of movement of mobile genetic elements, deletions or insertions or mutations that arise, particularly in culture.
Such variation does not affect the use of the ACs or their overall structure as described herein.
As used herein, ribosomal RNA (rRNA] is the specialized RNA that forms part of the structure.of a ribosome and participates in the synthesis of proteins. Ribosomal RNA is produced by transcription of genes which, In eukaryotic cells, are present in multiple copies. In human cells; the approximately X50 copies of rRNA genes per haploid genome are spread out in clusters on at feast five different chromosomes ;chromosomes 13, 14, 1 b, 21 and 22). In mouse cells, the presence of ribosomal D~tA (rDNA) has been verified on at least 11 pairs out of 20 mouse chromosomes jchromosomes 5, 6, 9, 11, 1 ~, 15, 16, '17, 18, 19 and X](see e.g., Rowe et al. (1996) Mamm. Genome 7:886-889 and Johnson ~t al. (1993) ll/tamm. Genome 4:49-52]. 6n eukaryotic cells, the multiple copies of the highly conserved rRNA genes are located in a tandemly arranged series of rDNA units, which are generally about 40-45 kb in length and contain a transcribed region and a nantranscribed region known as spacer (i.e., intergenic spacer) DNA which can vary in length S and sequence. .In the human and mouse, these tandem arrays of rDNA
units are located adjacent to the pericentric satellite DNA sequences (heterachromatinl. The raglans of these chromosomes in which the rDNA is located are referred to as nucleolar organi~:ing regions (NOR) which loop into the nucleolus, the site of ribosome production within the cell nucleus.
As used herein, the minichromosome refers to a chromosome derived from a multicentric, typically dicentric, chromosome (see, e~,a., FIG. 1 ] that contains mare euchromatic than heterochromatic DNA.
As used herein, a megachromosome refers to a chromosor~ie that, except far introduced heterologous DNA, is substaa~tially composed of heterochromatin. Megachromosomes are made of an array of repeated amplicons that contain two inverted megarepiicons bordered by introduced heterologous DNA (see, e'u., Figure 3 for a schematic drawing of a megachromosome]. For purposes herein, a megachromosome is about 50 to 400 Mb, generally about 250-400 Mb.
Shorter variants are also referred t~ as truncated rnegachrornosomes (about 90 to 120 or 150 Mb].,, dwarf megachromosomes [ --150-200 Mb]
and cell lines, and a micro-megachromosome [--a0-90 Mb, typically' 50-60 Mb]. For purposes herein, the term megachromosome refers to the overall repeated structure based on an array of repeated chromosomal segments (amplicons] that contain two inverted megareplicons bordered by any inserted heterologous DNA. The size will ide specified.

°~.3_ As used herein, genetic therapy involves the transfer .or insertion of heterolagous DNA into certain cells, target cekls, to produce specific gene products that are involved in correcting or modulating disease. The DNA is introduced into the selected target cells in a manner such that the heterokogous DNA is expressed and a product encoded thereby is produced. Alternatively, the heterokogous DNA may in some manner mediate expression of DNA that encodes the therapeutic product. It may encode a product, such as a peptide or RNA, that in sorrie manner mediates, directly or indirectly; expressie~n of a therapeutic product.
~ 0 Genetic therapy may atsa be used to introduce therapeutic compounds, such as TNF, that are not normatky produced in the host ar that are not produced in therapeutically effective amounts or at a therapeutieatly useful time. Expression of the heterokogaus DNA by the target cells within an organism affkicted with the disease thereby enables modulation of the disease. The hsterologous DNA encoding the therapeutic product may be modified prior to introduction into the cells of the affkicted host in order to enhance or otherwise akter the product or expression thereof.
As used herein, heterologous or foreign DNA and RNA are used interchangeably and refer to DNA or RNA that does not accur naturally as part of the genome in which it is present or which is found in a location or locations in the genome that differ.from that in which it occurs in nature. tt is DNA ar RNA that is not endogenous to the cell and has been exogenousty introduced into the cell. Examples of heterologous DNA include, but are not limited to, IDNA that encodes a gene product or gene pr~duct(s) ~f interest, introduced for purposes of gene therapy or for production of an encoded protein. ether examples of heterologous DNA include, but are not limited to, DNA that encodes traceable marker proteins, such as a pratein that confers drug resistance, DNA that encodes therapeutically effective substances, such as anti's cancer agents, enzymes and hormones, and DNA that encodes other types of proteins, such as antibodies. Antibodies that are encoded bra heterologous DNA may be secreted or expressed'on the surface of the cell in which the heterologous DNA has been introduced.
As used herein, a therapeutically effective product is a product that is encoded by heterologous DNA that, upon intcoduction of the DNA
into a host, a product is expressed that effectively ameliorates or eliminates the symptoms, manifestations of an inherited or acquired disease or that cures said disease.
1 ~ As used herein, transgenic plants refer to plants in which heterologous or foreign DNA is expressed or in which the expression of a gene naturally present in the plant has been altered.
As used herein, operative linkage of heterologous DNA to regulatory and effector sequences of nucleotides, such as promoters, '95 enhancers, transcriptional and translationaV stop sites, and other signet sequences refers to the relationship between such DNA and such sequences of nucleotides. For example, operative linkage of heterologous DNA to a promoter refers to the physical relationship between the DNA and the promoter such that the transcription of such 2~ DNA is initiated from the promoter by an RNA polymerase that specifically recognizes, binds to and transcribes the DNA in reading frame. Preferred promaters include tissue specific promoters, such as mammary gland specific promoters, viral promoters, such TIC, CMV, adenovirus promoters, and other promoters 8cnown to those of skill in the 25 art.
As used herein, isolated, substantially pure. DNA refers to DNA
fragments purified according to standard techniques employed by those skilled in the art, such as that found in Maniatis et ai. [('i 9~2) Molecular -~5°
Cloning: A Laboratory Manual, Coid Spring Harbor Laboratory Press, Cold Spring Harbor, NY].
As used herein, expression refers to the process by which nucleic acid is transcribed into mRNA and translated into peptides, polypeptides, or proteins. If the nucleic acid is derived from genomic DNA, expression may, if an appropriate eukaryotic host cell or organism is selected, include splicing of the mRNA.
As used herein, vector or plasmid refers to discrete elements that are used to introduce heterologous DNA into cells for either expression of the heterologous DNA or for replication of the dloned heterologous DNA.
Selection and use of such vectors 'and plasmids are well within the level of skill of the art.
As used herein, transformationftra~sfection refers tp the process by which DNA or RNA is introduced into cell. Tra~nsfection refers to the taking up of exogenous nucleic acid, e.g., an enpression vector~ by a host cell whetheP or not any coding sequences are in fact expressed.
fJumerous methods of transfection are known to the ordinarily skilled artisan, for example, by direct uptake using calcium phosphate [CaP04;
see,. Wigler et al. (1979) Proc. Natl. Acad. Sci. U.S.A. 76:1373-1376], polyethylene glycol (PEGl-mediated DNA uptake, electr~poration, lipofection [see, e.~., Strauss 11996) Meth. Mol. Biol. X4.:307-327], microcell fusion [see, EXAMPLES, see, ~iso Lambent (1991) Proc,~,IVatl.
Acad. Sci. U.S.A. 88:59D7-5911; IJ.S. Patent No. 5,396,767, Sawford et al. ( 1980 Somatic Cell Mol. Genet. 13:279-284; Dhar et al. (1984) Somatic Cell Mol. Genet. 10:547-559; and McNeill-ICillary et al. (1995) Meth. Enzvmal. 254:133-152.], lipid-mediated carrier systems (see, e.g., Teifel et al. (1995) Biotechniaues 19:79-8~; Albrecht et al. (1996) Ann.
Hematol. x:73-79; Holmen et af. (1995) In Vitro hell Dev: Biol. Anim.
31:347-351; REmy et al. (1994) Bioconiua. Chem..5:647-654; Le Bolch _~~»
et al. (1995) Tetrahedron P.ett. 36:6681-6884; Loe~filer et al. (1993) Meth. Enzymol. 217:599-618] or other suitable method. Successful transfection is generally recognized by detection of the presence of the heterologous nucleic acid within the te~ansfected ceVl, such as any indication of the operation of a vector within the. host cell.
Transformation means introducing DNA into an organism so that the DNPo is repiicable, either as ari extradhromosoma( element or by chromosomal integration.
As used herein, injected refers to the microinjection (use of a small 1~ syringes of DNA into a cell.
As used herein, substantially homologous DPVA refers to DNA that includes a sequence of nucleotides that is sufficiently similar to another such sequence to form stable hybrids under specified conditions.
it is well known to those of skill in this art that nucleic acid 'P5 fragments with different sequences may, under th;s same. conditions, hybridize detectably to the same "target" nucleic acid. Two nucleic acid fragments hybridize detestably, under stringent conditions over a sufficiently long hybridization period, because one fragment contains a segment of at least about 14 nucleotides in a sequence which is ZO cornpfernentary (or nearly complementary] to the sequence of at least one segment in the other nucleic acid fragment. If the time during ~,rhich hybridization is allowed to occur is held constant, at a value during which, under preselected stringency conditions, two nucleic acid fragments with exactly complementary base-pairing segr~nents hybridize 25 detestably to each other, departures from exact complementarity can be introduced into the base-pairing segments, and base-pairing will nonetheless occur to an extent sufficient to make hybridization detectable. As the departure from complementarity between the base-pairing segments of two nucleic acids becomes larger, and as conditions of the hybridization becocv9e more stringent, the probability decreases that the two segments will hybridize detestably to each other.
Two single-stranded nucleic acid segments have °'substantially the same sequence," within the meaning of the present specification, if (a) both form a base-paired duplex with the same segment, arid (b) the melting temperatures of said two duplexes in a solution of 0.5 X SSPE
differ by less than lOoC. If the segments being compared have the same number of bases, then to have "substantially the same sequence'°, they will typically differ in their sequences at fewer than 1 base in 10.
70. Methods far determining melting temperatures of nucleic acid duplexes are welt known [see, e.a., Meinkoth and Wahl (1984) Anai. Biochem.
i 38:267-284 and references cited thereinj.
As used herein, a nucleic acid probe is.a DNA or RIVA fragment that includes a sufficient number of nucleotides to specifically hybridize to DNA or RNA that includes identical or closely related sequences of nucleotides. A probe may contain -any number of nucleotides, from as few as about 10 and as many as hundreds of thousands of nucleotides.
The conditions and protocols for such hybridization reactdons are wetl known to those of skill in the art as are the effects of probe size, temperature, degree of mismatch, salt concentration and other parameters an the hybridization reaction. For exar°nnple, the lower the temperature and higher the salt concentration at which the hybridization reaction is carried out, the greater the degree of mismatch that may be present in the hybrid molecules.
To be used as a hybridization probe, the nucleic acid is generally rendered detectable by iabe.lling it with a detectable moiety or (abet, such as 32P, 3H and'4C, or by other mearss, including chemical labelling, such as by nick-translation in the presence of deoxyuridyiate biotinylated at the 5'-position of the uracii moiety. The resulting probe includes the biotinylated uridylate in place of thymidyiate residues and can be detected [via the biotin moieties] by any of a number of c;ommerciaily, available detection systems based on binding of streptavidin to the biotin.
Such commercially available detection systems can be obtained, for . example, from Enzo Biochemicals, inc. (New York, NY1. Any other label known, to those ~f skill in the art, including non-radioactive labels, rTaay be used as long as it renders the probes sufficiently detectable, which is a function of the sensitivity of the assay, the time available [for culturing cells, extracting DNA, and f~ybridization assays], the quantity of DNA or 1 ~ RNA available as a source of the probe, the particular fabe6 and the means used to detect the label_ Once sequences with a sufficiently high degree of homology to the probe are identified, they can readily be isolated bar standard techniques, which are described, for example, by Maniatis et al. ((19S~m (Molecular Cloning: A Laboratory IVlanual, Cold Spring Harbor Laboratory Press, C~Id Spring Harbor, NY?.
As used herein, conditions undei which DNA molecules form stable. hybrids and are considered substantially homologous are such that DNA molecules with at least about CO°lo complementarily farm stable hybrids. Such DNA fragments are herein considered to be "substantially homologous". Far example, DNA that encodes a particular protein is substantially homologous to another DNA fragment if the DNA forms stable hybrids such that the sequences of the fragments are at least about 60% complementary and if a protein encoded by the DNA retains 2S its activity.
For purposes herein, the following stringency conditions are defined:
1? high stringency: 0.1 x SSPE, 0.1 % ADS, 65°C
2) medium stringency: 0.2 x SSPE, 0.1 °!o SDS, 50°C

3) low stringency: 1.0 x SSPE, ~.1 °/a SDS, SO°C
or any combination of salt and temperature and other reagents that result in selection of the same degree of mismatch or matching.
As used herein, immunoprotective refers. to~the ability of a vaccine b or exposure to an antigen or immunity-inducing agent, to confer upoh a host to whom the vaccine or antigen is administered or introduced, the ability to resist infection by a disease-causing pathogen or to have reduced symptoms. The selected antigen is typically an antigen that is presented by the pathogen.
1 ~ As used herein, all assays and procedures, SLlch as hybridization reactions and antibody-antigen reactions, unless otherwiscs specified, are conducted under conditions recognized by those of skill in the art as standard conditions. -A. Preparation of cell lines containing MACS
3 S 1. The megareptiaon The methods, cells and MACs.provided herein are produced by virtcse of the discovery of the existence of a higher-order replication unit (megareplicon] of the centromeric region. This megarepiicon is delimited by a primary replication initiation site (megareplicatar]; and appears to ~~ facilitate replication of the centromeric h.eterochronnatin, and most likely, centrorneres. Integration. of heterologous ONA into the megarepBicator region or in close proximity thereto, initiates a large-scale amplification of megabase-size chramosomal segments, which leads to de nova chromosome formation in living cells.
25 DNA sequences that provide a preferred megareplicator are the rDNA units that give rise to ribosomal RiVA (rRNAl. fn rnammais, particularly mice and humans, these rDNA units captain specialized elements, such as the origin of replication (or origin of bidirectional replication, j,.e_, OBR, in tnouse3 and amplification promoting sequences (APS) and amplification control elements (ACE) (see, e.g., Gogei et aL.
(1996) Chromosoma 104:511-518; Coffman et al. (1993) Exo. Cell. lees.
209:123-132; Little et at. (1993? Mol. Cell. Biol. l3:fi600-fi613; Yoon ~t al. (1995) Mol. Cell. Biol. 15:2482-2489; Gonzalez and Sylvester (1995) Genomics 27:320-328; Miesfeld and Arnheim (1982) Nuc. Acids Res.
10:3933-394911; Maden et at. (1987? Biochem. J. 246:519-527). 1 As described herein, without being bound by any theory, these specialized elements may facilitate replication and/or amplification of megabase-size chromosomal segments in the de nowo forrnai:ion of chromosomes, such as those described herein, in cells. These specialized elements are typically located in the nontranscribed intergerric spacer region upstream of the transcribed region oi~ rDNA. The intergenic spacer region may itself contain internally repeated sequences which can be .classified as tandemly repeated blocks and nontanderrs blocks (see e~a., Gonzalez and Sylvester ('1995) Genomics 27:320-328?. Iry mouse rDNA, an origin of bidirectional replication may be found within a 3-kb initiation zone centered approximately 1.B kb upstream of the transcription start site (see, e~a., Gogel et al. (1996) Chromosoma 104:51 1-5181. The sequences of these specialized elements tend to have an altered chrorrqatin structure, which inay be detected, for example, by nuclease hypersensitivity ~or the presence of AT-rich regions that can give rise to bent DNA structures. An exemplary sequence encompassing an origin of replication is shown in SEQ IG N0. 16 and in GENBANK accession no. X82564 at about positions 2430-5435.
Exemplary sequences encompassing amplification-promoting sequences include nucleotides 590-1060 and 1105-1530 of SEQ IG NO. 16.
in. human rDNA, a primary replication initiation site may be found a few kilobase pairs upstream of the transcribed region and secondary , initiation sites may be found throughout the nontranscribed intergenic .31.
spacer region (see, e.g., Yoon et al. (1395) Mol. Cell: Biol. 15:2482-2489). A complete human rDNA repeat unit is presented in GENBANK
as.accession no. U133fi9 and is set forth in SEQ, ID NO. 1 7 herein.
Rnather exemplary sequence encompassing a repiicdtion initiation site a may be found within the sequence of nucleotides 35355-42486 in SEQ iD N~. 17 particularly within the seguence of nucleotides 37912-42486 and more. particularly within the sequence o~ nucleotides 37912-39288 of SEQ ID NO. 17 (see Coffman ~ al: (1993) Exo. Cell. es.
209°.123-132).
Celf lines containing MACS can be prepared by transforming cells, preferably a stable cell fine, with a heterologous DNA fragment that encodes a selectable marker, culturing under selective conditions, and identifying cells that have a multicentric, typically dicentric, chromosome.
These cells can then be manipulated as described herein to produce the minichromosomes and other MACs, particularly the heterochromatic SATACs, as described herein. , Development of a multicentric, particularly dicentric, chromosame typically is effected through integratian of the heterologous DNA in the pericentric heterochromatin, preferably in the centromeric regions of chromosomes carrying rDNA sequences. Thus, the frequency of incorporation can be increased by targeting to these regions, such as by including DNA, including, but not limited to, rDNA or satellite DNA, in the heterologous fragment that encodes the selectable marker. Among the preferred targeting sequences for directing the heteroiogous DNA to the pericentrameric heterochromatin are rDNA sequences that target centromeric regions of chromosomes that carry rRNA genes. Such sequences include, but are not limited to, the DNA of SECT iD NO. 1 fi and GENBANK accession no. X82564 and portions thereof, the DNA of SEQ
ID N~. 17 and GENBANK accession no. U13369 and portions thereof, and the DNA of SEQ fD NOS. 18-24. A particular vector incorporating DNA from within SEC2 ID N0. 15 for use in directing integration of heter~lvgous DNA into chromosomal rDNA is pTERPUD (see Example 12). Satellite DNA sequences can also be used to direct the heterologous DNA to integrate into the pericentric heterochromatin. For example, vectors pTENIPUD and pHASF'UD, which contain mouse and human satellite DNA; respectively, are provided herein øsee Example 12) as exemplary vectors for introduction of heterologvus DNA into, ce(!s for de novo artificial chrofnosome formation.
1 ~! The resulting cell lines can then be treated as the exemplifiec! .cells herein to produce cells in which the dicentric chromosome has fragmented. The cells can then be used to introduce additional selective markers into the fragmented dicentric chromosome (i.e., formerly dicentric chromosome), whereby amplification of t:he pericentric heterochromatin will produce the heterochromatic, chromosomes.
The following discussion describes this process with reference fo the EC3/7 lfne and the resulting cells. '1°he same procedures can be applied to any other cells, particularly cell Ifnes to create SATACs and euchromatic rninichror~iosdmes:
2~ 2. Formation of de novo chromos~nrees ~e n~d~ centrome~re formation in a transformed mouse I.MTK-fibroblast cell line [EC3d~l after cointegratiazn of ~ constructs , [~ICMB and ~toJVESneo~ carrying human and bacteriat.D-NA [Hadfac~ky et af. ( 1991 ) Rc'oc. Natf. Acad. Sci. U.S.A. 88:8106-81 'f ~ .
has been shovivn. Tlle integration of the "hete.rofogous" engineered human,. bacteria! and phage DNA, and the subsequerit amplification of mouse arid heterologous bIVA that led to the formation of a dfcentric chromosome, occurred at the centrvmeric region of the short arm of a mouse chromosome. 8y G-banding, this chromosome was identified as mouse chromosome 7. Because of the presence of two functionally active centromeres on the same chromosome,.~regular breakages occur between the centromeres. Such specific chromosome breakages gave rise to the appearance [in approximately 10% of the celBs] of a chromosome fragment carrying the neo-centromere. Erom the EC3/'~ cell line [see, U.S. Patent No.
5,288.525, deposited at the European C~Iiection of Animal Cel( Culture ' (hereinafter ECACC[ under accession no. 90051001; see, also Hadla~zky et aI. ~1991~ E'roc. Natl. Arad. Sci. tJ.S.A. 88:8108-8110 t 0 : . and the corresponding published European sppiicat'con EP 0 473 258, two sublines (EC3/7C5 and EC3/7C6),were selected by repeated single-cell cloning. In these veil lines, the neo-centromere was found exclusively on a minichromosome [neo-miniehromosome], while the formerly dicentric chromosome carried '! 5 tr aces of "heteralog.ous" DNA.
It has now been discovered that integration of DNA. encoding a -setectabie marker in the heterochromatic region of the cer~tromere led to formation of the dicentric chromosome.
8> Tt~e neo-minichrornosmane 20 The chromosome breakage in the EC3/7 cell s, which separates the neo-centromere from the mouse chromosome, occurred ir, the G-band positive '°heteroiogous" DNA region. This is supported by the observation of traces of a and human C7NA sequerices at the broken end of the formerly ciicentric chromosome. Comparing the G-hand pattern of the 25 chromosome fragment carrying the neo-centromere with that of the stable neo-miniahromosome, it is apparent that the neo-minichromosome i.s an inverted duplicate of the chromosome fragment that bears the neo-centromere. This fs supported by the observation that although the neo-rninichromc~some carries only -one functional centr~~mere, both ends of the minichromosome are heterochromatic, and mouse satellite DNA
sequences were found in these heterochromatic regions by in situ hybridization.
Mouse cells containing the minichromosorne, ~rhic)~ contains multiple repeats of the heterologous DNA, wf~ich in the exemplified embodiment is of DNA and the neomycin-resistance gene, can be used as recipient cells in cell transformation. Donor DNA, such a:; selected heterologous DNA containing i1 DNA linked to a second selectable marker, such as the gene encoding hygromycin phosphotransferase 1 ~ which confers hygromycin resistance [hygl, can be introduced into the mouse cells and integrated into the minichromosornes by horvoiogous recombination of ~i DNA in the donor DNA with that in the minichromosomes. Integration is derified by in situ hybridization and Sauthern blot analyses. Transcription and translation of the heterologous DNA is confirmed by primer extension and immunoblot analyses.
For example, DNA has been targeted into the neo-minichromosor~ie in EC3/7C5 Celts using a ~ DNA-containing construct (pNem1 ruc] that also contains DNA encoding hygromycin resistance and the l3enilla luciferase gene linked to a promoter, such as the cytomegalovirus [CMif]
~4 early promoter, and the bacterial neomycin resistance-encoding DNA.
Integration of the donor DNA into the chromosome in selected cells [designated PHN41 was confirmed by nucleic acid ampiifacation [PCR] and in sltca hybridization. Events that would produce a neo-minichromosome are depicted .in Figure 1.
The resulting engineered minichromosome that contains the heterol~gous DNA can then be transferred by cell fusion into a recipient cell line, such as Chinese hamster ovary cells [CHC] and correct expression of the heterologous DNA can be verified. Following production of the cells, metaphase chromosomes are obtained, such as by addition of colchicine, and the chromosomes purified by addition of AT- and GC-specific dyes on a dual laser beam based cell sorter (see Example 10 B for a description of methods of isolating artificial chrornomsomesl., Preparative amounts of chromosoa~r~es [5 x 10~ - 5 x 10' chromosomes/m!] at a purity of 95a/o or higher can be obtained. The resulting chromosomes are used for delivery to cells by methods such as microinjection and liposome-mediated transfer.
Thus, the neo-minichromosome is stabiy maintained in cells, replicates autonomously~ and permits the persistent long-term expression of the neo gene under non-selective culture conditions. !t also contains megabases of heteroiogous known DNA [~t DNA in the exemplified -embodimentsl that serves as target sites for homologous recombinbtion and integration of DNA of interest.. The neo-minichre~mosome is, thus, a vector for genetic engineering of cells. It has been introduced into SCID
mice, and shown to. replicate in the same manner as endogenous .
chromosomes.
The methods herein provide means to induce the events that lead to formation of the neo-minichromosome by introducing heterofogous DNA with a selective marker Ipreferably a dominant selectable marker]
into cells and culturing the cells under selective conditions. As a resutt, cells that contain a multicentric; e.g., dicentric chromosome, or fragments thereof, generated by amplification ace produced. Cells with the dicentric chromosome can then be treated to ,destabilize the chromosomes with agents, such as BrdU andlor culturing under selective conditions, resulting in cells in which the dicentric chromosome has formed two chromosomes, a so-caned minichromosome, and a formerly dicentric chromosome that has typically undergone ampkification in the a heterochromatin where the heterologous DNA has integ~°ated to produce a SATAC or a sausage chromosome tdiscussed below]. These cells can be fused with other cells to separate the minichromosome from the formerly dicentric chromosome into different cells_so that each type of MAC can be manipulated separately.
Preparation of SATA~s An exemplary protocol for preparation of SATACs is illustrated in Figure 2 [particularly D, E and F] and FIGURE 3 (see, also the EXAMPLES, particularly EXAMPLES ~.-7].
To prepare a SATAC, the starting materials are cells, preferably a stable cell line, such as a fibrobiast cell line, and a DNA fragment that '10 includes DNA that.encodes a selective marker. The DNA fragment is introduced into the cell by r~aethods of DNA transfer, inclcading but not limited to direct uptake using calcium phosphate, etectroporatioh, and lipid-mediated transfer. To insure integration of the DNA fragment in the heterochromatin, it is preferable to start with DNA that will be targeted 1 a to the pericentric heterochromatic region of the chromosome, such as ~tCMB and vectors provided herein, such as pTEl1lIPIJD [Figure b] and pHASPUD $see Example ~ 2) that include sdteilite DNA, or specifically into r~NA in the centromeric regions of chromoso~a~es containing rDNA
sequences. After introduction of the QNA, the cells are grown under 20 selective conditions. The resulting cells are examined arvd any that have multicentric, particularly dicentric, chromosomes [br heterochromatic chromosorvtes or sausage chromosomes or other such st:ructure~ see, Figure 2D, 2E and 2F] are selected.
tn particular, if a cell with a dicentric chromiosome is selected, it ~5 can be grown under selective conditions, or, preferably, additional DNA
encoding a second selectable marker is introduced, and the cells grown under conditions selective for the second marker. The resulting cells should include chromosomes that have structures similar to those depicted in Figures ZD, ~E, 2F. Cells with a structure, such as the sausage chromosome, Figure 2D, can be selected and fused with a second cell line to eliminate other chromosomes that are mot of interest.
If desired, cells with other chromosomes can be selected and treated as described herein. If a cell with a sausage chromosome is selected, it can be treated with an agent; such as BrdU, that destabilizes t:he chromosome so that the heterochrorcoatic arm forms a chromosome that is substantially heterochromatic fi.e., a rritegachromosome, see, Figure 2F]. Structures such as the gigachromsome in which the heterochromatic arm has amplified but not broken off from th'e euchromatic arm, well also be observed. The megachromosome is a stable chromosome. Further manipulatian, such as fusions and growth in selective conditions and/or- BrdU treatment or other such treatment, can lead to fragmentation of the rnegachromosome to form srnalier chromosomes that have the amplicon as the basic repeating unito The megachromosome can be further fragmented i~ viva using a chromosome fragmentation vector, such as pTEMF'UD Esee, Figure 5 and EXAMPLE 121, pH6~oSPUD or pTERPUD tsee Example 1 ~1 to ultimately produce a chrornasome that comprises a smaller sl:able replicable unit, about 15 Mb-60 Mb; containing one to four megarepticons.
Thus, the stable chromosomes formed de nc~v~ that originate from the short arm of mouse chromosome 7 have been analyxed. This chromosome region shows a capacity for amplification of large chromosome segments, and promotes de nova chromosome formation.
Large-scale amplification at the same chromosome region leads to the formation of dicentric and multicentric chromosomes, a cwinichromosome, the 150-200 Mb size rl neo-chromosome, the °'sausage'°
chromosome, the 500-1000 Mb gigachromosome, and the stable 250-X00 Mb megachromosome. .

-3g-A clear segmentation is observed along the arms of the megachromosome, and analyses show that the building units of this chromosome are amplicons of -30 Mb composed of mouse major satellite DNA with the integrated "foreign" DNA sequence's at both ends.
The - 30 Mb amplicoris are composed of two ~ 7 b Mb inverted doublets of -7.5 Mb mouse major satellite DNA blocks, which are separated from each other by a narrow band of non-satellite sequences (see, era., Figure 3]. The wider non-satellite regions at the amplicon borders contain integrated, exogenous f heterologousl DNA, vsihile the narrow 90 hands of non-satellite DNA sequences within the amplicons are integral parts of the pericentric heterochromatin of mouse chromosomes. These results indicate that the -7.5 Mb blocks flanked by non-satellite DNA
are the building units of the pericentric heterochrornatin cif mouse chromosomes, and the -15 Mb size pericentric regions of mouse chromosomes contain two - 7.5 Mb units.
Apart from the euchromatic terminal segments, the whose megachromosome is heterochromatic, and has structural homogeneity.
Therefore,,this large chromosome offers a unique possibility for obtaining information about the amplification process, and for andtyzing some basic characteristics of the pericentric constitutive heterochrornatin, as a vector for heterologous DNA, and as a target for further fragmentation.
As shown herein, this phenomenon is generalizabie and can be observed with other chromosomes. Also, although these de novo formed chromosome segments and chromosomes appear different, there are similarities that indicate that a similar amplificafion mechanism plays a role in their formation: (i) in each case, the amplification is initiated in the centromeric region of the mouse chromosomes and Large ~Mb size]
amplicons a.re formed; (ill mouse major satellite DNA sequences are constant constituents of.the arnpticonst either by providing the bulk of the heterochromatic amplicons [H-type amplification], or by bordering the aeuchromatic amplicons CE-type amplificationJ~ (iii) formatian of inverted segments can be demonstrated in the a neo-chromosome and megachromosome; (iv) chromosome arms and chromosomes formed by 6 the amplification are stable and functional.
The presence of inverted chromosame segments seems to be a common phenomenon in the chromosomes formed de raova~ at the centromeric region of mouse chromosome 7. During the formation of the neo-minichromosome, the event leading to the stabilization of the distal 1~ segment of mouse chromosome ~ that bears the neo-centromere may have been the formation of its inverted duplicate. Amplicons of the megachromosome are inverted doublets of --7.5 IVIb mouse major satellite DNA blocks.
5. Cell lines 15 Cel! lines that contain MACs, such as the min.ichrottiosome, the a-neo chromosome, and the SATACs are provided herein or can be produced by the methods herein. Such cell lines provide a convenient source of these chromosomes and can be manipulated, such as by cell fusion or production of microcells for fusion with selected ce91 lines, to 20 deliver the chromosome of interest into hybrid cell lines. Exemplary yell lines are described herein and some have been deposited with the ECACC.
a. EC3I7C5 and EC317C6 Cell lines EC3/7C5 and EC3/7C6 were produced by single cell 25 cloning of EC3/7. For exemplary purposes EC317C5 has been deposited with the ECACC. These cell lines contain a minichromosome and the formerly dicentric chromosome from EC~l7. The stable mini-chromosomes in cell lines EC3/'7C5 and EC3/7C6 appear to be the same and they seem to be duplicated derivatives of the ---10-15 Mb "broken-off°' fragment of the dicentric chromosome. Their similar size in these independently generated cell tines might indicate that --20-30 (Vib is the miwimal or close to the minimal physical size for a stable minichromosome:
6 ' b. T~F100~G19 Introduction of additions! heterotogous DNA, including DNA
encoding a second selectable marker, hygrornycin .phosphotransferase, i.e:, the hygromycin-resistance gene, anti. also a detectable marker, ji-galactosidase (i.e., encoded by the IacZ gene9, into the EC3!?'C5 cell line and growth under selective conditions produced cells designated TF 10046 19. In particular, this cell line was produced from the EC3/7C5 cell tine by cotransfection with pPasrrDids pH132, which contains an anti-HHV ribozyme and hygrarnycin-resistance gene, pCH110 (encodes ~-gaiactosidase] and .i phage (etci 875 Sam 7] DNA and selection with hygromycin B, Detailed analysis of the 1'F1~44G19 cell line by in situ hybridization with R phage and plasmid DNA sequences re~realed the formation of the sausage chromosome. The formerly dicentric chromosome of the EC3hC5 cell line translocated to the end of another acrocentric chromosome. The heterologous DNA integrated into the pericentric heterochromatin of the formerly dicentric chromosome and is amplified several times with megabases of mouse pericentric heterochrornatic satellite DNA sequences tFig. 2D] forming the "sausage"
chromosome. Subsequently the acroce.ntric mouse chromosome was substituted by a euchromatic telamere.
!n situ hybridization with biotin-labeled subfragments of the hygromycin-resistance and ,B-galactosidase genes resulted in a hybridization signal only in the heterochrornatic arm of the sausage -~,'( »
chromosome, indicating that in TF1004G19 transformant cells these genes are localized in the pericentric heterochromatin.
A high level of gene expression, however, was detected. In general, heterochromatin has a silencing effect in ~rosophila, yeast and on the HSV-tk gene introduced into satellite ~NA at the mouse centromere. Thus, .it was of interest to study the TF1004G19 transformed cell line to confirm that genes located in the heterochromatin were. indeed expressed, contrary to recognized dogma.
For this purpose, subciones of TF1004G19, containing_a different 1~ sausage chromosome (see Figure 2D1° were established by single cell cloning. . Southern hybridization of DNA isolated frorxz the subclones with subfragments of hygromycin phosphotransferase and lacZ: genes showed a close correlation between the intensity of hybridization and the length of the sausage chromosome. This finding supports the conclusion that these genes are localized in the heterochromatic arm of the .sausage chromosome.
( 1 ) T°F 10046-1905 TF1004G-1905 is a mouse LP~iTK fibroblast cell line containing neo-minichromosomes and stable "sausage°' chromosomes. It is a subclone. of TF10~4619 and was generated by single-cell cloning of the TF1004G19 cell line. It has been deposited with the ECACC as an exemplary cell line and exemplary source of a sausage cl°'romosome.
Subsequent fusion of this cell line with CH~ 1C20 ,cells and selection with hygromycin and 6418 and HAT (hypoxanthine, aminopteria, .and thymidine medium; see Szybaiski et al. (1962) Proc. Natl. Acad. Sci.
48:2026) resulted in hybrid cells (designated 19C5xHa4) that carry the sausage chromosome and the neo-minichromosome. BrdU treatme~it of the hybrid cells, followed by single cell cloning and selection with 6418 andlor hygromycin produced various cells 'Chat carry chrorr7osomes of interest, including GB43 and G3D5.
t2) other subctones Cell tines GB43 and G3D5 were obtained by treating 19C5xHa4.
cells with BrdU followed by growth in G4~18-containing selective medium and retreatment with BrdU. The two cell lines were isolated by single .
cell cloning of the selected cells. GB43 cells contain the neo-minichromasame only. G3D5, which has been deposited with the ECACC, carries the neo-minichromosome and the megach~~omosorne.
Single cell cloning of this celE line followed by growth of the subclones in 6418- and hygromycin-containing medium yielded subclones such as the GHB42 cell line carrying the neo-minichromosome and the megachromosome. H1 D3 is a mouse-hamster hybrid cell line carrying the megachromosome, but no neQ-minichromosome, and ~nras generated by treating 19C5xHa4 calls with BrdU follosrrved by growth in hygrornycin-containing selective medium and single cell subcloning of selected cells.
Fusion of this cell line with the CD4* HeLa cell line that also carries DNA
encoding an additional selection gene, the neomycin-resistance gene, produced cells [designated H~1 xHE41. cells] that carry the 2C1 megachromosome as well as a hurraan chromosome that carries CD4.neo.
Further BrdU treatment and single .cell cloning produced cell tines, such as.1 B3, that include cells with a truncated ~egachromosome.
5. D11EA constructs used to twansform the cells Heterofogcaus DIVA can be introduced into the cells by transfection or other suitable method at any stage during preparation of the chromosomes [see, e~~., F1G. 4j. In general, incorporation of such DNA
into the MACs is assured through site-directed integration, such as may be accomplished by inclusion of ~t-DNA in the heterotogous DNA (for the exemplified chromosomes, and else an additional selective marker gene.

-4 ~-For example, cells containing a MAC, such as the rninichromasome or a SATAC, can be cotransfected with a plasmid carrying the desired heterologous ONA, such as ONA encoding an HIV ribozyme, the cystic fibrosis gene, and DNA .encoding a second selectable marker, such as hygromycin resistance. . Selective pressure is then applied to the cells by exposing them to an agent that is harmful to Celts that do not express the new selectable rvaarker, tn this manner, cells that include the heteralagous DNA in the MAC are identified. Fusion with a second cell 4ine~ can provide 'a means. to produce Celt tines that contain one particular type of chromosomal structure or MAC.
Various vectors far this purpose are provided herein (see, Examples) and others can be readily constructed. The vectors preferably include DNA that is homologous to DNA contained. within a MAC in ~rder to target the DNA to the MAC for integration therein. The vectors also include a selectable marker gene and the selected heteroiog~us genes) of interest. Based on the disclosure herein and the knowledge of the skilled artisan, one of skill can construct such vectors.
Of particular interest herein is the vector pTEMPUO and derivatives thereof that can target DNA into the heterochromatic region of selected chromosomes. These vectors can also serve as fragmentation vectors (see, e~,a., Example 121.
Heterologous genes of interest include any gene that encodes a therapeutic product and DNA encoding gene products of interest. These genes and DNA include, but are not limited to: the cystic fibrosis gene [CFI, the cystic fibrosis transmembrane regulator (CF'TR) gene [see, ela., U.S: Patent No. 5,240,846; Rosenfeld et al. (1992) Cell, 6~F :14.3-755;
Hyde et al. (19931 Nature 362: 250-255; Kerem ~t al. (1989) Science 245:1073-1080; Riordan et ~l.(1989) Science 245:1065-1072;
Rommens et at. (1989) Science 245:1t~59-1065; ~sborr~e et al. (1991) Am. J. Hum. Genetics 48:6059-5122; lhJhite et al. X19903 Nature 344:665-667; Dean et al. (1990) Celi ~1:$B3-870; Erlich et al. (1991) Science 252:1543; and ~.S. Patent Nos. 5,453,357, 5,x'49,604, 5,434,086, and 5,240,546, which provides a retroviral vector encoding the normal CFTR gene).
~. Isolation of artificial chrornosorUaes The MACs provided herein can be isolated by any suitable method known to those of skill in the art. Also, methods are provided herein for effecting substantial purification, particularly of the SATACs. SATACs have been isolated by fluorescence-activated cell sorting (FAGS]. This ri~ethod takes advantage of the nucleotide base content of the SATACs, which, by virtue of their high heterochromatic DNA content, will differ from any other chromosomes in a cell. !n particular embodiment, metaphase chromosomes are isolated and stained with base-specific dyes, such as Hoechst 33258 and chromomycin A,3. 1=luorescence-activated cell sorting will separate the SATACs from the endogenocs~
chromosomes. A dual-laser cell sorter ~FACS Vantage Becton Dickinson Immunocytometry Systems) in which two lasers were set to excite the dyes se~parataly, allowed a bivariate analysis of the chromosomes by 2~ base-pair composition and size. Cells containing such SATACs can be Similarly sorted.
Additional methods provided herein for isolation of artificial chromosomes from endogenous chromosomes include procedures that are particularly well suited for large-scale isoiatiorv of artificial chromosomes such as SATACs. In these methods, the size and density differences between SATACs and endogenous chromosomes are exploited to effect separation of these two types of chromosomes. Such methods involve techniques such as swinging bucket centrifugation, tonal rotor centrifugation, and velocity' sedimentation. Affinity-, ~L~L°J~
particularly immunoaffinity-; based methods for separation of artificial from endogenous chromosomes are also provided herein.. For example, aATACs, which are predominantly heterochromatin, may be separated from endogenous chromosomes through immunoaffinity procedures involving antibodies that specif(cally recognize heterochromatin, and/or the proteins associated therewith, when the endogenous chromosomes contain 'elatively little heterochromatin, such as (n hamster cells.
C. In vitro construction of artificial chromosomes Artificial chromosomes can be constructed in vitro by assemb6ing 1~ the structural and functional elements that contribute to a compiete chromosome capable of stable replication and segregation alongside endogenous chromosomes in cells. The identification of the discrete elements that in combination yield a functional chromosome has made possible the in vitro generation of artificial chromosomes. The process of in vitro construction of artificial chromosomes, which can be rigidly controlled, provides advantages that may be desired in the generation of chromosomes that, for example, are required in large as~ounts or that are intended for specific use in transgenic animal systems.
For example, irs vitro construction may be advantageous when efficiency of time and .scale ar~ important considerations in the preparation of artificial chromosomes. Because in vitro construction methods do not involve extensive cell culture procedures, they may be utilized when the time and labor required to transform, feed, cultivate, and harvest cells used in in viva ,cell-based production systems is unavailable.
In vitro construction may also be rigorously controlled with respect to the exact manner in which the several elbmeni;s of the desired artificial chromosome are combined and in what sequence and proportions they are assembled to yield a chromosome of precise specifications. These aspects may be of significance in the production of artificial chromosomes that.will be used in live animals where it is desirable to be certain.that only very pure and specific DNA sequences in specific amounts are being intraduced into the host animal.
The following describes the processes involved in the construction of artificial chromosomes in vitro, utilizing a megachromosome as exemplary starting material<
1. Identification and isolation of the componebts of the artificial chromosome °90 The MACS provided herein, particularly the SATA~a, are elegantly simple chromosomes for u_se in the identification and isolation of COmpOnentS t0 be Used in the i~ vitro construction of artBfiCla!
chromosomes. The ability to purify MACS to a very higl-a level of purity, as described herein, facilitates their use for these purposes. For example, the megachromosome, particularly truncated forms thereof Ci.e.
cell lines, such as 1 B3 end mM2C1, which are derived from H 1 D3 (deposited at the European Collection of Animal Cell Culture (ECACC) under Accession No. J6~3~.0929, see EXAMPLES below) serve as starting materials.
20, For example, the mM2C1 cal! line contains a miceo-megachromosome ( - 5Q-6~ kB), which advantageously contains only one centromere, two regions of integrated heterologous DN,A with adjacent rDNA sequences, with the remainder of the chromosomal DNA being mouse major satellite DNA. Dther truncated megachromosomes can serve as a source of telomeres, or telomeres can bd provided (see, Examples below regarding construction ~f piasmids corataining tandernly repeated teiomeric sequences). The centromere of the mfVl2Ci cell line contains mouse minor satellite DNA, 4nrhich provides a useful tag for isolation of the centromeric DNA.

Additional features of particular SATACs provided herein, such as the micro-megachromosome of the mM2C1 cell line, that make them uniquely suited to serve as starting rhaterials in the isolation and.
identification of chromosomal components include the fact that the centromeres of each megachromosome within a single specific cell Dine are identical. The ability 'to begin with a homogeneous centromere source (as opposed to a mixture of different chromosome having differing centromeric sequences) greatly facilitates the cloning of the centromere DNA. By digesting purified megachromosomes, particularly ~ 0 truncated megachramosomes, such as the micro-megachrorrao so me, with appropriate restriction endonucleases and cloning the fragments into the commercially available and well known YAC vectors (see, e.ca., Burke et al. ('f 987) Science 236:806-87 2), BAC vectors (see, e_.cg., Shizuya et al.
(1992) Proc. Natl. Acad. Sci. U.S.A. 88~9: 8794-8797 bacterial artificial chromosomes which have a capacity of incorporating 0.9 - 1 Mb ~f ~NA) or PAC vectors (the P1 artificial chromosome vector which is a P1 piasmid derivative that has a capacity of incorporating 300 kb of DNA
and that.is delivered to ~. c~li host cells by electroporatiron rather than by bacteriophage packaging; see, e.~.; loannou et al., (1994) Nature Genetics 6:84-89; Pierce et a1. 11992) Meth. Enz~rrsol. 216:54.9-574.;
Pierce et a!. (1992) Proc. Natl. Acad. Sci. U.S.A. X9:2056-2060; U,S.
Patent No. 5,300,4.31 arid international PCT application No.
WO 92f 14819) vectors, it is possible for as few as 50 clones to represent the entire micro-megachromosome.
a. Centromeres An exemplary centromere for use in the construction of a mammalian artificial chromosome is that contained within the megachromosome of any of the megachrorraosome-containing cell lines provided herein, such as, for example, H1 ~3 and derivatives thereof, -4~~
Such as mM2C1 cells. Megachromosomes are isolated from such cell lines wtilizing, for example, the procedures described herein, and the centromeric sequence is extracted from the isolated megachromosomes.
For example, the rnegachromosomes may be separated into fragments utilizing selected restriction endanucleases that recognize. and cut at sites that, for instance, are primarily located in the replication andlor heterologous DNA integration sizes andlor in the satellite DNIa. Based on the sizes of the resulting fragments, certain undesired elements may be separated from the centromere-containing sequences. The centromere=
10. containing DNA, which could be as large as 1 Mb.
Probes that specifically recognize the centromeric sequences, such as mouse minor satellite DNA-based probes Esee, e~4., gong et al.
( 1988) Nucl. Acids Res. 1 x:11645-1.1661 ], may be used to isolate the centromere-containing YAC, BAC or PAC clones derived from the megachromosome. Alternatively, or in conjunction with the direct identification of centromere-containing megachromosomal DNA, probes .
that specifically recognize the non-centromeric elementsn such as probes specific for mouse major satellite DNA, the heterologous DNA andlor rDNA, may be used to identify and eliminate the non-centromeric DNA-containing clones.
Additionally, centromere cloning methods described herein may be utilized to isolate the centromere-containing sequence of the megachromosome. For example, Example 7 2 desoribes the use of YAG
vectors in combination with the murine tyrosinase gene and NMRI/lian mice for identification of the centromeric sequence.
Once the centrom~re fragment has been isolated, it may be sequenced and the sequence information may in turn be used in PCR
amplification of centromere sequences from megachromosornes or other sources of centromeres. isolated centromeres may also be tested for function in vivo by transferring the DNA into a host mammalian cell.
Functional analysis may include, for example, examining the ability of the centromere sequence t~ bind centromere-binding proteins. The cloned centromere wil! be transferred to mammalian cells with a selectable marker gene and the binding of a centromere-specific protein, sueh as anti-centromere antibodies (ego., l.U851, see, Hadlaczky ~t a!. ( 1885) IExc~. Cell Res. 167:1-15) can be used to assess furyction of the centromeres.
b. Tetomeres 'l~ E'referred telorrDeres are the 7 kB synthetic telomere provided herein (see, Examples). A double synthetic telomere construct, which contains a 1 k8 synthetic telorvere linked to a dominant selectable marker gene that continues in an inverted orientation may be used for ease of manipulation. Guch a double canstruct contains a series of TTAGGG repeats 3' of the marker gene and a series of repeats of the inverted sequence, i.e., GGGATT, 5' of the marker gene as follows: ' (GGGATTT?~ --dominant marker gene---t'TTAGGG)~. Using an inverted marker provides an easy means for insertion, such as by blunt end 1lgation, since only properly oriented fragments will be selected.
c. IUlegareplicator The megareplicator sequences, such as the rDNA, provided herein are preferred for use in irs vitro constructs. The rE~~IA provides an origin of ceplication and also provides sequences that facilitate amplification of the artificial chromosors~e in vivo tc~ increase $he size of the chromosome to, for example accommodate increasing copies of a heteroiogous gene of interest as well as continuaus high levels of expression of the heterologous genes.

-J~_ d. FiCler heterochromatin Filler heterochromatin, particularly satellite DNA, is included to maintain structural integri~~y and stability of the artificial chromosome arid provide a structural base for carrying genes within the chromosome. The satellite DNA is typically A!T-rich DNA sequence, such as mouse major satellite DNA, or GJC-rich DNA sequence, such as hamster natural sate6lite DNA. Sources of such DNA include any eukaryctic organisms that carry non-coding satellite DNA with sufficient A/T or G/C
composition to promote ready separation by sequence, such as by FACS, or by density gradients. The satellite DNA may also be synthesized by generating sequence containing monotone, tandem repeats of highly AJT-or GlC-rich DNA units.
The most suitable amount of filler heterochromatin for use in construction of the artificial chromosome may be empirically determined by, for example, including segments of various lengths; increasing in size, in the construction process. Fragments that are t~o small to be suitable for use will not provide for a functional chromosome, which may.
be evaluated in cell-based expression studies, or will result in a chromosome of limited functional lifetime or mitotic and structural stability.
e. Selectable marker Any convenient selectable marker may be used and at any convenient locus in the II~AC:
2. Combination of the isolated chromosomal elervoents Once the isolated elements are obtained, they may be combined to generate the complete, functional artificial chromosome. This assembly can be accomplished for example, by icy vitro ligation either in solution, LMP aga~ose or on microbeads. The Iigation is conducted so that one end of the centromere is directly joined to a telomere. The other end of the centromere, ~rvhich serves as the gene-carrying chromosome arm, is built up from a combination of satellite DNA and rL)NA sequence and may atso contain a selectable marker gene: Another telomere is joined to the end of the gene-carrying chromosome arm, The gene-s carrying arm is the site bt vrhich any heterologous genes of interest, for example, in expression of desired proteins encoded thereby, are incorporated either during in vitro construction of the chromosome or sometime thereafter.
3. - Analysis and testing of the artificial chroosorrce Artificial chromosorr~es'constructed in vitro may be tested for functionality in in vivo mammalian cell systems, using any of the methods described herein for the SATACs, minichromosomes, or kno~nrn tv those of skill in the art:
4~. Introduction of desired heterolagous DNA into the in vitro synthesized chromosome Heterologous DNA may be~lntroduced into the ids vitro synthesized chromosome using routine methods of molecular biology, may be introduced using the methods described herein for the SA'T~,Cs, or may be incorporated into the in vitr~ synthesized chrorxiosome as part of one of the synthetic elements, such as the heterochrocnatin. The heterologous DNA may be linked to a selected repeated 'fragment, and then the resulting construct may be amplified In vitro using the methods for such in vitro amplification provided herein (see the Examples?.
D. ~ Introduction of artificial chromosomes into cells, tissues, animals and plants Suitable hosts for introduction of the NIACs provided herein, include, but are not limited to, any anima! or ptant, cell or tissue thereof, including, but not limited to: mammals, birds, reptiles, amphibians, insects, fish, arachnids, tobacco, tomato, wheat, plants and algae. The MACs, if contained in cells, may be introduced by cell fusion or microcell -~2_ fusion or, if the MACs have been isolated from cells, they may be introduced into host cells by any method known to those of skill in this art, including but not limited to: direct I~NA transfer, electroporatian, lipid-mediated transfer, e.g=, iipofection and liposomes, microprojectile bombardment, micro injection in cells and embryas, protoplast regeneration for plants, and any other suitable method (see, e~c~., Weissbach et al. 11988) Methods for Plant Molecular l3ioiogy, Academic Press, N.Y., Section VIII, pp. 421-463; (~rierson ~~ al. (1988) Plant Molecular Biology, 2d Ed., Blackie, London, Ch. ~7-9P see, also IJ.S.
Patent Nos. 5,4.91,075; 5,482,928; and 5,424,499; see, also; e~c~., i.J.S.
Patent No. 5,47~,708, which describes-particle-mediated transformation of mammalian unattached cells].
Other methods for introducing GNA into cells include nuclear microinjection and bacterial protoplast fusion with intact cells.
1 a Polycations, such as polybrene and polyornithine, may also be used. For various techniques for transforming mammalian cells, see e~,a., Keown et al. Methods in Enzymvloay (19901 Vol. 185, pp. 527-537; and Mansc~ur et al. ( 1988) Nature 336:348-352.
For example, isolated, purified artificial chromosomes can be 2f~ injected into an embryonic cell line such as a human kidney primary embryonic cell line [ATCC accession number CRL 1573] or embryonic stem cells [see, e-a., Hogan ~t ai. (1994) IUlanipulating the Motesa Embryo, A :Laboratoryr Manual, Cold Spring Harbor Laboratory Press, Cotd Spring Harbor, NY, see, especially, pages 255-2~~4 and 25 Appendix 3].
Preferably the chrom~somes are introduc~ad by microinjection, using a system such as the Eppendarf automated micrainjection system, and grown under selective conditions, such as in the presence of hygromycin B or neomycin.

-~3-1. Ntethods for introduction of chromosomes unto hosts Depending on the host cell used, transformation is done using standard techniques appropriate to such cells. These methods include any, including those described herein, known to those of skill in the art.
a. ~N~. uptake For mammalian cells that do not have cell walls, the calcium phosphate precipitation method for introduction o~ exogenous DNA [see, e~a., Graham et al. (1975) Viroloay 52:4.56-457; Wigler ~t al. (1979) Proc, Natl. Acad. Sci. U:S.A. 75:1373-'1376; and ~rurrent Protocols in Molecular Bioloay. Voi. 1, Wlley Inter-Science, Supplement 14; Unit .
9.1.1-9.1.9 (1990)] is often preferred. DNA uptake can be accomplished by DNA alone or in the presence of polyethylene glycol IP~G-mediated gene transfer], which is a fusion agent, or by any variations of such methods known to those of skill in the art [see, e.~., U.S. Pat. No.
4;6H4.D6111.
Lipid-mediated carrier systems are also among the preferred methods for intcoduction of DNA into cells [see, e.g., Teifel et al. (1995) Biotechniaues 19:79-80; Albrecht et al. (1996) Ann. He~-natol. 72:73-79;
Holmen _~t al. (1995) (_n 4/itro.Ceil Dev. Biol. Anim, 31:347-351; Remy et al. (1994) Bioconiua. Ghem. 5:F4~~-654 Le Bolc°h et al. (1995) Tetrahedron Lett. 36:6681-6654; Loeffler et al. ('I 9931 Meth. Enzymol.
217:599-6181. Lipofection [see, e,~a., Strauss ( 1996) 1y61eth. Mol. Biol.
54:3~7-327] may also be used to introduce DNA into cells. This method is particularly well-suited for transfer of exogenous DN~e into chicken cells (e~4., chicken blastodermal cells and primary chicken fibrablasts;
see f~razolot et a!. (1991 ). Mol. R~pro. Dev. 30:304-31 ~:). In particular, DNA of interest can be introduced into chickens ~iri operative linkage with promoters from genes, such as lysozyme and ovalbumin, that are -54~
expressed in the egg, thereby permitting expression of the heterologous DNA in the egg.
Additional methods useful in the direct transfer of DNA into cells include particle gun electrofusion [see, ,e.,~r .,, IJ.S. Patent Nos.
4,965,378, 4,923,814, 4,470,004, 4,90fi,576 and 4,441,972] and virion-mediated gene transfer.
A commonly used approach for gene transfer in land plants- involves the direct introduction of purified DNA into protoplasts. The three basic methods for direct gene transfer into plant cells include: 1) polyethylene gtycol [PEGI-mediated DNA uptake, 2) electroporation-rrsediated DNA
uptake and 3) ~micrain~ection. In addition, plants may be transformed using ultrasound treatment [see, elca., International PCT application publication No. WO 91!00358).
b. Electroporateon Electroporation invol~res providing high-voltage electrical pulses to a solution containing a mixture of protoplasts and foreign DNA to create reversible pores in the membranes of plant protoplasts as well. as other cells. Electroporation is generally used for prokaryotes or other cells, such as plants that contain substantial cell-wall barriers. Methods for effecting electroporation are well knovwn [see, e.,g~,, U.S. Patent Nos.
4,784,737, 5,501,967, 5,501,6fi2, 5,019,034, 5,503,899; see, also Fromrnet al. f l 985) Proc. Nat!. Acad. Sci. U.S.A. 82:5824-5828].
For exampte, electroporation is often used for transformation of plants [see, e~a., Ag Biotechnology News 7:3 and 17 (September/October 1990)]. !n this technique, plant protopiasts are electroporated in the presence of the DNA of interest that also includes a phenotypic marker. Electrical impulses of high field strength reversibly permeabilize biomernb~anes allowing the introduction of the plasmids.
)=lectroporated plant protoplasts reform the cell oval!, divide, and form ~Jwt'9_ plant callus. Transformed plant cells will be identified by virtue of the expressed phenotypic marker. The exogenous DNA may be added to thte protoplasts in any form such as, for example, naked linear, circular or supercoiled DNA; DNA encapsulated in liposomes, DNA in spheroplasts;
DNA in other plant protoplasts, DNA complexed with salts, and other methods.
c. Microcells The chromosomes can ~be transferred by preparing microcells containing an artificial chromosome and then fusing with selected target cells. Methods for such preparation and fusion of microcells are well known [see the Examples and also see, e~g., U.S. Patent Nos.
5,240,840, 4,806,476, 5,298,429, 5,396,767, Fournier (7981) Pr~c.
Natl. Ac ad. Sci. U.S.A. '78:6349-6353; and Lambert et al. (1991) Proc.
Natl. Acad. Sci. U.S.A. 88:5307-59J. Microcell fusion; using microcells 75 that contain an artificial chromosome, is a particularly useful method for introduction of MACS into avian cells, such as DT40 chicken pre-B veils (for a description of DT40 cell fusion, see, e.a., Dieken ~t al. ( 19961 Nature Genet. 12:174-182J.
2. t~losts Z~ Suitable hosts include any host known to be useful for introductian and expression of heteroiogous DNA. Dl particular °snterest herein, animal and plant cells and.tissues, including, but riot limited to insect cells and larvae, plants; and animals, particularly transgenic anon-human) .
animals, and anima! cells. ~ther hosts include, but are not limited to 25 mammals, birds, particularly fowl such as chickens, reptiles, amphibians, insects, fish, arachnids, tobacco, tomato, wheat, monocots, divots and algae, and any host into which introduction of heterologous DNA is desired. Such introduction can be effected using the MACs provided herein, or, if necessary by using the MACS provided herein to identify species-specific centromeres ar~dlor functiona6 chr~amosomal units and then using ttte resulting centromeres or chromosomal units as artificial chromosomes, or alternatively, using the methods exemplified herein for production. of MACs to produce species-specific artificial chromosomes.
a. Introduction of DIlt~1 into embryos for production of transgenic (non-hurt~an? animals and introduction of DNA into animal cells '10 Transgenic (non-human) animals can be produced by introducing exogenous genetic material into a proriucleus of a mammalian zygote by microinjection (see, e~c~., I.J.S. Patent IVos. 4,873,191 and 5,354,674;
seer also, International PCT ~pptication publication lVo. 1~!VO 95114769..
The zygote is capable of development into a mamrrtial. The embryo or zygote is transplamed into a host female uterus and' allowed to develop. Detailed protocols and exarriples are set forth below.
Nuclear transfer (see., Wilr~iufi et a1. (1997? Nature 385:810-813, International PCT application Nos. WO 97107669 and WO 97/Q~7668i.
24 Briefly in this- method, the SATAC containing the genes of interest is introduced by any suitable method, into an appropriate donor cell, such as a mammary gland cell, that contains totipotent nuctei. The diploid nucleus of the cell, which °rs either in GO.or G1 phase, is then introduced;
such as by cell fusion or microinjection; into an unactivated oocyte, preferably enucfeated cell, which is arrested in this metaphase of the second meiotic division. Enucieation may be effected by any suitably method, such as actual removal, or by treating with means, such as ultraviolet light, that functionalty remove the nucleus. The oocyte is then activated, preferably after a period of contact; about 6-20 hours for cattle, of the new nucleus with the cytoplasm, while maintaining correct ~ 57-ptoidy, to produce a reconstituted embryo, which is then introduced into a host. Ploidy is maintained during activation, for example, by incubating the reconstituted cell in the presence of a microtubtale inhibitor, such as nocodazole; colchicine, cocemld, and taxol, whereby the DNA replicates once.
Transgenic chickens can be produced by injection ~f dispersed blastodermal cells from Stage X chicken embryos into recipient embryos at a similar stage of development Isee e.~., Etches et at. I1993y Po~ltry Sci. 72:882-889 Petitte et al. (1990) Develotarr~ent 108:785-189].
Heterologous DNA is first introduced into the donor blastodermal cells using methods such as, -for example, iipofection (see, e.~.,, Brazo.lot et al.
I1991 ) Mol. Reraro. Dev. 30:304.-312] ar microcetl fusion Isee, e.~., Dieken et al. (1996) Nature Cenet. 12:174-1$2]: 1'he transfected donor cells are then injected into eecipient chicken embryos Isee e~4., Carsience '15 et at. 11993) Development 1 17: 669-675]. The recipient chicken embryos within the shell are candled ahd allowed to hatch to yield a germline chimeric chicken.
DNA can be introduced into animal cells using any known procedure, including, but not limited to: 'direct upt2ike, incubation v~Aith 2~ polyethylene glycol IPEG], microinjection, electroporation, lipofectlon, cel!
fusion, microcell fusion, particle bombardment, including microprojectife bombardment Isee, e~ca., IJ.S. Patent No. 5.4.70,708, which provides a method for transforming unattached marrimalian cells via particle bombardment], and any other such method. For exempts! the transfer of 25 plasmid DNA in liposomes directly to human cells rs~ sitr~ has been approved by the FDA for use in humans Isee, e.g_, Nabei, et al. (1990) Science 249:1285-1288 and l~.S. Patent No. 5,461 ~m32].

.~-b. Introduction ~f heterolog~us ~I~A into p9ants Numerous methods for producing or developing transgenlc plants are available to those of :skill in the art. The method used is primarily a function of the species of plant. These methods include, but are not limited to: direct transfer of DNA by processes, suich as P6G-induced DNA uptake, protoplast fusian, microinjection, electroporation, and microprojectile bombardment [see, e~4., Uchimiya ~t ai. (°1989) J. of Biotech. '! 2: 1-20 for a review of such procedures; see, also, e.a., U.S.
Patent Nos. 5,436,392 and 5,~.89,52CD and many othersj. For purposes 1 ~ herein, when introducing a i~IIAC, microinjection, protoplast fusion and particle gun bombardment are preferred.
Plant species, including tobaccos rice, maize, rye, soybear9, Brassica n~u_s_, cotton, lettuc~, potato and tomato, have been used to produce transgenic plants. Tobacco and other species, such as petunias, often serve as experiments! models in which the methods have been developed and the genes first introduced and eicp~ressed.
DNA uptake can be accomplished by DNA alone or ire the presence of PEG, which is a fusion agent, with plant protoplasts or by any variations of such methods known t~ those of skill in the art [seep e.~., 2~ U.S. Patent No. 4,684,1"r11 to Schilperoot et al.). Electroporation, which involves high-voltage electrical pulses to a solution containing a mixture of protoplasts and foreign DNA to create reversible pores, has been used, for example, to successfully introduce foreign genes into rice and Brassica nraaus, lVlicroinjection of DNA into plant cells,, including cultured cells and cells in intact plant organs and embryoids in tissue culture end microprojectile bombardment [acceleration of small high density particles, which contain the DNA, to high velocity with a particle gun apparatus, which forces the particles to penetrate plant celli.walls and membranes) have also been used. All plant cells into which DNA can be introduced and that can be regenerated frog the transformed cells can be used to produce transformed whole plants which contain the transferred artificial chromosome. The particular protocol and means for introduction of the DNA into the plant host may need to be adapted or refined to suit the particular plant species or cuttivar.
c. Insect cells Insects are useful hosts for introduction of arttfICial chromosomes for nurrierous reasons, including,, but nat lirraited to: la) amplification of genes encoding useful proteins can be accomplished in the artificial °I~ chromosome to obtain higher protein yields in insect cells; (b) insect cells support required post-translationat modifications, such as glycosylation and phosphorylation, that can be required for protein biological functioning; (c) insect cells do not support mammalian viruses, and, thus, eliminate the problem of cross-contamination of products with such '15 infectious agents; (d) this technology circumvents traditional recombinant baculovirus systems for production of nutritional, industrial or medicinal proteins in insect cell systems; (e) the low temperate~re optimum for insect cell growth (28° C) permits reduced energy cost of production;
(f) serum-free growth medium for insect cells permits lower production 2~ costs; (g) artificial chramosome-containing cells,can be stored indefinitely at low temperature; and (h) insect larvae will be biological factories for production of nutritional, medicinal or industrial proteins by microinjection of fertilized insect eggs [see~ e~..4., Joy et al. ~ 1991 ) Current Science 66:145-150, which provides a method for microinjecting heterotogous 25 DN~1 into Bombyx mori eggs].
Either II~ACs or insect-specific artificial chromosomes [BUGACs]
will be used t~ introduce genes into insects. As described in the Examples, it appears that MACS .will function in insects to direct expressian of heterologous DNA contained thereos~a For example, as ~~7~.
described in the Examples, a MAC containing the B. marl actin gene promoter fused to the IacZ gene has been generated by transfection of EC3l7C5 cells with a plasmod containing th~ fusion gene. Subsequent fusion of the B. marl ceps with the transfected EG3~A'7C5 cells that survived selection yielded a MAC-containing insect-mouse hybrid cell fine in which ~-galactosidase expression was detectable.
Insect host cells include, but are not limited to, hosts such as Spodoptera frugiperda [caterpillar], Aeries aegypti l~rnosquito], Aeries albopictus (mosquito), ~rosphila efanr~gaster Efruitflyl, gom,byx marl [silkworm), Mancluca sexta [tomato horn worm] and Trichoplusia ni [cabbage looper): Efforts have been directed toward propagation of insect cells in culture. Such efforts have focused on the fall armyworm, Spodoptera frugiperda: Cell lines have been developed also from other insects such as the cabbage looper, Trichoplusia ni and the silkworm, 1 a Borr~byx rraor:. It has also bran suggested that analogous cell lines can be created using the tomato hornworm, Manduca sexta. To introduce DNA into an insect, it should be intre~duced into this larvae, and allowed to proliferate, andthen the hemolymph recovered from the larvae so that the proteins can be isolated therefrom.
The preferred method herein for introduction of artificial chromosomes into insect cells is microinjection [see, e~c~., Tamura et al.
(19911 Bio Ind. _5:26-31 a Nikolaev et ate. 119891 Mot. Biol. (Moscow) 23:i 177-S7e and methods exemplified and discussed herein].
I=. Applications for, and Uses of Artificial chrornosore~es Artificial chramosomes provide convenient and useful vectors, and in some instances [e.a., in the case of very I~rge heterologous genes) the only vectors, for introduction of heterologous genes into hosts. Virtually any gene of interest is amenable to infiroduction into a host via artificial chromosomes. Such genes include, but are not limited to, genes that Y
encode receptors, cytokines, enzymes, professes, hormones, growth factors, antibodies, tumor suppresser genes, therapeutic products and multigene pathways.
The artificial chromosomes provided herein will be used in methods of protein and.gene product production, particularly using insects as host cells for production of such products, and in cellular ie~c~., mammalian cell) production systems in which the artificial chromomsomes (particularly MACs) provide a reliable, stable and efficient means for optimizing the biomanufacturing of important compounds for medicine and industry. They are also intended for use in methods of gene therapy, and for production of transgenic plants arid, animals (discussed above, below and in the EXAMPLES).
1. Gene Therapy Any nucleic acid encoding a therapeutic. gene: product or product °~5 of a multigene.pathway may be introduced into a heat animal, such as a human, or into a target c~Il line for introduction into an animal, for therapeutic purposes. Such therapeutic purposes include genetic therapy to cure or to provide gene products that are missing or defective, to deliver agents, such as.anti-tumor agents, to targeted cells or to an animal, and to provide gene products that will confer resistance or .
reduce susceptibility to a pathogen or ameliorate symptoms of a disease or disorder. The following are some exemplary genes and gene products.
Such exemplification is not intended to be limiting.
a. Anti-lily ribozymes As exemplified below; DNA encoding anti-HlV ribozymes can be introduced and expressed in cells using MACs, including the euchromatin-based minichrorrtosomes and the SATACs. These MACS
can be used to make a twansgenic mouse that expresses a ribozyme and, thus, serves as a model for testing the activity of such ribozymes or from which ribozyme-producing cell lines can be made. also, introduction of a M,~C that encodes an anti-hllV ribo~yme into human cells will serve as treatment for HIV infection. Such systems further demonstrate the viability of using any disease-specific ribozyme to treat or ameliorate a particular disease.
b. 'tumor Suppresser Genes Tumor suppresser genes are genes that, in their wi~d-type alleles, express proteins that suppress abnormal cellular proliferation. When the gene coding for a tumor suppresser protein is mutated or deleted, the 1 ~ resulting mutant protein or the complete lack of tumor suppresser protein expression may result in a failure to correctly regulate cellular proliferation. Cansequentiy, abnormal cellular proliferation may take place, particularly if there is already existing damage to the cellular regulatory mechanism. A number of well-studied human tumors and tumor cell tines have be~n shown to have missing azr nonfunctional turraor suppresser genes.
Examples of tumor suppression genes include, but are not limited to, the retinobtastoma susceptibility gene or Ft8 gene, the p53 gene, the gene that is deleted in colon carcinoma [i:e., the DCC gene) and the neurofibromatosis type 1 [NF-1 ~ tumor suppresser gene [see, ~.g_, tJ.S.
Patent No. 5,496,731; Weinberg et al. X1991) 254~a1135-7145. Loss of function or inactivation of tumor suppresser genes may play a central role in the initiation andJor progression of a significant number of human cancers.
The p53 Geese Somatic cell mutations of the p53 gene are said to be the most frequent of the gene mutations associated with human cancer [see, e~g., Weinberg eat at. ( 1991 ) Science 254:1138-11461. The r~ormai or wild-type p53 gene is a negative regulator of cell growth, which, when damaged, favors cell transformation. The p53 expression product is found in the-nucleus, where it may act in parallel or cooperatively with other gene products. Tumor cell fines in which p53 has been deleted have been successfully treated with wild-type p53 vector to reduce tumorigenicity [see, Baker et ai. [1990) Science 249:912-915].
DNA encoding the p53 gene and plasmids containing this DNA are well known [see, e.g_, U.S. Patent No. 5Y260,191; see, also Chen et al.
[ 1990) Science ,x:1576; Farrel et al. ( 1991 ) EM Bt? J. 1 t~:2879-288 A
plasmids containing the gene are available from the ATCC~ and the 1~ sequence is in the GenBank Database, accession nos. X5415f, X60020, M14695, M16494, KQ3199].
c. The CFTR gene Cystic fibrosis [CF] is an autosomal recessive disease that affects epithelia of the airways, sweat glands, pancreas, and other organs. It is a lethal genetic disease associated with a defect in chl~ride ion transport, and is caused by mutations in the gene coding for the cystic fibrosis transmembrane conductance regulator [~FTR], a 1480 arraino acid protein that has been associated with the expression of chloride eaonductance in a variety a~ eukacyotic cei! types. Defects in CFTR destroy or reduce the ability of epithelial cells in the airways, sweat glands, pancreas and other tissues to transport chloride ions in response to CAMP-mediated agonists and impair activation of apical membrane channels by cAl'//IP-dependent protein kinase A [PKA]. Given the high incidence and devastating nature of this disease, development of effective CF treatrnents is imperative.
~5 The CFTR gene [-250 kb] can be transferred into a MAC for use, for example, in gene therapy as follaws. A CF-YAC [see Green et al.
Science 250:94-98) may be modified to include a selectable marker, such as a gene encoding a protein that canfers resistance to puromycin or hygromycin, and ~l-DNA for use in site-specific integration into a neo-t ..~L~,_ minichromosome or a SA~"AC. Such a modified CF-YAC can be introduced into MAC-containing cells, such as EC3/'7C5 or 19CaxHa~
ce!!s, by fusion with yeast protoplasts harboring the modified CF-YAC:or microinjection of yeast nuclei harboring tile modified CF-'SAC into the cells. Stable transformants are them selected on the basis of antibiotic resistance. These transformants will carry the modified CF-YAC within the fVIAC contained in the cells.
2. Animals, birds, dish arid plants that are genetically altered t~
possess desired traits such as resistance to disease 't0 Artificial chromosomes are ideally suited for preparing animals, including vertebrates and invertebrates, including birds and fish as well as mammals, that possess certain desired traits, such as,. for example;
disease resistance, resistance to harsh environments! conditions, altered growth patterns, and enhanced physical characteristics.
~ne example of the use of artificial chromosomes in generating disease-resistant organisms involves the preparation of multivalent vaccines. Such vaccines include genes encoding multiple antigens that can be carried in a MAC, or species-specific artificial chromosomes and either delivered to a host to induce immunity', or incorporated into embryos to produce transgenic (non-human! animals and plants that are immune or less susceptible to certain diseases.
Disease-resistant animals and plants may also be prepared in which resistance or decreased susceptibility to disease is conferred by introduction into the host organism ar embryo of artificial chromosomes containing DIVA encoding gene products (e'o,, ribbzymes and proteins that are toxic to certain pathogensD that desfiroy or attenuate pathogens or limit access of pathogens to the host.
Animals and plants possessing desired trails that might, for example, enhance utility, processibility and commercial value of the organisms in areas such as the agricultural and ornamental plant industries may also be generated using artificial chromosomes in the same manner as described above for production of disease-resistant animals and plants. In such instances, the artificial chromosomes that b are introduced into the organism or embryo contain DNA erncoding gene products that serve to confer the desired trait in the organism.
Birds, particularly fowl such as chickens, fish arid crustaceans will serve as model hosts for production of genetically alfered organisms using artificial chromosomes.
i D 3. Use of MACS and other artificial chromosomes for preparation and screening of libraries Since large fragments of DIVA can be incorporated into each artificial chromosome, the chromosomes are well-suited far use as cloning vehicles that can accommodate entire gen~mes in the preparation .
15 of genomic DNA libraries. which then can be readily screened. For example, MACs may be used to prepare a gertomic DNA library useful in the identification and isolation of functional centrorneric DNA from different species of organisms. In such applications, the MAC used to prepare a genomic ~NA library from a particular organisrr~ !s one that is 20 not functional in cells of that organism. That is. the MAC doss not stably replicate, segregate or provide for expression of genes contained within it in cells of the organism. Preferably, the I~IIACs contain an indicator gene (e~c:, the IacZ gene encoding ,Q-galactosidase or genes encoding products that confer resistance to antibiotics such as 25 neomycin, puromycin, hygromycin? linked to a promoter that is capable of promoting transcription of the indicator gene in cells of the organism.
Fragments of getsomic DNA from the organism are incorporated into the MACs, and the MACS are transferred to cells from, the organism: Cells that contain MACs that have incorporated functions! cer~tromeres contained within the genomic ~NA fragments are identified by detection of expression of the marker gene.
4. Use of MACS arid other artificial chromosomes for stable, high-level protein production b Cells containing the MACS andlor other artificial chromosomes provided herein are advantageously used for production of proteins, particularly several proteins from one cell line,-.such as multiple protein involved in a biochemical pathway or multivalent vaccines. The genes encoding the proteins are introduced into.the artificial chromosomes which are then introduced into cells. Alternatively, the heterologous genes? of interest are transferred into a production calf line that already contains artificial chromosomes in a rrianner that targets the genes' to the artificial chronrsosomes. The cells are cultured under conditions whereby the heterologous proteins are expressed. l3ecause the proteins will be expressed at high levels in a stable permanent extra-gertomic chromosomal system, selective conditions are not required.
Any transfectable cells capable of serving a:~ recombinant hosts adaptable to continuous propagation in a cell cult~,~re system (see, elo., McLean ( 1993) Trends lc~ Siotech. 1 1:232-238] are suitable for use in an 2~ artificial chromosome-based protein production system. Exemplary host cell lines include, but are not limited to, the following: Chinese hamster ovary (CHOI cells (see, e~a., fang et al. (1995) Siotechnolo4y 13:389-392], HEK 293, Ltk-, C~S-~, ~G4.4, and SHiC cells. CHt3 cells are particularly preferred host cells. Selection of host cell lines for use in 2S artificial chromosome-based protein production systems is within the skill of the art, but often udiil depend on a variety of factors, including the properties of the heterologous protein to be produced, potential toxicity of the protein in the host cell, any requirements for post-transtational modification (elc~., glycosylation, amination, phosphorylation) of the protein, transcription -factors available in the cells, the type of promoter elements) being used to drive expression of the heterologous gene, whether production will be completely intracellular or the heterologous protein will preferably be secreted from the cell, and the types of processing enzymes in the cell.
The artificial chromosome-based system for heterologous protein production has many advantageous features. !=or example, as described above, because the heterologous DNA is located in an independent, extra-genomic artificial chromosome (as opposed to randomly inserted in 70 an unknown area of the host cell genome or located as extrac.hromosomal elements) providing only transient expression) it is stably maintained in an active transcription unit and is not subject to ejection via recombination or elimination during cell division.
Accordingly, it is unnecessary to include a selection gene in the host '15 cells and thus growth under selective conditions is also unnecessary.
Furthermore, because the artificial chromosomes are capable of incorporating large segments of DNA, multiple copies of the heterologous gene and linked promoter elements) can be retained in the chromosomes, thereby providing for high-level expression of the foreign 20 protein(s). Alternatively, multiple copies of the gene cari be linked to a single promoter element and several different genes may be linked in a fused polygene complex to a single promoter for expression of, for example, all the key proteins constituting a complete metabolic pathway [see, e~c~., Beck von Bodman et al. 41995) Biotechn Vocv 13:587-691 ].
25 Alternatively, multiple copies of a single gene can be operatively linked, to a single promoter, or each or one or several copies maybe linked to different promotes or multiple copies of the same promoter.
Additionally, because artificial chromosomes have an almost unlimited capacity for integration and expression of foreign genes, they can be used not only for the expression of genes encoding end-products of interest, but also for the expression of genes associated with optima!
maintenance and metabolic management of the host cell, ela., genes encoding growth factors, as well as genes that may facilitate rapid synthesis of correct form of the desired heteroiogous protein product, e'a., genes encoding processing enzymes and transcription factors.
The MACS are suitable for expression of any proteins or peptides, including proteins and peptides that require in vivo posttranslationai modification for their biological activity. Such proteins include, but are not limited to antsbody fragments, full-length antibodies, and multimeric antibodies; tumor suppressor proteins, naturally occurring or.
artificial antibodies and enzymes, heat shock proteins, arid others.
Thus, such cell-txased '°protein factories" employing MACS can generated using MACs constructed with multiple copses [theoretically an unlimited number or at least up to a number such l:hat the resulting MAC
is about up to the size of a genomic chromosome (i.e., endogenous)1 of protein-encoding genes with appropriate promoters, or o°r~ultiple genes driven by a single promoter, i.e., a fused gene complex [such as a complete metabolic pathway in plant expression system; see, ew., Beck vvn Bodman (1995) Biotechnology 13:58'7-591]. Once such MAC is constructed; it can be transferred to s suitable cell culture system, such as a CHO cell line in protein-free culture medium [see, e~a., (19951 Biotechnoloav 13:389-39] or other immortalized cell lines (see, e~cs., ( 1993) TIBTECH 11:232-238 where continuous production can be established.
The ability of MACs to provide fvr high-level expression of heterologous proteins in host cells is demonstrated. for example, by analysis of the H1 D3 and G3D5 cell lines described herein and deposited with the ECACC. Northern blot analysis of mRNA obtained from these 'b° 9_ cells reveals that expression of the hygromycin-resistance and ~3-galactosidase genes in the cells correlates with the amplicon number of the megachromosome(s1 captained therein.
F. Methods for the synthesis of DIVA sequences containing repeated DNA units Generally, assembly of tandemly repeated DIVA poses difficulties such as unambiguous annealing of the complementary oligos. For .
example, separately annealed products may ligate in an inverted orientation. Additionally, tandem ar inverted repeats are particularly susceptible to recombination and deletion events that may disrupt the sequence. Selection of appropriate host organisms ie~a., rec' strains) for use in the cloning steps of the synthesis of sequences of tandemly repeated DNA units may aid in reduction and elimination of such events.
Methods are provided herein for the synthesis of extended DNA
sequences containing repeated DNA units. These rnethods are particularly applicable to the synthesis of arrays of tandemly repeated DNA units, which are generally difficult or not possible to construct utilizing other known gene assembly strategies. A specific use of these methods is in the synthesis of sequences of any length captaining simple (e.g., ranging from 2-6 nucleotides) tandem repeats (such as telomeres and satellite DNA repeats and trinucleotide repeats of possible clinical significance? as well as complex repeated DNA sequences. An particular example of the synthesis of a telomere sequence containing over 1 aA
successive repeated hexamers utilizing these methods is provided herein.
The methods provided herein for synthesis of arrays of tandem DNA repeats are based in a series of extension steps in which successive doublings of a sequence of repeats results in an exponential expansion of the array of tandem repeats. These methods provide several advantages over previously known methods of gene assembly. For instance, the starting oligonucfeotides are used only once. The intermediates in, as well as the final product off the construction of the DIVA arrays described herein may be obtained in cloned form in a ~microbiai organism (e.c~., E.
coli and yeast). Of particular significance, wine regard to these methods is the fact that sequence length increases exponentially, as opposed to linearly, in each extension step of the procedure even though only two oiigonucieotides are required in the methods. The construction process does not depend on the compatibility of restriction enzyme recognition sequences and the sequence of the repeated DNA because restriction sites are used only temporarily during the assembly procedure. No adaptor is necessary, though a region of sirniiar function is located between two of the restriction sites employed in the process. The only limitation with respect to restriction site use is that the two sites employed in the method must not be present elsewhere in the vector utilized in any cloning steps. These procedures can also he used to construct complex repeats with perfectly identical repeat units, such as the variable number tandem repeat (VNTR) 3' of the human apolipoprotein B100 gene (a repeat unit of 30 6p. ~ 00% .AT) or alphoid satellite DNA.
The method of synthesizing DNA sequences containing tandem repeats.
may generally be described as follows.
1. Starting materials Two oligonucleo.tides are utilized as starting materials.
Oiigoriucleotide 1 is of length k of repeated sequence (the flanks of which are not relevant) and contains a relatively short stretch (60-90 nucleotides) of the repeated sequence, flanked with appropriately chosen restriction sites:
5'-S1»»»»»»»»»»»>~»»S2 -3' wherein S1 is restriction site 1 cleaved by E1 [preferably an enzyme producing a 3'-overhang 4e.9_. Pact, Pstl, ~1, N;sil, etc.) or blunt-end3, S2 is a second restriction site cleaved by E2 (preferably an enzyme producing a 3'-overhang or one that cleaves outside the recognition sequence, such. as TsnRl), > represents a simple repeat unit, and '~' denotes a short (8-10) nucleotide flanking sequence complementary to oligonucleotide 2:
3'- S3-5' wherein S3 is a third restriction site for enzyme E3 and which is present 1 ~ in the vector to be used during the construction.
Because there is- a large variety of restriction enzymes that recognize many different ~NA sequences as cleavage sites, it should always be possible to select sites and enzymes (preferably those fihat yield a 3'-protruding end) suitable far these methods in connection with the synthesis of any one particular repeat arrary. !n most cases, anly 1 for perhaps 2) nucleotidets) has of a restriction site is required to be present in the repeat sequence, and the remaining nucleotides-of the restriction site can be removed, far example:
.. Pacl; TTAAT/TAA-- (KtenowldNTP1 TAA--Pstl: CTGCAIG-- (KienowldNTP) G--mil: ATGCA/T-- (KlenowldNTP) T--Koni: GGTAC/C-- (KlenowIdNTP) C--Though there is no known restriction enzyme leaving a single A
behind, this problem can be salved with enzymes leaving behind none at all, for example;
Tail: ACCT/ (Kfeno.wIdNTP) --N[alll: CATG/ (Klenow/dNTP) __ Additionally, if mung bean nuclease is used ihstead of Klenow, then the following , -~x-Xbal; TICTAGA Mung bean nuclease A--Furthermore, there are a number of restriction enzymes that cut outside of the recognition sequence, and in this case, there is no limitation at all:
Ts~sRl NNCAGTGNN%-- (KtenowIdNTP) --Bsm! GAATG CNl-_ . 4KlenowIdNTP) _-' CTTAC/GN -- iKlenowldNTP) -2. Step 1 = Annealing Otigonucleotides 1 and 2 are annealed at a temperature selected 1D depending on the length of overlap (typically in the range of,30_65 °C).
3. Step 2 - Generating a double-stranded molecule The annealed oGgonucleotides are filled-in with Kienoviv polymerise in the presence of dNTP to produce a double-stranded (ds) sequence:
5°-S1»»»»»»»»»»»»»»>a»>S2 S3~~3' 1J 3°-S1«««««««««««<e««««<s2 53-5' Step 3 - Incorporation of double-stranded ~IiiFi into a vector The double-stranded DNA is cleaved with restriction enzymes E1 and E3 and subsequently ligated into a vector ~e~a., pUCl9 or a yeast vector) that has been cleaved with the same enzymes ~1 and E3. The 20 ligation product is used to transform competent host cells compatible with the vector being used (e~o., whery pUCl9 is used, bacterial cells such as E. coli DH5ar are suitable hosts) which are'then plated onto selection plates. Recombinants can be identified either by color (e~,ca., by ,X-gal staining for ~B-galactosidase expression) or by colony hybridization Zb using 32P-labeled oligonucleotide 2 (detection by hybridization to oligonucleotide 2 is preferred because its sequence is removed in each of the subsequent extension. steps and thus is present only in recombinants that contain DNA that has undergone successful extension of the repeated sequence).
t 5. Step 4 - isolation of insert from the piasrhid An aliquot of the recombinant plasmid containing k nucleotides of the repeat sequence is digested with restriction .enzymes E1 and E3, and the insert is isolated on a get (native polyacrylamide while the insert is 'S short, but agarose can be used for isolation of longer inserts in subsequent steps?. A second aliquot of the recombinant plasmid is out with enzymes E2 (treated with Kienow and dNTP to remove the 3°-overhang) and E3, and the large fragment (plasmid ~NA plus tfie insertl is isolated. .
°I0 . 5. Step 5 - Extension of the ~NA sequence of k repeats The two DNAs (the S'!-S3 insert fragment and the vector plus insert) are ligated, plated to selective plates, and screened for extended recombinants as in Step 3. Now the length of the repeat sequence between restriction sites is twice that of the repeat sequence in the 15 previous step, i.e., 2xk.
7. Step 6 - Extension of the DtllA sequence of 2xk repeats Steps 4 and 5 are repeated as many times as needed to achieve the desired repeat sequence size. In each extension cycle, the repeat sequence size doubles, i.e., if m is the number of extension cycles, the 20 size of the repeat sequence will be k x ~"' nucleotides.
The following examples are included for illustrative purposes only and are not.intended to limit the scope of the invention.

Genera! Materials and Methods 25 The following materials and methods are exemplary of methods that are used in the following Examples and that can be used to prepare cel! lines containing artificial chromosomes. ~ther suitable materials and methods knav~in to those of skill in the art may used. Modifications of these materials and methods known to those of skill in the art may also be employed.
A. Culture of cell lines, cell fusion, and transfection of cells 1. Chinese hamster K-2Q cells and mouse A9 fibroblast cells were cultured in F-12 medium. EC3/7 [see, U.S. Patent No.
5,288,625, and deposited at the European Collectian of Animal cell Culture (ECACC) under accession no. 90051001; see, also Hadlaczky et ai. 41991) Frog. Natl: Acad. Sci. U.S.A. 88:8106-8110 and U.S.
application Serial No. 08/375,271] and EC317C5 [see, tJ.S. Patent No.
1~ 5,288,625 and Praznovszky et al. (1991) Proc. Nati. Acad. Sci. U.S.A.
88:1 1 ~42-1 10461 mouse cell lines, and the KE 1-2/4 hybrid cell line were ri~aintained in F-12 medium containing 400 ~rg/mf 6418 [SIGMA, St.
Louis, MO].
2. TF10~4619 and .TF1004G-1905 mouse cells, described belaw, and the 19C5xHa4 hybrid, described below, and ids subtines were cultured in F-12 medium containing up to 400 ~g/ml Hygromycin B (Cait~iocheml. LP11 cells were maintained in F-12 medium containing 3-15 pglm! Puromycin (SIGMA; St. i_ouis, M01.
3. ~ Cotransfection of EC3/7C5 ceils-with plasmids [pH132, pCH110 available from Pharmacia, see, also Half et al. (1983) J. Mol. Appl. Ci~n. _2:101-109] and with ~t DNA was conducted using the calcium phosphate DNA precipitation rrtethod [see, e.,g_, Chen et al.
(1987) Mot. Cell. Blot. '?:2746-27521. using 2-5 beg plasmid DNA and 20 ,ug a phage DNA per 5 x .10~ recipient cells.
4. Cel! fusion Mouse and hamster cells were fused using polyethylene glycol (Davidson et al. (1976) Som. Cell Genet. 2:165-176]. Hybrid cells were selected in HAT medium containing x.00 Nglrnl ~iygromycin B.

.75_ Approximately 2x107 recipient and 2x106 donor cells. were fused using polyethylene gkycok [Davidson et al. (1976) Sorr,Z Celi Genet.
2:165-176]. Hybrids were selected and maintained in F-1 /HAT medium [Szybaksky e~ aP. !1962) Natl. Cancer Inst. Monogr. 'x:75-89] containing 1~°~6 ECS and 400 ~rg/rnl 6418. The presence of °'p~arental"
chromosomes in the hybrid cell lines was verkfied by in situ. hybridization wkth species-specific probes using biotin-labeled human and harr~ster g!enomic DNA; and a mouse long interspersed repetitive DNA
[pMCPEI .51 ).
5. Microcelt fusion Microcell-mediated transfer of artificial chrorreosomes.from EC3I7C5 cells to recipient cells was done according to Saxon et al.
[( 1985) Mol. Cell. Biol. ~ :140-146] with the modifications of Goodfellow et al. [(1989) Techniques for mammalian genome transfer. in Ger~ome Analysis a Practical A,o~rmach. K.E. Davies, ed., !R!. Press; Oxford, Washington DC. pp.1-17] and 1'amada ~t al. [(1990) ~ncogene 5:1141-11471. Briefly, 5 x 106 EC3/7C5 cells ire a T25 flask were treated first with 0.05 Nglml colcemid for 48 hr and then with '10 ~rglml cytochalasin B for 30 min. The T25 flasks v~rere centrifuged on' edge and the pelleted microcells were suspended in serum free DME medius~r9. The microcells were filtered through fkrst a 5 micron and then a 3 micron polycarbonate filter, treated with 50 ~glml of phytohemagglutin, and used for polyethylene glycol mediated fusion with recipient cekls. Selection of cells containing the MMCneo was started 48 hours after fusion in , medium containing 400-800 ~glml 6418.
Microcells were also piepared from 1 B3 and GNB42 donor cells as . follows in order to be fused with E2D6K cells (a Ci-i0 !C-20 cell line carrying the puromycin N-acetyltransferase gene, ~~y.e., the puromycin resistance gene, under the control of the SV40 early promoter]. The -7s-donor cells were seeded to achieve 60-7S% confiuency within 24-36 hours. After that time, the cells were arrested in mitosis by exposure to colchicine (10 ~g/ml) for 12 or 24 hours to induce imicronucleatlon. To promote micronucleation of GHB42 cells, the cells were exposed to hypotonic treatment (10 min at 37aC). After.cotchicine treatment, or after colchicine and hypotonic treatment, the cells were grown in colchicine-free rriedium:
The donor cells were trypsinized and centrifuged and the pellets were suspended in a 1:1 Percoil medium and incubated for 30-40 min at 37°C. After the incubation. 1-3 x 10' cells (60-7(3°~
micronucieation index) were loaded onto each Percoll gradient leach fusion was distributed on 1=2 gradients?. The gradients vsiere centrifuged at 19,000 rpm for 80 min in a Sorvall SS-34 rotor at 34-37°C. After centrifugation, two visible bands of cells were removed, centrifuged at 2000 rpm, 10 min at 4°C, resuspended and filtered through 8 pm pore size nucleopore filters.
The microcells prepared from the 1 B3 and t~HO342 cells were fused with E2DfiK. The E2D6K cells were generated by CaPG4 transfectior~ of CHt~ K-20. cells with pCHTV2. Plasmid pCl-iTV2 contains the puromycin-resistance gene linked to the SV40 promoter and poiyadenylation signal, the Saccharom3rces cerevisiae URA3 gene, 2.4- and 3.2-kb fragments of a Chinese hamster chromosome 2-specific satellite ~NA (HC-2 satellite;
see Fatyol et al: (1994) Nuc. Acids Res. 22:3728-3?36), two copies of the diptheria toxin-A chain gene (one linked to the herpes simplex virus thymidine kinase (HSV-TK) gene promoter and SV40 polyadenylation signal and the other linked to the HSV-TK promoter without a polyadenylation signal, the ampicitlin-resistance gene and the ColE1 origin of replication. Following transfection, puromycin-resistant colonies were isolated. THe presence of the pCHTV2 piasmid in the E2D6K cell line was confirmed by nucleic acid amplification of DhIA isolated from the cells.
The purified microcelfs were centrifuged as described above and' resuspended in 2 mt of phytohemagglutinin-P (PHA-P, 100 ~glml). The microcell suspension was Then added to a 50-'70% confluent recipient culture of E2D6K cells. The preparation was incubated at room temperature for 30-40 min to agglutinate the microcells. After the PFIA-P
was removed, the cells were incubated with 1 mi of 5~% polyethylene-glycol (PEG) for one min. The PEG was. removed and the culture was 1~ washed three times with F-12 medium v~rithout serum. The cells were incubated in non-selective medium, for 48-60 hours. After this time, the cell culture was trypsinized and plated in F-12 medium containing 4.00 ug/ml hygromycin B and 10 g/rrbl puromycin to select against the parental cell lines.
Hybrid clones were isolated from the cells that had been cultured in selective medium, These clones were then analyzed for expression of j3-galactosidase by the X-gal staining method. Four of five hybrid clones analyzed that had been generated by fusion of GH54~ microcelis with E2D6K cells yielded positive staining results indicating expression of ~3-galactosidase from the tact gene contained in the megachromosame contributed by the GH1342 cells. Similarlyo a hybrid clone that had been generated by fusion of 153 microcells with E~D6K cells yielded positive staining results indicating expression of ~3-gaiactosidase from the lack gene contained in the megachromosome contributed by the 113 cetls. In situ hybridization analysis of the hybrid clones is also performed to analyze the mouse chromosome content of the mouse-hamster hybrid cells.

B. Chromosome banding Trypsin G-banding of_chromasornes was performed using the method of Vifang & Fednroff ((1972) Nature 235:52-54], arid the detection of constitutive heterachromatin with the BSG: C-banding method was done according to Sumner ((1972) Exn. Cell. Res. 7~5:304-306]. Far the detection of chromosome replication by bramodeoxyuridine (BrdU] incorporation, the Fluorescein Plus Giemsa (FPG] staining method of Perry & Wolff ((1974) Nature 251:156-1681 was used.
C. Imrnunofabefiing of chromosomes and in s~cc~ hybridization Indirect immunofluarescence label8ing with human anti-centromere serum LU851 (Eiadlaczicy et al. (i986) 6xp. Cell Res. 167:1-15], and indirect immunofluorescence and in situ hybridization on the same preparation were performed as described previously (see, Hadlaczky et ate. (1991) Proc. IVatf. Acad. Sri. U.S.A. $8:8106-13110.
immunoEabellung witn fluorescein-conjugated.anti-BrdU r~tor~oclona! antibody [Boehringer] was performed acCOrding to the procedure recommended by the rnanufaaturer, except that for treatment of mouse A9 chromosomes, 2 M hydrochloric acid was used at 37° C for 25 min, and for chromosomes of hybrid cells, 1 M
hydrochloric acid was cased at 3?° C fQr 30 min.
Scanning electron microscopy Preparation of mitotic chromosomes for scanning electron microscopy using as.mium irnpregvation was performed as described .
previously (Sumreer (199?) Chromosoma 100:410-418], The chromo-somes were abserued with a Hitachi S-80Q field emission scanning electron microscope operated with an accelerating voltage of 25 kV.

E. DNA maniputations, plasrnids and probes 1. General methods -Alt general DNA manipulations were performed by standard procedures [see, eT4., Sambraok et a!. (1989) Molecular alonirag: ~4 Laboratory Manual Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY). The mouse major satellite probe was provided by Dr. J. B.
Rattner [University of Calgary, Alberta, Canada]. Cloned rouse satellite DNA probes[see Wong et al. (1988) Nucl. Acids Res. 16:11f45-11581], including the mouse major satellite probe, were gifts from Dr. J. B.
Rattner, University of Calgary. Hamster chromosome painting was done with total hamster genomic DNA, and a cloned repetitive sequence specific to the centromeric region of chromosome 2 [Fatyoi renal. (194) Nucl. Acids Res. 22:3728-3736] was ats~ used. Mouse chromos~me painting was done with a cloned long interspersed repetitive sequence CpMCPI .51 ] specific for the mouse euchromatin.
For cotransfection and for ire situ hybridization, the pCH1 10 ~i-galactosidase construct CPharmacia or Invitrogen], and etch 875 Sam7 phage DNA [New England Biolabsl were used.
2. Construction of Plasrnid pi~uroT~et 2~ Plasmid pPuroTet, which carries a Puromycin-resistance gene and a cloned 2.5 kb human telomeric sequence [see SEC3. !~ No. 3], was constructed from the pBabe-puro retroviral vector FM~rgenstern et ate.
(1990) Nucl. Acids Res. 18:3587-3596; provided by Dr. ~.. Szekely (Microbiology and Tumorbiology Center, Karolinska Institutet, Stockholm); see, also Tonghua et al. (19951 Chin. Med. J. (Beijing, Engl.
Ed.) 108:653-659; Couta et ail. (1994) Infect. Immun. 62:2375-2378 ~unckley --et a(. (1992) FEBS Lett. 29fi:128-34; French et al. (1995) Anal.
Biochem. 228:354-355; Liu et al. (1995] Blood 85:1095-1103;

-g~.
international PCT application Nos. WO 9520044; WO 9500178, and Wa 9419456) .
F. Deposited cell lines Cell lines KE1-214, EC3/?C5, TF1004G19C5, 19C5xHa4, G3D5 and H 1 D3 have been deposited in accord with the Budapest Treaty at the European Collection of Animal Cell Culture (ECACC) under Accession Nos. 96040924, 96040925, 96040926. 9604092?, 96040928 and .
96040929, respectively. The cell lines were deposited on April 9, 1996, at the European Collection of~ Animal Cell Cultures (ECACCy Vaccine Research and Production Laboratory, Public Health 1_aboratory Service, Centre for Appliced Microbiology and Research, Po~rton ~owns Salisbury, Wiltshire SP4. OJG, United Kingdom: The deposits were made in the name at Gyuia Hadlaczky of H: 6?23, SZEGED, SZAMOS U.1.A. !X. 36.
HUNGARY, who has authorized reference to the deposited ce81 lines in this application and who has provided unreserved and irrevocable consent to the deposited cell lines being made available to the public !n accordance with Rule 28'1)(d) of the European Patent Convention.
EXI~iVfPLE 2 Preparation of EC317, EC3/?C5 and related cell lines The EC3l7 cell line is an i_MTiC' mouse cell line that contains the neo-centramere. Tlie EC3/?C5 cell line 'is a single-cell subclone of EC3/7 that contains the neo-minichromosome.
A. EC317 Cel! line As described in-U:S. Patent No.~ 5,288,625 (see, also Praznovszky et ai. (1991) Proc. f~atl. Acad. Sci. U.S.A. 88:11042-11046 and Hadlaczky et al. (1991) Proc. Nati. Acad. Sci. U.S.A. 88:8106-8110] de novo centromere formation occurs in a transformed mouse LMTK' fibro-blast cell line [EC3/?] after coiritegration of A constructs [~ICM8 and rlgtWESneo] carrying human and bacterial DNA.

13y cotransfection of a 14 kb human DNA fragment cloned in o!
(~CMB~ and a dominant marker gene (~gtlnlESneo~, a selectable centromere linked to a dominant marker gene [neo-centromerel was formed in mouse LMTK'cell line EC~I'7 f~adl~czky et ai. (1891) I'roc.
S Natl. Aced. Sci. U.S.A. 8:8106-81 10, see Figure 1 ]. integration of the heterologous ~NA [the A DNA and marker gene-encading ~NA] occurred into the short arm of an acrocentric chromosome [chromosome 7 (see, Figure 1 B)7, where an amplification process resulted in the formation of the new centromere [neo-centromere (see Figure 1 C)]. Un the dicentric 1~ chromosome (Figure 1 C~, the newly formed centromere region contains al! the heteroiogous DNA (human, ~R, and bacterial? introduced into the cell and an active centromere.
Having two functionally active centromeres on the earns chrorrvosome causes regular breakages between the csntromeres [see, 15 Figure 1 El. The distance between thr two centromeres on the dicentric chromosome is estimated to be -10-15 Mb, and the breakage that separates the minichromosome occurred between 'the two centromeres.
Such specific chromosome breakages result in the appearance [in approximately 10/0 of the cells] of a chromosome fragrgtent that carries 2~ the neo-centromere [Figure 1 F]. This chromosome fragment is princi~atly composed of human, rl, p(asmid, and neomycin-resistance gene ~NA, but it also has some .mouse chromosome! C~NA. Cytological evidence suggests that during the stabilization of the MMCneo, -there was an inverted duplication of the chromosome fragment bearing the 25 neo-centromere. The size of minichro~osomes in cell fines containing the MMCneo is approximately 2~-3C Mb; this finding indicates a two-fold increase in size.
From the EC317 cell line, which contains the dicentric chromosome [Figure 1 E~, two sublines [EC317C5 and ECS/7C6~ ware selected by -8x-repeated single-cell cloning. In these cell lines, the nea-centromere was found exclusively on a small chromosome [neo-minichromosome], while the formerly dicentric chromosome carried detectable amounts of the exogenously-derived DNA sequences but not an active neo-centromere [Figure 1 F and 1 G1.
The minichromosomes of cell lines EC3/7C5 and EC3l7C6 are similar. No differences are detected in their architectures at either. the cytological ar molecular level. The minichromosomes were indistinguishable by conventional restriction endonuclease mapping or by long-range mapping using pulsed field electrophoresis and Southern hybridization. The cytoskeleton of cells of the EC8/7C6 line showed an increased sensitivity to colchicinea so the EC3l7C5 line was used for .
further detailed analysis.
B: Preparation of the EC3l7C5 ar<d EC3l7C6 cell lines The EC3l7C5 calls, which contain the nea-rninichromosome, were produced by subcloning the EC3l7 cell line in high concentrations of 6418 [40-fold the lethal dose] for 360 generations. Two single cell-derived stable cell lines [EC317C5 arid EC3/7C6] were established.
These cell fines carry the neo-centrornere on minichromosomes and also contain the remaining fragment ~f the dicentric chromosome. Indirect immunoftuorescence with anti-centromere antibodies and subsequent in situ hybridization experiments.demonstrated that the minichromosomes derived from the dicentric chromosome. fn EC3lJC5 and EC8I7C6 cell lines (140 and 128-metaphases, respectively) n~ intact dicentric chromosomes were found, and minichromosorraes were detected in 97.2°!o and 98.1 % of the cells, respectively. The minichromosomes have been maintained for over 150 cel4 generations. They do contain the remaining portion of the formerly dicentric chromosome.

Multiple copies of telomeric DNA sequences were detected in the marker centromeric region of the remaining portion of the formerly dicentric chromosome by in situ hybridization. This indicates that mouse telomeric sequences were coamplified with the foreign ~NA sequences.
These stable minichromosome-carrying cell lines pro~ride direct evidence that the extra centromere is functioning and is capable of maintaining the minichromosomes [see, U.S. patent No. 5;288,6253" ' The chromosome breakage in the EC31'~ cells, which separates the neo-centromere from the mouse chromosome, occuawed in the G-band '!~ positive "foreign°' DNA region. This is supported by the observation of traces of .i and human DNA sequences at the broken end of the formerly dicentric chromosome. Comparing the G-band pattern of the chromosome fragment carrying the neo-centromere with that of the stable neo-minichromosome, reveals that the neo-minichromosome is an S inverted duplicate of the chromosome fragment tha?~ bears the neo_ centromece. This is also evidenced by the observation that. although the neo-miniehromosome carries only one Functional eentromere, both ends of the minichwomosome are hater~chromatic, and mouse satellite DNA
sequences were found in these heterochromatic regions by ir~:situ 2~ hybridization.
These two cell lines,. EC3/7C5 and EC3I'7C6, thus carry a selectable mammalian minichromosome [MMCneol with a centromere linked to a dominant marker gene C~adiaczky et al. ('1991 ) Proc, Natl.
Acad. Sci. U.S.A. 88:8105-8110]. MMCneo is intended to be used as a ~5 vector for minichromosdme-mediated gene transfer and has been used as model of a minichromosome-based vector system.
Long range mapping studies of the MMCneo indicated that human DNA and the neomycin-resistance gene constructs integrated into the mouse chromosome separately, followed by the amplification of the chromosome region that contains the exogenous DNA. The MMGneo contains about 30-50 copies of the ,AGM3 and rIgtWESneo DNA in the form of approximately 16~ kb repeated blocks, which together cover at teast a 3.5 Mb region. In addition to these, there are mouse telomeric sequences [Praznovszky et at. (1991 ) Proc. Natl. Acad. Sci. U.S.A.
89:11 ~42-11046) and any DNA of mouse origin necessary far the correct higher=ordered structura! organization of chromatids.
Using a chromosome painting probe mCPE1.51 [mouse long interspersed repeated ~NA~, which recognizes exclusively euchromatic mouse DNA, detectable amounts of interspersed repeat sequences were found on the MMCneo by in situ hybridization. ,The neo-centromere.is associated with a small but detectable amount of satellite DNA. The chromosome breakage that separates the neo-centromere from ttae mouse chromosome occurs in the °'foreign" 1~NA region. This is demonstrated by the presence of a and human DNA at the broken end of the formerly dicentric chromosome: At both ends of the MMCneo, however, there are traces of mouse major satellite DNA as evidenced by in situ hybridization. This observation suggests that tha doubling in size of the chromosome fragment carrying the neo-centromere during the stabilization of the MMCneo is a result of an inverted duplication.
Although mouse telomere sequences, which coampiified with the exogenous DNA sequences during the neo-centromere formation, may provide sufficient telomeres for the MMCneo, the duplication could have supplied the functional telomeres for .the minichromosome.
The nucleotide sequence of portions of the neo-minichrocnosomes was determined as follows. Total DNA was isolated from EC3I7C5 cells according to standard procedures. The DNA was subjected to nucleic acid amplification using 'the Expand Long Template PCR system [Boehringer Mannheiml according to the manufacturer's procedures. The -~5~
amplification procedure required only a single 33-mer oligonucleotide primer corresponding to sequence in a region ~f the phage ~ right arm, which !s contained in the neo-minichromosorne. The sequence of this otigonucleotide is set forth as the first 33 nucleotides of aEG ID No. 18.
Because the neo-minichromosome contains a series of inverted repeats ~f this sequence, the single oligonucieotide was used as a forward and reverse primer resulting in amplification of DNA positioned between sets of inverted repeats of the phage rt DNA. Three products.were obtained from the single amplifseation reaction, which suggests that the sequence 1~ of the DNA located between different sets of inverted repeats may differ.
in a repeating nucleic acid unit within an artificial chromosome, minor differences may be present and may occur during culturing of cells containing the artificial chromosome. For example, base pair changes may occur as wet! as integration e~f mobile genetic elements and deletions of repeated sequences.
Each of the three products was subjected to DIVA sequence analysis. The sequences of the three products are set forth In 5EQ l~
Nos. 13, 14, and 15, respectively. To be certain that the sequenced products were amplified from the neo-minichromosome, control 2~ amplifications were conducted using the same primers on DNA isolated from negative control cel6 lines imouse t-t!c° cells) lacking minichromosomes and the formerly dicentric chromosome;, and positive control cell lines [the mouse-hamster hybrid cell line GB43 generated by treating 19C5xHa4 cells isee Figure 41 with BrdU followed by growth in 6418-containing selective medium and retreatment with BrdU] containing the neo-minichrom~some only. Dnly the positive contr~I cell Line yielded the three amplification products; no amplification product was detected in the negative control reaction. The results obtained in the positive control amplification also demonstrate that the neo-minichromosome °e6-DNA, and not the fragment of the formerly dicentric mouse chromosome, was amplified.
The sequences of the three amplification products were compered to those contained in the Genbank/EMBL database. SEQ~ID Nos. 1.3 aryl 14 showed high(-96°I6) homology to portions of DIVA from intracisternal A-particles from mouse. SEC ID No. 15 showed no significant homology with sequences available in the database. All three of these sequences may be used for generating gene targeting vectors as homologous D(VAs to the neo-minichromosome.
C. Isotation.and partial purification of minichromosomes Mitotic chromos~mes of EC3l~C5 cells were isolated as described by Hadlaczicy et ai. (( 1981 ) Chromosome 81:537-555, using a glycine-hexylene glycol buffer system IHadtaczf~y et al: ( 1982) Chromosome 86:643-6591. Chromosome suspensions were centrifuged at 1, 200 x g for 30 minutes. The supernatant containing.
minichromosames was centrifuged at 5,000 x g for 30 minutes and the pellet was resuspended in the appropriate buffer. Partially purified minichromosomes were stored in 50°Jo glycerol at -20° C.
~. Stability of the MMCneo maintenance and neo expression EC3l7C5 cells grown in non=selective mediurra for 284 days and then transferred to selective medium containing 400 ~rglml 6418 showed a 96% plating efficiency (colony formation) compared to control cells cultured permanently in the presence of 6418. Cytogenetic analysis indicated that the MMCneo is stabty maintained at one copy per cell under selective and non-selective culture conditions. ~nly two metaphases with two MMCneo were found in 2,270 metaphases analysed.
Southern hybridization analysis showed no detectable changes in DNA restriction patterns, and sismilar hybridization intensities were _g7-observed with a neo probe when DNA from cells grown under selective or non-selective culture conditions were compared.
Northern analysis of RNA transcripts from the rreo gene isolated from cells grown under selective and non-setective~conditions showed only minor and not significant differences. Expression of the neo gene persisted in E,C317C5 cells maintained in F-12 medium free of 64:18 for 290 days under non-selective culture conditions. The tong-term expression of the neo genes) from the minichromosome may be influenced by the nuclear location of the MMCneo. In situ hybridization experiments revealed a preferential peripheral location of the MMCneo in the interphase nucleus. Iri more than 60n/o of the 2,5~0 nuclei analyses;
the minichromosome was observed at the perimeter of the nucleus near the nuclear envelope.
EXAfrIIPLE ~
Minichromosome transfer and production of the .1-nea-chromosome 61. Minichromosome transfer The neo-miniehromosome [referred to as MMCheo, FtG. 2C] has been used for gene transfer by fusion of minichromosome-containing cells (EC317C5 or EC~l7C6] with different mamrnaliain cells,.inctuding 2~ hamster and human.. Thirty-seven stable hybrid cell lines have been produced. All established hybrid cell lines proved to be true hybrids as evidenced by in situ hybridization using biotinytated i~uman, and hamster genomic, or pMCPE1,.51 mouse long interspersed repeated DNA probes for "chromosome painting". The MMCneo has also been successfully transferred. into mouse A9, 1.929 and pluripotent F9 teratocarcinoma cel6s by fusion of microcetts derived from EC3f7Ca cells. '~'ransfer was confirmed by PCR, Southern blotting and in situ hybritiiZation with minichromosome-specific probes. The cytogeneti~ analysis confirmed -that, as expected for micracell fusion; a few cells I1-5%J received for retained] the MMCneo.
These results demonstrate that the MMCneo is.tolerated by a ~nride range of cells: The prokaryotic genes and the extra dosage for the' human and ~l sequences carried on the minichromosome seem to be not disadvantageous for tissue .culture cells. . ' The MMCneo is the smallest chromosome of the EC~17C5 genome and is estimated to be approximately 20-3~ Mb, which is significantly smaller than the.majority of the host cell imouse) chromosomes. By virtue of the smaller size, mlnichromosomes can be.,partially purified from a suspension of isolated chromosomes by a simple differential centrifugation, In this way, minichromasome suspensions of 15-20%
purity have been prepared. These enriched minichromosome preparations can be used to introduce, such as by microinjection or 1,5 lipofection, the minichromosome int~ selected target cells. Target cells include therapeutic cells that can be use in methods of gene therapy; ared also embryonic cells for the preparation of transgenic tnon-human) animals.
The MMCneo is capable of autonomous replication, is stably maintained in cells, and permits persistent expression of the neo genei~), even after long-term culturing under non-selective conditions. It is a non-integrative vector that appears to occupy a territory near the nuclear envelope. Its peripheral localization in the nucleus may have an important role in maintaining the functional integrity and. stability of the MMCneo. Erunctional compartmentalization of the host nucleus may have an effect. on the function of foreign sequences. In addition. MMCneo contains megabases of a DNa4 sequences that should serve as a target site for homologous recombination and thus integration ~of desired genets) into the MMCneo. !t can be transferred by cell. and microcell CA 02429726 2003-O(i-09 fusion, microinjection, electroporatisan, lipid-mediated starrier systems or chromosome uptake. The neo-eentromere of the MMCneo is capable of maintaining and supporting the normal segregation of a larger 150-200 Mb .ineo-chromosome. . This result demonstrates that the MMCneo chrorrzosome should be useful for carrying large fragments of heterologou,s DNA.
B. Production of the ~ineo-chromosorne-4n the hybrid cell line KE1-2/4 made by fusion of EC3/7 and Chinese hamster ovary cells [FIG 2], the separation of the neo-1~ centromere from the dicentric chromosome was associated with a further amplification process. This amplification resulted in the, formation of a stable chromosome of average size [i.e., the ~Ineo-chromosome; see, Praznovszky et al. (1991 ) Proc. Natl. Acad. Sci. U.S.A. X8:1 1042-11046]: The .tneo-chromosome carries a terminally located functional 75 centromere and is composed of seven large amplicon~s containing multiple copies of a, .human, bacterial, and mouse DNA sequences [see FIG 2].
The amplicons are separated by mouse major satellite DNA [PraTnovszky e~ al. (1991D Proc. Natl. Acad. Sci. U.S.,A. 85:1104-110461 which forms narrow bands of constitutive heterochromatin between the 21D arnplicons.
E%flIMPLE 4 Formation of the "sausage chromosome" [SC]
The findings set forth in the above EXAMPLES demonstrate that the centromeric region of the mouse chromosome 7 has the capacity for 25 large-scale amplification [other results indicate that 'this capacity is not unique to chromosome 71. This conclusion is further supported by results from cotransfection experiments, in which a second dominant selectable marker gene and a non-selected marker gene were introduced into EC3I7C5 cells carrying the formerly dicentric chromosome 7 and the . -neo-minichromosome. The EC3/7C5 cell line was transformed with r1 phage DNA, a hygromycin-resistance gene construct (pH132j, and a ~3-galactosidase gene construct (pCHl1~]. Stable transformants were selected in the presence of high concentrations (40C~ p~glmlj Hygromycin B, and analyzed by Southern hybridization. Established transfo'mant cell lines showing multiple copies of integrated exogenous DNA were studied by in situ hybridization to localize the integration site(s), and by Lack staining to detect ,B-galactosidase expression.
A. tlAaterials and methods 1. Construction of pH'E3 The pH132 plasmid carries the hygromycin B resistance gene and the anti-HIV-1 gag ribozyme (see, ScQ !D NO. 6 for DNA sequence that corresponds to the sequence of the ribozyme) under control of the ,~-actin promoter. "this plasmid was constructed from pHyg plasmid (Sugden et al. ]1985) EVIoI. Cell. t3iai. 5:410-413p a gift from Dr. A. D.
Riggs, Beckman Research institute, Duarte; sees also, e~o.~ 11.5. Patent No. 4,997,?64j, and from pPC-RAG12 plasmid (see, Chang et al. (199~) Clin Biotech 2:2.3-31; provided by Dr. J. J. Rossi; Beckman Research Institute, Duarte; see, also 1.!.S. Ratent Nas. 5P2?282.62, 5;1.49,796 and 2~ 5,144,~19, which describes the anti-HIV' gag ribozyme and construction of a mammalian expression vector containing the ribozyme insert lir'ked to the :actin promoter grad SV40 late gene transcriptional termination and polyA signals]. Construction of pPC-RAG12 involved insertion of the ribozyme insert flanked by BamHl linkers was into ~aml-ll~~digested pl-l~-~pr-1gpt (see, Gunning et al. (1987) Proc. NatE. Ae:ad. Sci. l~.S.~,.
X4:4831-4835, see, alses IJ.S. Patent No. 5,14.4;0191.
Plasmid pH 132 was constructed as follows. First, pPC-RAG 12 (described by Chang et al. (i 99D) Ciin. Biotech. 2:~3-31 j was digested with BamHl to excise a fragment containing an anti-HIV ribozyme gene [referred to as r~bozyme D by Char~g~et al. [(1990) Clin. Biotech. 2:23-311; see also U.S. ,Patent No..5,144~019 to Rossi et al.., particularly Figure 4 of the patent] flanked by the human ~-actin promoter at the 5°
end of the gene and the SV40 (ate transcriptional termination and polyadenylation signals at the 3' end of the gene. As described by .
Chang et al. [(1990) Clin. Biotech. 2:23-31], ribozyme ~ is targeted for cleavage of the translational initiation region of the HIV gag gene. This fragment of pPC-RAG12 was subcloned into pBluesceipt-KS(+]
[Stratagene, La Jolla., CA] to produce plasmid 132. Plasmid 132 was then digested with Xhol and EcoR! to yield a fragment containing the ribozyme D gene flanked by the ,l3 actin promoter at the 5' end and the SV40 termination and polyadenylation signals at the 3' end of tha gene.
This fragment~was ligated to the largest fragment generated by'digestion of pHyg [Sugden et al. ( 1985) Niol. Cell. Biol. 5:410-~13] with EcoRl and 1 a Sal1 to yield pH 132. Thus, pH 132 is an -- 9.3 kb plasmid containing the following elements: the R-actin promoter linked to are anti-filV ribozyme gene followed by the SV40 termination and polyadenylation signals. the thymidine kinase gene promoter linked to the hygrocnycin-resistance gene followed by the thymidine kinase gene polyaclenylation signal, and the _E.
cdli ColE1 origin of replication and the ampicillin-resistance gene.
The plasmid pHyg [see, e.~., U.S. Patent Nos. 4,997,7f4, 4,686,186 and 5,162,215], which confers resistance to hygcomycin g using transcriptional controls from the HSV-1 tk gene, was originally constructed from pKan2 gates et ai.~(1984) Proc. IVatl. Acad. Sci.
U.S.A. 81:3806-3810] and pLG89 (see, Gritz ~t al. (1983' Gene 25:179-'188). Briefly pKan2 was digested with Smai and Bgtli to remove the sequences derived from transposon TnS. The hygrort~ycin-resistance hph gene was inserted into. the digested pKan2 using blunt-end ligation at the Snal site and "sticky=end" ligation [using 1 V~eiss unit of T4. ~NA

-92,-ligase (BRL) in 20 microliter volume] at the _Bg_Iil site. The Srr~al and X1,1 sites of pKan2 were lost during ligation.
The resulting plasmid pH132, produced from introduction of the anti-HBV ribozyme construct with promoter and polyA site into pHyg, includes the anti-HIV ribozyme under control of the ~-actin promoter as welt as the hygromycin-resistance gene under control of the TK
promoter.
Z. Chromosome Landing Trypsin G-banding of chromosomes was performed as described in EXAMPLE 1.
3. Cell cultures TF1004G19 and TF1004G-19C5 mouse cells and the 19C5xHa~4 hybvid, described below, and its sublines were cultured in F-12 cntedium containing 400 ,ugim) Hygromycin B (Calbiochemj.
B. Cotransfection of EC317C5 to produce TF1~Ci4G19 Cotransfection of EC3/7C5 cells with plasmids (pH132, pCH1 10 auaitable from Pharmacia,. see, also Hall et al. (1983; J. Nlol. Appl. Gen.
2: i 01-1091 and with ~ DNA Cacl 875 Sam 7(New England Bioiabs)j was conducted using the calcium phosphate CNA precipitation rrtethod (see, 2~ ela., Chen et a1_ (1987?! Mol. Cell. Blot. x:2745-27a2j, using 2-5 E.rg plasmid DNA and 20 Ng ~ phage DIVA per 5 x 10& recipient cells.
G. Cell lines containing the sausage chromosome Analysis of one of the transformants, designated TF3004G19, revealed that it has a high copy n.umbec of integrated pH132 and pCH 110 sequences, and a high level of ~-galactosidas.e expression. G-banding and in situ hybri~diz~tioct with a human prabe~CM$
revealed une~pe~t.~c~iy that integwati,on had occurred in the formerly clicentric chromosome 7 of the EC317C5 cell line. Furthermore, this chromosome carried a newly fo~rr~ed °~3_ heterochrornatic chromosome arm. The size of this heterochromatic arm varied between ~ 150 and -~ 800 iVib in individual metaphases.
By single cell cloning from the TF1004G19 cell line, a subctone °-TF1004G-19C5 (FIG 2D], which carries a stable chrorrtosorc~e 7 with a --100-150 Mb heterochromatic arm (the sausage chromosome] was obtained. This cell tine has been deposited in the BCACC under Accession No. 960.40926. This chromosome arm is composed of four to five satellite segments rich in satellite L7NA, and evenly spaced integrated heterologous °'foreign" D~IA sequences. At the end of the compact 1~ heterochrorriatic acm of the sausage chromosome, a less condensed euchromatic terminal segment is regutarly observed. This subclone was used for further analyses.
D. Demonstration that the sausage cfiromosmme is derived fr~rva ties formerly dicentric chromasome In situ hybridization with a phase arid phi i 32 DNA on the TF1004G-1905 cell tine showed positive hybridization only on the minichromosome and on the heterochr~matic arm of the °°sausage"
chromosome [Fig. 2D]. !t appears that the "sausage's chromosome (herein afro referred to as the SC] developed from the formerly dicentric cheomosome (FD) of the ~C317C5 cel9 tine.
To establish this, the integration sites of pCH 1 10 and pH 132 piasmids were determined. This was accomplished by in ssru hybridization on these cells with biotin-labeled subfragments of the hygromycin-resistance gene and the ~-gaiactcisidasa gene. Both - experiments resulted in narrow hybridizing bands on the heterochromatic arm of the sausage chromosome. The same hybridization pattern was detected on the sausage chromosome using a mixture of biotin-labeled ~1 probe and pf-1132 plasmid, proving the cointegration of A phases, pF#132 and pCH1 10 plasmids.

-To examine this further, the cells were cultured in the presence of the DNA-binding dye Hoechst 3355. Culturing of mause cells in the presence of this dye results in under-condensation of the pdricentric heterochromatin of metaphase chromosomes, thereby permitting better S observation of the hybridization pattern. Using this technique, the heterochromatic arm of the sausage chromosome of TF1~C~4G-1906 cells shawed regular under-condensation revealing the details of the structure of_the "sausage°' chromosome by fn situ hybridization. Results of in situ hybridization on Hoechst-treated TF1~~4G-1906 cells with biotin-labeled 1~ subfragments of hygromycin-resistance and R-galactosidase genes shows that these genes are localized only in the heterochromatic arm of the sausage chromosome. In addition, an equal bandirig hybridization pattern was observed. This pattern of repeating units [ampiiconsl clearly indicates that the sausage chromosome was formed by an amplification 16 process and that the a phage, phi 1 ~2 and pCFi11 ~ piasmid DNA
sequences border the amplicons.
In another series of experiments using fluorescence in situ hybridization [FISHl carried out with mouse major satellite DNA, the main component of the mouse pericentric heterochromatin, the results 2~ confirmed that the amplicons of the sausage chromosome are primarily composed of satellite DNA.
E. The sausage chromosome has one ce~ntrornere To determine whether mouse centromeric sequences had participated in the amplification process farming the: "sausage'°
~5 chromosome and whether or not the ampficons carry inactive centromeres, in situ hybridization was carried out with mouse minor satellite DNA. Mouse minor satellite DNA is localized specifically near the centromeres of all mouse chromosomes. Positive hybridization was detected in all mouse centromeres including the sacssage chromosome, -which, however, only showed a positive signal at the beginning of the heterochromatic arm.
indirect immunofluorescence with a human anti-centramere antibody (LU 1351 ] which recognizes only functional centromeres i.see, ela., Hadlacaky et at. ~(i 989 Chraniosoma 97:282-288) proved that the sausage chromosome has only one active centromere. The .centromere comes from the formerly dicentric part of the chromosome and co-localizes with the in situ hybridization signal of the mouse minor ~~NA
probe.
F. The selected and non-selected heteroiogous G~fVA in the heterochromatin of the sausage chromosome is expressed 1. ~ High levels of the heterologous genes are expressed The TF100~4G-1905 cei! line thus carries multiple copies of hygromycin-resistance and ~3-galactosidase genes localized only in the heterochromatic~ arm of the sausage chromosome. .The TF1004G-1905 cells can grow very wsll in .the presence of X00 ,ug/ml or even 400 Nglml hygromycin B. CThe level of expression was determined by tVorthern hybridization with a subfragment of the hygromycin-resistance gene and single copy gene.]
The expression of the non-selected ~i-galactos:idase gene in the TF1004G-19C5 transformant was detected with La~;Z staining of the cells. By this method one hundred percent of the cells stained dark blue, showing that there is a high, level of ,E gaiactosidase expression in all of TF 10046-19C5 cells.
2. The heterologous genes that are expressed are in the heterochromatin of the sausage chromosome To demonstrate that the genes localized in the constitutive heterochromatin of the sausage chromosome provide the hygromycin resistance and the LacZ staining capability of TF1004G-19C5 transformants (i.e. ~B-gal expression], PEG-induced cell fusion between TF1004G-19C5 mouse cells and, Chinese hamster ovary ce!!s was performed. The hybrids were selected and maintained in HAT medium containing 6418 [4.00 Ng/m!) and hygromycin [2~0 ~rg/mt], Two hybrid clones designated 19C5xH~3 and 19CaxHa4, which have been deposited in the ECACC under Accession i'to. 96040927,~were selected.
Both carry the sausage chromosome and the minichromosome.
Twenty-seven single ce6l derived colonies of the 13C5xHa4 hybrid were maintained and ~anatyzed as individual subclones. In situ hybridization with hamster and m~use chromes~me painting probes and hamster chromosome 2-specific probes verified that the 19C5xHa4 clone contains the complete Chinese hamster genome and a partial. mouse genome. A!l 19C5xHa4 subclones retained the hamster genome, but different subetones showed different numbers of mouse chromosomes indicating the preferential elimination of mouse chromosomes.
. To promote further elimination of mouse chromosorrAes, hybrid cells were repeatedty treated with BrdU. The BrdU treatments, which destabilize the genorne, result in significant toss of mouse chromosomes.
The t3rdU-treated 19C5xHa4 hybrid ce!!s were divided to three groups.
One group of the hybrid cells (GHI were maintained in the presence of hygromycin (200 Nglrret) and 6418 l400,uglml), and the other two groups of the cells Were cultured under 6418 (G) or hygromycin (H) selection conditions to promote the etiinination of the sausage chromosome or minichromosome.
One month later, single cell derived subctones were established 2a from these three subcultures of the 19C5xHa4- hybrid line. The subclones were monitored by in situ hybridization with biotin-labeled ~
phage and hamster 'chromosome painting probes. Four individual clones [G2B5, G3C5, G~.D6, G2B41 selected in the presenne of 6418 that had lost the sausage chromosome but retained the minichromosome were -found. lJnder hygromycin selection only one subclone (H1D3] lost the minichromosome., In this clone the megachromosome [see Example 5J
was present.
Since hygromycin-resistance and ~3-galactosidase genes were thought to be expressed from the sausage chrom~same, the expression of these genes was analyzed in the four scsbclones that hoc! lost the sausage chromosome. In the presence of ADO ~sg/m! F~ygromycin, one hundred percent of the cells of four individual subclones died. !n order to detect the J3-galactosidase expression hybrid, subclones were analyzed 1~ by LacZ staining. One hundred percent oi; the cells of the four subclones that lost the sausage chromosome also lost the LacZ staining capability.
All of the other hybrid subclones that had not lost the sausage chromosome under the non-selective culture conditions showed positive LacZ staining.
'15 These findings demonstrate that the expression of hygromycin-resistance and /3-galaetosidase genes is linfced to the presence of the sausage chromosome. Results of ire sitar hybridizations show that the heterologous DNA is expressed from the constitutive heterochromatin of the sausage chromosome.
2~ In situ hybridization studies of three other hybrid subclones (C205, G2D1, and G4D5] did not detect the presence of the sausage chromosome. By the LbcZ staining method, some stained cells were ' detected in these hybrid lines, and when these subclones were transferred to hygromycin selection some colonies aurvived. Gytoiogical 25 analysis and in situ hybridization of these hygromycin-resistant colonies revealed the presence of the sausage chromosome, suggesting that only the cells of G2C6, G2D1 and G4D5 hybrids that had not lost the sausage chromosome were able to preserve the hygromycin resistance and ~-galactosidase expression. These results confirmed that the expression of -these genes is linked to the presence of ,the sausage chromosome. The level of ,B-galactosidase expression was determined by the immunobiot technique using a monoclonal antibody.
Hygromycin resistance and ~B-galactosidase expression of the cells which contained the sausage chromosome were provided by the genes localized in the mouse pericentric heterochromatin. This was demonstrated by performing Southern ~NA hybridizations on the hybrid cells that lack the sausage chromosome using PCR-amplified subfragments of hygromycin-resistance and ~-galactosidase genes as probes. None of the subctones showed hybridization with these probes;
however, all of the~anatyzed clones contained the minichrornosome.
Other hybrid clones that contain the sausage chromosome showed intense hybridization with these I~N,~ probes. These results lead to the conclusion that hygromycin resistance and ,8 galactosidase expression of the cells that contain the sausage chromosome were provided by the genes localized in the mouse pericentric heterochromatin.

The gigachromosome As described in Example 4, the sausage chromosome was transferred into Chinese hamster cells by cell fusion. Using Hygromycin 5/HAT and 6415 selection, two hybrid clones 19C5xHa3 and 19C5xHa4 were produced that carry the sausage chromosome. ~n situ hybridization, using hamster and mouse chromosome-painting probes and a hamster chromosome 2-specific probe, verified that clone 7 9C5xHa4 contains a . complete Chinese hamster genome as well as partial mouse gerromes.
Twenty-seven separate colonies of 19C5xHa4 cells were maintained and analyzed as individual subcEones. Twenty-six out of 2~ subclones contained a morphologically unchanged sausage chromosome.

In one subcfone.of the 19C5xHa3 cell line, 19C5xHa~.7 (see Fig.
2E1, the heterachromatic acm of the sausage chromosome became unstable and showed continuous intrachromosoma! growth. In extreme cases, the amplified chromosome arm exceeded 1000 Mb in size dgigachromosome).
EXAMPE.E 6 The stable megachromosome A. Generation of cell lures captaining the megachromosame AI! 19C5xHa4- subclones retained ,a compete hamster genome, but different subclones showed, different numbers of mouse chromosomesy indicating the preferential elimination of mouse chromosomes. As described in Example 4, to promote further elimination of mouse chromosomes, hybrid cells were treated with BrdU, cultured under 64.18 (G) or hygromycin (H) selection conditions followed lay repeated treatment with 10-~ M BrdU far 16 hours and single cell subciones were, established. The BrdU treatments appeared to destabilize the genome, resulting in a change in the sausage chromosome as well. A gradual increase in a cell population in which a further amp6ification had occurred was observed: in addition to the ~ 100-150 Mb heterachromatic arm of the sausage chromosome. an extra centromere and a --150-250 Mb heterochromatic chromosome arm were formed, which differed from these of mouse chromosome 7. By the acquisition afi another euchramatic terminal segment, a new submetacentric chromosome 4megachromosome) was farmed. Seventy-nine individual subclanes were established from these BrdU-treated cultures by single-cell cloning: 4.2 subclones carried the intact megachromosome, 5 subcianes carried the sausage chromosome, and in 32 se~bclones fragments or translocated segments of the megachromosome were observed. Twenty-six subclones that carried the megachromosome were dultured under. non--1 ~0-selective conditions over a two-month period. In 19 out of 26 subclones, the megachromosome was retained. Those subclones which lost the megachromosomes all became sensitive to Hygromycin B and had no ~-galactosidase expression, indicating that both markers were linked to the megachromosome.
Two sublines (G3D6 and H 1 D31; which were chosen far further experiments, showed no changes in the morphology of the megachromosome during more than 1 ~0 generations under selective conditions. The G3D5 cells had been obtained by growth of 19C5x1~a4 cells in 6418-containing medium fiollowed by repeated BrdU treatment, whereas H 1 D3 cells had been obtained by culturing 19C5xHa4 cells in hygromycin-containing medium followed by repeated BrdU .treatment.
B. Structure of the rnegachromosome The .following results demonstrate that, apart from the euchromatic terminal segments, the integrated foreign DNA (and as in the exemplified embodiments, rDNA sequence), the whole megachromosome is constitutive heterochromatin, Containing a tandem array oi~ at least 4.0 t ~ 7.5 Mbl blocks of mouse major satellite DNA [see Figures 2 and 31:
four satellite DNA blocks are organized into a giant palindrome 2~ [ampliconl carrying integrated exogenous DNA sequences at each end, The long and short arms of the submetacentric megachromosome contains 6 and 4 amplicons, respectively. It is of course understood that the specific organization and size of each component can vary among species, and also the chromosome in which the amplification event initiates.
1. The megachromosome is composed primarily of heterochromatin Except for the terminal regions and the integrated foreign DNA, the megachromosome is composed primarily of heterochromatin. This was -1~1-demonstrated by C-banding of the megachromosomen which resuited in positive staining characteristic of constitutive heterochromatin. Apart from the terminal regions and the integrated foreign DNA, the whole megachromosome -appears to be heterochromatic. Mousy major satellite DNA is the main component of the pericentric, constitutive heterochromatin of rr~ouse chromosomes and represents --10~/0 of the total DNA [blaring et ~ (1966] Science 154:791-794-]. Using a mouse major satellite DNA probe for ~n s~tu hylsridiz~ation, str°dng hybridization was observed throughout the megachromosome, except for its terminal 1~ regions. The hybridization showed a segmented pattern: four large blocks appeared on the short arm and usually 4-7 blocks were seen on the long arm. By comparing these segments with the pericentric regions of normal mouse chromosomes that carry --15 Mb tsf major satellite DNA, the size of the blocks of major sateilite DNA on the megachromosome was estimated to be --30 lib.
Using a mouse probe specific to euchromatin [pMCPE1.51; a mouse long interspersed repeated DNA probe]o positive hybridization eneas detected only on the terminal segments of the megachromosome of the H1 D3 hybrid subline. in the G3D5 hybrids, hybridization with a hamster-2~ specific pr~be revealed that several megachromosornes contained terminal segments of hamster origin on the long arm. This observation indicated that the acquisition of the terminal segments on these chromosomes happened in the hybrid cells, and that the Ibng arm of the megachromosome was the recently formed one arm. When a mouse miner satellite probe was used, specific to the centromeres of mouse chromosomes IWong at al. (19991 Nucl. Acids Res. 16°11645-11661], a strong hybridization signal was detected only at the primary constriction of the megachromosome, which colocatized with the positive 102°
immunofluorescence signs! produced with human anti-centromere serum [LU 851 ].
!n situ hybridization experiments with pH132, pCH110, and ~i DNA
probes revealed that all heterologous DNA was located in the gaps between the mouse major satellite DNA segments. each segment of mouse major satellite DNA was bordered by a narrow band of integrated heterologous DNA, except at the second segment of the long arm where a double band of heterologous DttA existed, indicating that the major satellite DNA segment was missing or considerably reduced in size here.
This chromosome region served as a useful cytological marker in identifying the long arm of the megachromosome. At a frequency of 1t~~4, '°restoratian" ofi these missing satellite DNA blocks was observed in one chromatid, when the formation of a whole segtrmnt ~r~ one chromatid occurred.
After Hoechst 33258 treatment (50 Ng/ml for 16 hours), the ' megachromosome showed undercondensation throughout its length except for the terminal segments. This made it possible to~ study the architecture of the megachromosome at higher resolution. In situ hybridization with the m~use major satettgte probe on und~rcondensed megachromosomes demonstrated that the ~ 30 Mb majar satellite segments were composed of four blocks of -7.5 Nib separated from each other by a narrow band of nan-hybridizing sequences [Figure 3].
Similar segmentation can be observed in the large block of pericentric heterochromatin in metacentric mouse chromosomes from the LMTIC' and A9 cell lines.
2. The megachromosome is composed of segments containing two tandem --7.~a l~ltb blocks followed by two inverted blocks Because of the asymmetry in tl7ymidine content between the two strands of the DNA of the mouse major satellite, when mouse cells are grown in the presence-of BrdU-for a single S phase, the constitutive heterochromatin, shows laterat asymmetry after FPG staining. Also, in the 19C5xHa4 hybrids, the thymidine-kinase [Tkj deficiency of the mouse fibroblast cells was complemented by the hamster Tk gene, permitting BrdU incorporation experiments.
A striking structural regularity in the megachromosome was detected using the FPG technique. In both chromatids, alternating dark and fight staining that produced a checkered appearance of the megachromosome was observed. A similar picture was obtained by labelling with fluoi'escein-conjugated anti-BrdU antibody. Comparing , these pictures to the segmented appearance of the. «egachromosome showed that one dark and one light FPG band corresponded to one ~- 3~
Mb segment of the megachromosome. These resulta suggest that the two halves of the -30 Mb segment have an inverted orientation. This Was verified by combining In situ hybridization and immunolabelling of the incorporated BrdU with ffuorescein-conjugated anti-BrdU antibody on the same chromosome. Since the - 30 Idlb segments for ampficons] of the megachromosome are composed of four blocks of mouse major satellite DNA, it can be concluded that two tandem -7.5 111? blocks are followed by two inverted blocks within one segment, Large-scale mapping of megachromosome DP~IA by pulsed-field electrophoresis and Southern hybridization with "foreign°' DNA probes revealed a simple pattern of restriction fragments. Using endonucleases with none, or only a single cleavagd site in the integrated foreign DNA
sequences, foliowed.by hybridization with a hyg probe, 1-4 predominant fragments were detected. Since the megachromosome contains 10-12 amplicons with an estimated 3-g copies of hyg. sequences per amplicon (30-90 copies per megachromosome?. the small number of hybridizing-fragments indicates the hbmageneity of DNA !n the amplified segments.

3. Scanning electron microscopy ~f the rn~gachrosnosome confirmed the above findings The homogeneous architecture of the heterochromatic arms of the w megachromosome was confirmed by high resolution scanning electron microscopy. Extended arms of megachramosomes, and the pericentric heterochromatic region of mouse chromosomes, treated with f-foechst 33255, showed similar structure. The constituti~re heterochrorriatic regions appeared mare campact than the euchromatic segments. Apart from the terminal regions, both arms of the megachroanosome were completely extended, and showed faint grooves, which should correspond to the border of the satellite DIVA blocks in the non-amplified chromosomes and in the' megachromosome. VVithm;xt Hoechst treatment, the grooves seemed to correspond to the ampiicon borders on the megachromosome arms. In addition, centromeres showed a more compact, finely fibrous appearance than the surrounding heterochromatin:
4. The megachra~mosorne of 113 cells contains rRNA gene sequence The sequence of the megachramosome in the regior5 of the sites of integration of the heterolagous DIVA was investigated by isolation of these regions through using cloning methods and-sequence analysis of the resu9ting clones. The results of this analysis repealed that the heterologous DNA was located near mouse ribosomal DIVA gene (i.e., rDf~iA~ sequences contained in the megachromosome.
a. Ctoning of raglans of the megacl~ramosomes in which heterologous DNA had integrated Megaahromosornes were isolated from 1 B3 cells (which were generated by repeated Brdl7 treatment and single cell cloning of H 1 xHE~.1 cells (see Figure 4~ and which contain a tr~rncated 3~ megachromosome) using fluorescence-activated cell sorting methods as described herein (see Example 10). Following separation of the SATACs (megachromosomes) from the endogenous chromosomes, the isolated ~megaahromosomes were stared in CH buffer (100 ml~Jf giycine, 1 °/6 hexylene glycol, pH 8_4-8.6 adjusted with saturated calcium , hydroxide solution;, see Example 1 ~) and centrifuged into an agarose bed in 0.5 M EDTA:
Large-scale mapping of the megachramosome around the area of the site of integration of the heteralogaus DNA revealed that it is enriched in sequence containing rare-cutting enzyme sites, such as the recognition site for Notl. Additionally, mouse major satellite DNA (which makes up the majority of the megachrarnosome) does not contain Notl recognition sites. Therefore, to facilitate isolation of regions of the megachsomosorrte associated with the site of integration of the heteralogous DNA, the isolated megachromasomes were cleaved with otl, a rare cutting restriction endonuciease with an 8-by CSC recognition site. Fragments of the megachromosome were inserted into plasmid pWE15 (Stratagene, La Jolla, California) as follows. Half of a 10~-NI lom melting point agarose'black (mega-plugl containing the isolated SATACs was digested with Notl overnight at 37°C. Plasmid pWE15 was simitar4y 2Q digested with Notl overnight. The mega-plug was then melted and mixed with the digested plasmid, ligation buffer and T4 ligase. t_igation was conducted at 16°C overnight. Bacterial C3H5g cells were transformed with the ligation product and transformed cells ware plated onto LBIAmp plates. Fifteen to twenty colonies were grown on each-plate for a total of 189 colonies. Plasmid DNA was isolated from colonies that survived growth on t_BlAmp medium and was analyzed by Southern blot hybridization for the presence of-DNA that hybridized to a pUCl9 probe.
This screening methodology assured that all c9canes, even clones lacking an insert but yet containing the pWE15 plasmid, wa~ld be detected. Any clones containing insert DNA would be expected to contain contain non-satellite GC-rich rnegachromosome l3iVA sequences located at the site of integration of the heterologous DNA. Ail colonies were positive for .
hybridizing DNA.
Liquid cultures of alt 189 transformants were used to generate cosmid minipreps for analysis of restriction sites within the insert DNA.
Six of the original 189 cosri~id clones conatained.an insert. These clones were designated as follows: 28 (-9-kb insert), 30 (-- 9-kb insert), 60 ( -4-kb insert), 113 ( -- 8-kb insert). 157 ( -9-kb insert) and 161 ( -- 9-kb insert). Restriction enzyme analysis indicated .that three of the clones (1 13; 157 and 1 &1 y contained the same insert.
b. !n situ hybridization experiments using isola~tett segrraents of the megachromosoene as probes Insert ~NA from clones 30, 913, 157 and 161 was purified, labeled and used as probes in in situ hybridization studies of several cell lines. Counterstaining ofi the cells with propidium iodide facilitated identification of the cytological sites of the hybridization signals. The locations of the signals detected within the cells are summarized in the following table:
CELL 'TYPE PR~BE L~CA'1"!ON ~F 51(;NAI.

Human Lymphocyte No. 161 4-5 pairs of acrocentic chromosomes (mate) at centrameric regions.

Mouse Spleen No. 161 Acracentric ends of 4 pans of chromosomes.

EC317C5 Cells No. 161 Minichramosame and the . end of the formerly dicentric chromosome.

Pericentric. heterochromatin of one of the metacentric mouse chromosomes.

Centromeric region of some of the other mouse chromosomes.

K2p . Na. 3t7 Ends .of at least 6 pairs of Chinese Namster chromosomes. An interstitial ~ signs!

Cells . on a short chromosome.

-1 ~?' HB31 Cells No. 30 Acrocentric ends of at least 12 pairs imouse-hamster of chromosomes. Centromeres hybrid of cells derived certain chromosomes and from H1 D3 the cells by repeated megachrmmosome. Borders l3rdU of the treatment and ampiicons of the megachromosome.
single cell cloning which carries the megaciiromosomeD

Mouse Spleen No. 30 Similar to signal observed Cells for probe no. 161.: Centromeres of 5 pairs of chromosomes. Weak cross-hybridization to periceritric heterochromatin.

HB31 Cells No. 113 Similar to signal observed fos~ probe no. 30.

Mouse Spleen No. 113 Centromeric region of 5 Cells pairs of chromosomes.

K20 Cells iVo. At least 6 pairs of chromosomes.

'. Weak signal at some telomeres and several interspersed signals.

Human Lymphocytehlo. Similar to signal observed 157 for probe I

Cells frnale) no, 161.

c. Southern blot hybridization using isolated segments of the megachromosome as probes DNA was isolated from mouse spleen tissue, mouse LMTK' cells, S K20 Chinese hamster ovary cells, EJ30 human fibroblast ce!!s and H1D3 cells. The isolated DNA and Lambda phage DNA, ores subjected to .
Southern blot hybridization using inserts isolated from megachromosorrge clone rios. 30, 113, 157 and 161 as probes, Plasmid pWElS was used as a negative control probe. Each of the four megachramosome clone inserts hybridized in a mufti-copy manner (as demonstrated by the intensity of hybridization and the number of hybridizing bands to all of the DNA samples, except the lambda phage DNA. Plasmid pWE15 hybridized to lambda DNA only.

ct. Sequence analysis of megachromosome clone no. 161 Megachromosome clone no. 16i appeared to_show the strongest hybridization in the in situ and Southern hybridization experiments and' was chosen for analysis of the insert sequence. 'Fhe sequence analysis was approached by first subctoning the insert of cosmid clone no. 161 to obtain five subclones as follows. ~ _ To obtain the end fragments of the insert of clone no. 161, the clone' was digested with Nott and BamHl and ligated with Notl/BamHl-digested pBluescript KS (Stratagene, La Jolla, Californial. Two fragments of the insert of clone no. 161 were obtained: a 0.2-kb and a 0.7-kb insert fragment. To subcione the interrial fragment of the insert of clone no. 161, the same digest was ligated with BamHl-digested plJC19.
Three fragments of the insert of clone no. 16t were obtained: a 0.6-kb, a 1.8-kb and a 4.8-kb insert fragment.
The ends of all the subcloned insert fragments were first sequenced manually. However; due. to. their extremely high GC content, autoradiographs were difficult to interpret and sequencing was repeated using an ABI sequencer and the dye-terminator cycle protocol. A
2~ comparison of the sequence data to s~quences in the GENBANK
database revealed that the insert of clone r~o. 161 corresponds to an internal section of the mouse ribosomal F~NA gene ti'ONA1 repeat unit between positions 7551-i 5670 as set forth in GENBANK accession no.-X82564, which is provided as SEO ID NO. 16 herein. The sequence data obtained for the insert of clone no..161 is set forth in SEQ. ID NOS.
18-24. Specifically, the individual subclones corresponded to the following positions in GENBANK accessian no. X82564 fi.e., SEQ ID
N~. 161 and in SEC3 1~ N~s. 18-24:

-1 ~9~
SubcloneStart End Site SEe -t~ ~lo.

in 161 7679 7755 Wit, BamHl7 8 k1 t 61 7756 8494 BamH! ! 9 m6 161 8495 10231 l3amHl 20 (shows only sequence m?

corresponding to nt. 8485-8950), 21 (shows only sequence corresponding to nt. 9851-10231!

161m12 10232 15000 BamHl 22 (shows only sequence corresponding to nt. 10232-106001, 23 (shows only sequence corresponding to nt. 14267-150001, 161 ~ 15001~ 15676' lVatl, ~ 24 k2 8amH1 The sequence set forth in SEG !D NUs. 18-24 diverges in some positions from the sequence presented in positions 7;151-1567~ of GENBANK accession no. X82564. Such divergence rnay be attributable to random mutations between repeat units of rDNA. The results of the sequence analysis of clone no. 161, which reveal that it corresponds to rDNA, correlate with the appearance of the in situ hybridization signal it 16 generated in human lymphocytes and mouse spleen cells. -fhe hybridization signal was clearly observed on acrocentric chromosomes in these cells, and such types of chromosomes are known to include rDNA
adjacent to the pericentric satellite DNA on the short arm of the chromosome. Furthermore, rRNA genes are highly conserved in 2~ mammals as supported by the cross-species hybridization of clone no.
161 to human chromosomal DNA.
To isolate amplification-replication control regions such as these found in rDNA, it may be possible to subject DNA isolated from megachromosome-containing cells, such as H1 D.3 cells~ tn nucleic acid 2.5 amplification using, e.g., the polymerese chain reaction (PDR~ with the following primers:

-11 ~-amplification control element forward primer (1-30) 5'-GAGGAATTCCCCATCCCTAATCCAGATTGGTG-;3' (SEA ID NO. 25) amplification control element reverse primer d2142-2111 ) 5'-AAACTGCAGGCCGAGCCACGTCTCTTCTGTGTTTG-3' {SEA (D I~!~. 26$
origin of replication region forward primer (2116-2141 ) 5'-AGGAATTCACAGAAGAGAGGTGGCTCGGCCTGC-3' ISEQ ID NO. 2'7) origin of replication region reverse primer (5~4-6-552'11 5'-AGCCTGCAGGAAGTCATACCTGGGGAGGTGGCCC-3' (SEQ ID NO. 28) C, Summary of the formation of the megachromosome 1~ Figure 2 schematically sets forth events leading to the formation of a stable megachromosame beginning with the generation of a dicentric chromosome in a mouse Lli~fTK~ cell line: (A) A single E-type amplification in the centromeric region of the mouse chromosome ~ following transfection of LMTK- cells with ~CNl8 and ~IgtVIIESneo generates the neo-centromere linked to the integrated foreign DNA, and forms a dicentric chromosome. N6ultiple E-type amplification forms the ~lneo-chromosome, which was derived from chromosome 7 and stabilized in a mouse-hamster hybrid cell line; (S) Specific breakage between the centromeres of a dicentric chromosome ~ generates a chromosome ' fragment with the neo-centromere, and a chromosome ~ with traces of foreign DNA at the end; tC) Inverted duplication of the fragment bearing the neo-centromere results in the formation of a stable neo-minichromosome; (~) Integration of exoe~enous DNA into the foreign DNA
region of the formerly dicentric chromosome 7 initiates H-type 2S amplification, and the formation of a heterochromal:ic arm. Sy capturing a euchromatic terming! segment, this new chromosome arm is stabilized in the form of the °°sausage" chromosome; (E) Brdld treatment andJor drug selection appears to induce further H-type amplification, which results in the formation of an unstable gigachromosome: (F) F~epeated BrdU treatments and/or drug selection induce further ~°i-type amplification including a centromere duplication, which leads to the formation of another heterochromatic chromosome arm. It is split off from the chromosome 7 by chromosome breakage and acquires a terminal segPnent to form the stable megachromosome.
t7. Expression of Q-galactosidase and hygromycin transfarase genes in cell lines carrying the megachromosome or derivatives thereof The Level of heterologous gene (i.e., ~3-galactosidase and hygromycin transferase genesl expression in cell fines containing the 1~ megachromosome or a derivative thereof was quantitatively measured The relationship between the copy-number of the heterologous genes, and the level of protein expressed therefrom was also determined.
1. Materials and methods a. Cell lines Heterologous gene expression levels of H 1 DB cells, carfying a 25C~-4~0 Mb megachrom~scime as decribed abo~.re, and rriM2C1 cells, carrying a 50-60 Mb micro-megachromosome, were quantitatively evaluated, mM2C1 cells were generated by repeated BrdU treatment and single cell cloning of the HlxHe41 cell tine (mouse-hamster-human hybrid 2~ cell line carrying the megachromosome and a single human chromosor~re with CD4 and neoP genes; see Figure 4). The cell lirees were grown under standard conditions in F12 medium under selective ('12~ ~glml hygromycin) or non-selective conditions.
b. E~reparation of cell extract for /3-galactosidase assays Monolayers of mM2~1 or H1 DO rail cultures were washed three times with phosphate-buffered saline (PBB'. Cells were scraped by rubber policemen and suspended and washed again in PBB. Washed cells were resuspended into 0.25 M Tris-HC1, pH 7.8, and disrupted by three cycles of fceezing in liquid nitragen and thawing at 37°C. The extract was clarified by centrifugation at 12,000 rpm for 5 min. at 4°C.
c. ~-galaco~iid~se assay The ~-galactosidase assay mixture contained 1 mM MgCl.a, 45 mM ~-mercaptoethanol, 0.8 mglml o-nitrophenyl-~3-D-galactopyrano-side and 66 mM sodium phosphate, pH 7.a. After incubating the reac-tion anixture with the cell extract at 37 ° C for increasing time, the reac-Lion was terminated by the addition of three volumes of 1 M Na2C03, and the optical density was measured at 420 nm. Assay mixture incubated without cell extract was used as a control. The linear range of the reac-tion was determined to be between 0.1-0.8 OD4zo, One unit of ~'-galac-tasidase activity is defined as the amount of enzyme that will hydrolyse 3 nmofes of o-nitropheny!-,B-D-galactopyranoside in 1 minute at 37°C, d. Preparation of cell extract for hygromycin phosphotransferase assay C~Ils s~,~ere v~aashed as ~tescribed above and resuspended into 20 mM Hepes buffer, pH 7..3, 100 mM potassium acetate, a mM Mg acetate and 2 mM dithiothraitol~. Cells were disrupted at OGC by six 10 sec bursts in an MSE ,ultrasonic disintegrator using a microtip probe. Cells were allowed to cool for 1 min after each ultrasonic burst. The extracts were clarified by centrifuging for 1 min at 2000 rpm in a microcentrifuge.
e. Hygromycin phosphotrarsferase assay Enzyme activity was measured by means of the phosphocelluiose paper binding assay as described by Haas and ~owding [(°1975). Meth.
EnZymol. 43:611-628]. The cell extract was uppiemented with 0.1 M
ammonium chloride and 1 mM adenosine-y-32P-triph~osphate 4specific activity: 300 Cilmmol). The reaction:was initiated by the addition of 0.1 .
mg/ml hygromycin and incubated for increasing time at 37°C. The reaction was terminated by heating the samples for 5 min at 75°C in a -31 ~~
water bath, and after removing the precipitated proteins by centrifugation for 5 min in a microcentrifuge, an aliquot of the sdpernatant was spotted on a piece of Whatman P-S~ phosphocellulose paper ~~ cmz). After 3~
sec at room temperature the papers are pieced into 5C7~ ml of hot (75~~) distilled water for 3 min. White the radioactive ATP remains in solution under these conditions, hygromycin phosphate binds strongly and quantitatively to phosphocellulose. The papers are rinsed 3 times in 5Dt7 ml of distilled water and the bound radioactivity was measured in toluene scintillation cocktail tn a Beckman liquid scintillation counter. Reaction 11g mixture incubated without added hygromycin served as a control.
f: Deterrraination of the copy-number of the heterotogous genes DNA was prepared from the H1 D3 and mM2C~ cells using standard purification protocols involving SDS lysis of thd cells followed by Proteinase K treatment and phenollchlbroform extractions. The isolated DPdA was digested with an appropriate restriction endonucfease, fractionated on agarose gels, blotted to nylon filters and hybridized with a radioactive probe derived either from the (3-galactosidase or the hygromycin phosphotransferase genes. The level of hybridization way quantified !n a Molecular Dynamics Phosphorlmage Analyzer. To control the total amount of ~NA loaded from the different cells tines, the filters were reprobed with a single copy gene, and the hydridization of ,B-gaiactosidase and hygromycin phosphotransferase genes was normalized to the single copy gene hybridization.
~5 g. ~etermination of protein concentration The total protein content of the cell extracts was rneasured by the Bradford colorimetric assay using bovine serum albumin as standard.

2. Characterization of the j3-gatactosiidase and hygrumycin phsophotransferase activity expressed in H1D3 and mM2C1 cells In order to establish quantative conditions, the most important kinetic parameters of ~ galactosidase and hygromycin phosph~transferase activity have been studied. The ~3-galactosidase activity measured with this colorimetric assay was linear between the 0.1-0.8 OD4zo range both for the nM2C1 and H1D3 a~ell lines. The /3-gafactosidase activity was also proportional in both cell lines with the amount of protein added to the reaction mixture within 5-100 erg total protein'concentration range. The hygromycin phosphotransferase activity of nM2C1 and H1 D3 cell lines was also proportional with the reaction time or the total amount of added cell extract under the conditions described for the ,B galactosidase.
a. Comparison of //3-galactosidase activity of rnM2G1 and H 1 D3 cell lines Celt extracts prepared from logarithmically growing mM2Cl and H1D3 call lines were tested for ~i-galactosidase activity, and the specific activities were compared in 10 independent experiments. The ,B-galactosidase activity of H1D3 cell extracts was 44f3~25 Ulmg total protein. Under identical conditions the,l3-galactosidase activity of the mM2C1 cell extracts was 4.8 times lower: 92~13 U/mg tataf protein.
~~Q-galactosidase activities of highly subconfluent, subconfluent and nearly confluent cultures of H 1 D3 and mM2C 1 cell lines were also compared. In these experiments different numbers of logarithmic H1D3 and mM2C1 cells were seeded in constant volume of culture medium and grown for 3 days under standard conditions. No significant difference was found in the ~-galactosidase specific activities of cell cultures grown at different cell densities, and the ratio of HlD3ImM2C1 ~-galactosidase specific activities was also similar for all three cell densities. In confluent, stationary cell cultures of H103 or mM2C1 cells, however, the expression of (3-galactosidase significantly decreased due fil~ely to cessation of cell division as a result of contact inhibition.
b. C~mparison of hygromyoin phosphotransferase activity of H1D3 and mM2C1 cell lines The bacterial hygromycin phosphotransferase is present in a membrane-bound form in 1i1D3 or mM2C1 cell fines. This follows from the observation that the hygromycin phosphotransferase activity can be completely removed by high speed centrifugation of these cell extracts, 1~ and the enzyme activity can be recovered by resuspending the high speed pellet.
The ratio of the enzyme's specific activity in H1D3 and mM2C1 cell lines was similar to that of /3-galactosidase activity, i.e., H1 D3 cells have 4.1 times higher specific activity compared with mM2.C1 cells.
~. Hygrdrrnycln phosphotransferase activity in H1D3 and mM2C1 cells grown under n~n-selective conditions The level of expression of the hygromycin phosphotransferase gene was measured on the basis of quantitation of the specific enzyme activities in H1D3 and mM~C1 cell lines grown under non-selective conditions for 3~ generations. The absence of hygromycin in the medium did not influence the expression of the hygromycin phosphotransferase gene.
3. ~tuantitation of the number of O3-galactosidase and hygromycin phosphotransferase gene copies in H1D3 and mM2C1 cell lines As described- above., the Q-galactosidase and laygromycin phosphotransferass genes are located only withsn thse megachramosome, or micro-megachromosome in H'! D3 and mM2C1 cells. Quantitative analysis of genomic Southern blots of Dl~~l isolated from H1 D3 and 3~ mM2Cl cell lines with the phosphorlmage Anaiy~erT~' revealed that the copy number of (3-gaiactosidase genes integrated into the megachromo-some is approximately 1C) times higher ifs H1D3 cells than in mM2C1 cells. The copy-number of hygre~mycin phosphotransferase genes is approximately 7 times higher in X11 D3 cells than in rnM2C1 cells.
4. Summary and conclarsions of results of quantitation of heterologous gene exg~ression in cells containing megachremeosomes or derivatives thereof Quantitative determination of ~-galactosidase activity of higher eukaryotic cells (e~g., H1 D3 cellsy carrying the bacterial ,S-galactosidase 1~ gene in heterochromatic megachromosomes confirmed the observed high-level expression of the integrated bacterial gene detected by cytological staining methods, It has generally been established in reports of studies of the expression of foreign genes in transgenic animals that although transgene expression shows correct tissue and developmental specificity, the level of expression is typically low and shows extensive position-dependent variabliity (i.e., the level of transgene expression depends on the site of chromosomal integration). ft is generally assumed that the low-level transgene expression may be due to the absence of special DNA sequences which can insulate the transgene from the inhibitory effect of the surrounding chromatin and promote the forrt9atie~n of active chromatin structure required for efficient gene expression.
Several cis-activing DNA sequence elements have been identified which can abolish this position. dependent variability, and can ensure high-level expression of the transgene locus activing region iLAR) sequences in higher eukaryotes and specific chromatin structure uses) elements in lower eukaryotes (see, e.g., Eissenberg and Elgin 11991) Trends in Genet. 7:335-340). If these cis-acting DIVA sequences are absent, thb level of transgene expression is low and copy-nurciber independent.
Although the bacterial ~3-galactosidase reporter gene contained in the heterochromatic megachromosomes of ~1D3 and mM2C1 cells is - °! 1 '7-driven by a potent eukaryotic promoter-enhancer element, no specific cis-acting DNA sequence element was designed and incorporated into the bacterial DNA construct which could functior9 as a boundary element.
Thus, the high-level ~-gaVactosidase expression measured in these cells is of significance, particularly because the Q-galactosidase gene in the megachromosome is located in a long, compact heterochromatic environment, which is known to be able to block gene expression. The megachromosome appears to contain DNA' sequence element(sl in association with the bacterial DNA sequences that function to override the inhibitory effect of heterochromatin on gene expression.
The specificity of the hater~logous gene expression in the megachromosome is further supported by the observation that the level of ~3-gafactosidase expression is copy-number dependent. In the H1D3 cell line, which carries a full-size megachromosome, the specific activity of;t3-galactosidase is abaut 5-fold higher than in mM~C'I cells, which carry only a smaller, truncated version of the megachromosome. A
comparison of the number of ~B-galactosidase gene copies in H1D3 and mM2C1 cell lines by quantitative hybridization techniques confirmed ttaat the expression of ~3-galactosidase is copy-number dependent. The number of integrated J3-galactosidase gene copies is approximately 1 ~-fold higher in the H1 D3 cells than in mM~C1 cells: Thus, the cell line containing the greater number of copies of the (~-galactosidase gene also yields higher levels of ,B-galactosidase activity, which supports the copy-number dependency of expression. The copy number dependency of the ,~-gaiactosidase and hygromycin phosphotransferase enzyme levels in cell lines carrying different derivatives of the megachrorraosome indicates that neither the chromatin organization surrounding the site of integration of the bacterial genes, nor the heterochrornatic environment of the megachromosome suppresses the expression of the genes.

-1.18-The relative amount of (3-gaiactosidase protein expressed in H1 D3 cells can be estimated based on the V,~ax of this enzyme [500 for homogeneous, crystallized bacterial (3-galactosidase ~Naider et al. (19721 Biochemistry 1 1:3202-3210] .and the specific activity of H1 D3 cell protein. /~ V~,a% of 500 means that the homogeneous ~ galactosidase protein hydrolyzes 500 ~rmoles of substrate per minute per mg of enzyme protein at 37°C. ~ne mg of total H1 D3 cell protein extract can hydrolyze 1.4. ~rmoies of s~sbstrate per minute at 37°C, which means that 0.28% of the protein present in the H1 D3 cell extract is ~-galactosidase.
10~ The hygromycin phosphotransferase is present in a membrane-bound form in H1 D3 and rx~M2C1 cells. The tendency of the enzyme to integrate into membranes in higher eukaryotic cells may be related to its periplasmic localization-in prokary~tic cells. The bacterial hygromycin phosphotransferase has not beerB purified to homogeneity; thus, its V,~a ~5 has not been determined. Therefore, no estimate can be made on the total amount of hygromycin phosphotransferase protein expressed in these cell lines. The 4-fold higher specific activity of hygromycin phosphotransferase in H 1 D3 cells as compared to m2C1 cells, however, indicates that its expression is also copy number dependent.
20 The constant and high level expression of the J3-galactosidase gene in H1D3 and mM2Cl cells, particularly-in the absence of any selective pressure for the expression of this gene, clearly indicates the stability of the expression of genes carried in the heterochromatic megachromo-somes. This conclusion is farther supported by the observation that the 25 level of hygromycin phosphotransferase expression did not change when H1D3 and mM2C1 cells ~rvere.grown under non-selective conditions. 'The consistent high-level, stable, and copy-number dependent expression of bacterial marker genes clearly indicates that the megachromosome is an ideal vector system far expression of foreign genes.

1 ~"
EXAnlIPLE 7 Summary of soFne of the cell lines with SATACS and minlchromosomes that have been constructed 6 1. EC3/7-Derived cell lines The LMTK'-derived cell line, which is a mouse fibrobiast cell line, was transfected with ~Cf~it8 and agtV~IESneo DNA'[see, EXAMPLE 21 to produce transformed cell lines. Among these cell lines was EC3/7, deposited at the European Collection of Animal cell Culture tECACC) .
under Accession No. 900151001 [see, U.S. Patent Nca. 5,288,625; see, also Hadlac~ky et al. 11991] Proc. Nati. Acad. Sci. U.S.A. 88:8106-811fl.
This cell line contains the dieentric chromosome with the neo~centrr~mere. Recrloning and selection produced cell lines such as ~03/705, which are cell lines with.the stable neo-minichromosorne and the formerly dicentcic chromosome tree, Fig.
2C] .
2. KE 1-2/4 Celts Fusion of EC3/7 with CHO-K20 cells and selection with G4.18/HAT
produced hyhrid cell lines, among these was KE1-21~, which has been deposited with the ECACC under Accession .No. 96040392. K1'1-2!4' is a stable cell line that contains the ~lnea-chromosome see, Fig. 2D; see, also U.S. Pateht No, 5,288,625]o produced by E-type amplifications.
KE1-2l4 has been transfected with vectors containin g ~1 D~IA, selectable markers, such as the puromycin-resistance gene, and genes of interest, such as p53' and the anti-!-iiV ribozyme gene. These vectors target the gene of .interest into the ~neo-chromosome by virtue of homologous reeombinatio~t with the heteroiogous DNA in the chromosome.

-12~-3. CSpMCT53 Cells The EC3I7C5 cell fine has been co-transfected with pH 132, pCH110 and a ~NA [see, EXAMPLE 21 as welt as oi:her constructs.
Various clones and subclones have been selected. For example transformation with a construct that includes p53 encoding ~NA, produced cells designated CSpMCT53.
4. TF1004G24.Cells As discussed above, cotransfection of EC317C5 cells with plasmids [pH~132, pCH110 available from Pharmacia, see, also Half et al.
(1983) J. Mol: Aar~l. Gen. 2:101-109] and with ~1 DNA [~1c1 13T5 Sam '~
flew England Biolabs)1 produced transformed cells. Among these is TF1004G24, which contains the ~NA enc~ding the.anti-F11V ribozyme in the neo-minichromosome. l~ecloning of TF1004G24 produced numerous cell lines. Among these is the NHHL24 cell line. This cell line also has the anti-HIV ribozyme in the neo-minichromosome and expresses high levels of ,B-gal. It has been fused with CHI-1C20 cells to produce various hybrids.
6. TF1004G19-derived cells Recloning and selection of the TF1004G transformants produced the cell line TF1004G19, discussed above in EXAMi~LE 4, which contains the unstable sausage chromosome and .the neo-minichromosome. Single cell cloning produced the TF1004G-19C5 [see Figure 4] cell line, which has a stable sausage chromosome and the neo-minchromosome. TF1004G-19C5 has been fused with CHO cells and the hybrids-grown under selective conditions to produce the 19C5xHa4 and 19C5xHa3 cell lines [see, EXAMPLE 41 and others. Recloning of the 19C5xHa3 cell line yiefded a cell fine containing a gigachromosome, i.e., cell line 19C5xHa47, see Figure 2E. BrdU treatment of l9CaxHa4 cells and growth under selective conditions [neomycin (G) andlor hygromycin (H91 has produced hybrid cell lines such as the G3D5 and G4D8 cell lines and others. G3D5 has the neo-minichromosome and the megachromosame. G4D6 has only the neo-minichromosome.
Recl~ning of 19C5xHa4 cells in H medium produced numerous a ciones. Among these is H1 D3 Csee Figure 4[,which has the stable megachromosome. Repeated BrdU treatment and recloning of H 1 D3 cells has .produced the HB31 cell line, which has been used for transformations with the pTEMPUD, pTEMPU, pTEMPU3, and pCEPIlR-132 vectors [see, Examples 12 arsd 14, below).
H1D3 has been fused with a CD4'' Heia cell line that carries DNA
encoding CD4 and neomycin resistance on a plasmid [see, e.c~., U.S.
Patent Nos. 6,413,91 ~, 5,409,810, 5,268.600, ~ 5,2.23, 263, 5,216,914 and 5,144,019, which describe these Hela cells]. Selection with GH has produced hybrids, including H1xHE41 [see Figure 4~, which carries the megachramosome and also a single human chromosome that. includes the CD4neo construct. Repeated BrdU treatment and .single cell cloning has produced cell Lines with the megachromosome [cell line 1B3, see Figure 4l. About 25% of the 1 B3 cells have a truncated megachromosome [ - 90-120 Mb3. Another of these subclones, designated 2Cb, was cultured on hygromycin-containing medium and megachromosome-free cell lines were obtained and grown in 64.18-containing medium. Recloning of these cells yielded cell lines such as 1B4 and others that have a dwarf megachromosome [ ~ 150-200 Mbl, and cell lines, such as 11 C3 and mtVi2Cl, which hare a micro-megachromosome C - 50-90 Mb]. The micro-megachromosome of cell -122- .
,.
line mM2C1 has no teiomeres; however, if desired, synthetic telomeres, such as those described and generated herein, may be added to the mM2C1 cell micro-megachromosomes. Cell lines containing smaller truncated megachromosomes, such as the mM2C1 cell line containing 6 the micro-megachromosome, can be used to generate even smaller megachromosomes, e.g., ~ 1 O-30 Mb in size. This may be accomplished, for example, by breakage and fragmentatiow of the micro-megachroinosome in these cells through exposing the cells t~ X-ray irradiation, BrdU or telomere-directed in vivo chromosome fragmentation.

Replication of the megachromosome The homogeneous architecture of the megachromomes provides a unique opportunity to perform a detailed analysis of the replication of the constitutive heterochromatin.
A. Materials and methods 'I. Culture of cell limes ' H1D3 mouse-hamster hybrid cells carrying.the megachromosome (see, EXAMPLE 4J were cultured .in F-12 medium containing 10°6 fetal calf serum (FCS) and 400 ~uglml Flygromycin 8 [Calbs'ochem~. G3D5 hybrid cells [see, Example 4l were maintained in F-12 mecfrum containing 109~o FCS, 400 pglml Hygromycin B (Calbiochem), and 400 ~uglml 6418 [SIGMA]. Mouse A8 fibrobtast Cells were cultured un F-12 medium supplemented with 10~/o FCS.
2. BrdU labelling In.typical experiments, 20-24 parallel semi-confluent cell cultures were set up in 10 cm Petri dishes. , Bromodeoxyuridine (BrdU) (Fluke) was dissolved in distilled water alkalized with a drop of NaOH, to make a 1 O'2 M stock solution. Aliquots of 1 ~-5O ~I of this l3rdL3 stock solution were added to each 10 ml culture, to give a final.BrdU concentration.of 10-50 ;uM. The cells were cultured in the presence c~f .BrdU for 30 min, and then washed with warm complete medium, and incubated without &dU until required. At this point, 5 lrglml cofchicine was added to a sample culture every 1 car 2 h. After 1-2 h colchicine treatment, mitotic cells were collected by "shafts-off" and regular chromosome preparations were made for immunolabelling.
3. linmunolabelling of chrorno.somes and ire sitar hybridization tmmunolabelling with fluorescein-conjugated. anti-BrdU monoclonal antibody (8oehringer) .Was done according to the manufacturer's recommendations, except that for mouse- A9 chromasomes, 2 iVl hydrochloric acid was used at 37° C for 25 min, while for chromosomes of hybrid cells, 1 6V1 hydrochloric acid was used at 37° C
.
for 30 min. In silo hybridization with biotin-labelled probes, and indirect immunofluorescence and in silo hybridization on the same preparation;
were performed as descritaed previously fHadiaczky et al. (1991) Proc.
Natl. Acad. Sci. U.S.A. 88;$106-8110, see, also U.S. Patent No.
5,288.625:
4. Microscopy All observations and microphotergraphy were made by using a Va.noxTM AH-BS EOEympu~s) microscope. l;ujicolorT~' 400 Super G or FujicolorT""
1600 Super NC high-speed colournegatives were used for photographs.
8. Results The replication of the megachromosorrte was analyzed by BrdU
pulse labelling followed by immunolabelling. The basic parameters for DNA labelling in viva were first established. Using a 30-train pulse of 50 ,~M BrdU in parallel cultures, samples were. taken and fixed at 5 mia~
intervals from the beginning of the pulse, and every 15 min up to 1 h after the removal of BrdU. Incorporated >3rdU was.detected by ° 124' imrnunolabelling with fluorescein-conjugated anti-SrdU monoclonal antibody. At the first time point (5 rein) 3800 of the nuclei were labelled, and a gradual increase in the number of labelled nuclei was observed during incubation in the presence of BrdU, culminating in 46°i6 in the min sample, at the time of the removal of BrdU. At further time points (60, 75, and 90 min] no significant changes were observed, and the fraction of labelled nuclei remained constant [44.5-46%].
These results indicate that (i) the incorporation of the BrdU is a rapid process, (ii) the 30 min pulse-time is sufficient for reliable Labelling of S-phase nuclei; and (iii) the BrdU can be effectively rernoved from the cultures by washing:
The length of the cell cycle of the H1 D3 and G3D5 cells was estirriated by measuring the time between the appearance of the earliest BrdU signals on the extreme late replicating chromosome segments and the appearance of the same pattern only on one of the chromatids of the chromosomes after one completed cell cycle. The length of Ca2 period was determined by the time of the first detectable BrdU signal ow prophase chromosomes and by the tabetled mitoses method [fl.astler et al. (1959? Exp. Cell Res. 17:420-438]. The length of the S-phase was determined in three. ways: (i) on the t~asis of the length of cell cycle and the fraction of nuclei labelled during the 3C~-120 min pulse; (ii! bY
measuring the time between the very end of the replication of the extreme late replicating chromosomes and the detection of the first signal on the chromosomes at the beginning of S phase; (iiij by the labelled mitoses method. In repeated experiments, the duration of the cell cycle was found to be 22-26 h; the S phase 10-14 h, and the G2 phase 3.5-4.5 h.

' Analyses of the replication of the megachromosome were made in parallel cultures by collecting mitotic cells at two hour intervals following two hours of colchicine treatment. in a repeat experiment, the same analysis was performed using one hour sample intervals and one hour colchicine treatment. Although the two procedures gave comparable results, the two hour sample intervals were viewed as more appropriate since approximately 30% of the cells were found to have a considerably shorter or~longer cell cycle than the average. The characteristic replication patterns of the individual chromosomes, especially same of the late repiicating hamster chroei~rosomes, served ass useful internal markers far the-different stages of S-phase. T~ minimize the error caused by the different lengths of cell cycles in the different experiments, samples were taken and analyzed throughout the whole cell cycle until the appearance of 'the first signals on one chromatic at the beginnirig of the second S-phase.
The sequence of replication in the megachromosome is as follows.
At the very beginning of the S-phase, the replication of the megachrornosome starts at the ends of the chromosomes. The first initiation of replication in an interstitial position can usually tae detected at ZO the centromeric region. Soon after, but still in the first quarter of the S-phase, when the terminal region of the short arm has almost completed its replicatian, discrete initiation signals appear along the chromosome arms. in the second quarter of the S-phase, as replication proceeds, the BrdU-labelled zones gradually widen, and the checkered pattern of .the megachromosome becomes clear [see, e~a., Fig. 2F]. At the same time, pericehtric regions of mouse chromosomes also show intense.
incorporation of BrdU. The replication of the megachromosome peaks at the end of the second quarter and in the third quarter of the S-phase. At the end of the third quarter, and at the very beginning-of the last quarter 126' of the S-phase, the megachramosome and the pericentric heterochromatin of the mouse chromosomes coriipiete their replication.
By the end of S-phase, only the very late replicating segmments of mouse and hamster chromosomes are still incorporating BrdtJ.
The replication of the inrhole genome occurs in distinct phases.
The signs( of incorporated BrdU increased continuously until the end of the first half of the S-phase, but at the beginning. of the third quarter of the S-phase chromosome segments other than the heterochromatic regions hardly incorporated BrdU.. fn the last quarter of the S-phase, the BrdU signals increased again when the extreme late replicating segments showed very intense incorporation.
Similar analyses of the cepiication in mouse A9 cells were performed as controls. To increase the resolution of the immunolabelling pattern, pericentric regions of A9 chromosomes were decondensed by treatment v~,r'sth Hoechst 33258. Secause of the intense replication of the surrounding euchromatic sequences, precise localization of the initial BrdU signal in the heterochromatin was normally difficul;, even on undercondensed mouse chromosomes. On those chromosomes where the Initiation signat(s) were localized unambiguously; the replication of the pericentric heterochromatin of A9 chromosomes v~ras similar to that of the megachromosome. Chromosomes of A9 cells also exhibited replication patterns and sequences similar to those of the~mouse chromosomes in the hybrid cells. These results irtdicate that the replicatars of the megachromosome and mouse chromosomes retained their original timing and specificity in the hybrid cells.
. By comparing the pattern of the initiation sites obtained. aftdr BrdU
incorporation with the location of the integration sites of the "foreign°' DNA in a detailed analysis of the~first quareer of-the S-phase, an attempt was made to identify origins of replication (initiation sites) in relation to the amplicon structure of the megachromosome,. 'The double band of integrated DNA on the long arm of the megachromosome served as a cytological marker. The results showed a colocali~atiow of the BrdU and in situ hybridization signals found at the cytological level, indicating that the "foreign" DNA sequences are in close proximity to the origins of replication, presumably integrated into the non-satellite sequences between the replicator and the satellite sequences [see, Figure 37. As described in Example 6.B.4., the rDNA sequences detected in the megachromosome are also localized at the amplicon borders at the site of integration of the "foreign" DNA sequences, suggesting that the origins of replication responsible for initiation of replication of the megachromosome involve rDNA sequences. In the pericentric region of several other chromosomes, dot-like BrdU signals can also be observed that are comparable to the initiation signals on the megachromosome.
1 a These signals may represent similar initiation sites in the heterachroi~atic regions of normal chromosomes.
At a frequency of ifl'4, °'uncontrofied°°
amplification of the integrated DNA sequences was observed in the megachromosome.
Consistent with the assumption (above) that '"foreign"' sequences are in proximity of the replicators, this spatially restricted amplification is likely to be a consequence of uncontrolled repeated firings of the replication origin(s) without completing the replication of the whole segment.
C. Discussion It has generally been thought that the constitutive heterochromatin of the pericentric regions of chromosomes is late replicating [see, e~a., Miller ( 1976) Chromosome 55:165-'I Z(5j. ~n the contrary, these experiments evidence that the replication of the heteroehromatic blocks starts at a discrete initiation site in the first half ~f the S-phase and continues through approximately three-quarters of S-phase. This difference can be explained in the fotlo~nring ways: [l) in normal chromosomes, actively replicating euchromatic sequences that surround the satellite DNA obscure the initiation signals, and thus the precise localization of initiation sites is obscured; (ii) replication of the heterochromatin can only be detected unambiguously in a period during the sea~nd half of the S-phase. when the bulk of the heterochromatin replicates and most other chromosomal regions have already completed their replication, or have not yet started it. Thus, tow resolution cytological techniques, such as analysis of incorporation of radioactively labelled precursor's by autoradiography; only detect prominent replication signals in the heterochramatin in the second half of S-phase, when adjacent euch~omatic segments are no longer replicating.
in the megachromosome, the primary initiation sites of replication colocalize with the sites where the "foreign" DNA sequences and rDNA
sequences are integrated at the amplicon borders. SimIIar initiation signals were observed at the same time in the pericentric heterochromatin of some of the mouse chromosomes that do not have "foreign" DNA. indicating that the replication initiation sites at the borders of amplicons may reside in the non-satellite flanking sequences of the satellite DNA blocks. The presence of a primary initiation site at each satellite DNA doublet implies that this large chromosome segment is a single huge unit of replication [megaceplicon] delimited by the primary initiation site and the termination point at each end of the unit. Se~reral fines of evidence indicate that, within this higher-order replication unit, "secondary" origins and replicons contribute to the complete replication of the megarepficon:
1. . The total replication time of the heterochromatic regions of the megachromosome was --9-11 h. At the rate of movement of replication forks, 0.5-5 kb per minute, that is typical of eukaryotic chromosomes (Kornberg et al. (1992) ~11h4 Replication, end. ed.., i~iew York: 1ALH. Freeman and Co, p. x.74.],' replication of a --15 Mb replicon would require 50-500 h. Alternatively, if only a single replication origin was used, the average replication speed would have to be 25 kb per minute to complete replication within 10 h. By comparing the intensity of the BrdU signals on the euchromatic and the heterochromatic chromosome segments, no evidence for a 5- to 50-fold difference in their replication speed was found.-2. Using short BrdU pulse labelling, a single origin of replication 1 ~ would produce a replication band that moves along the replicon, reflecting the movement of the replication fork. In contrast, a widening of the replication zone that fina3ly gave rise to the checkered pattern of the megachromosome was observed, and within the replication period, the most intensive i3rdU incorporation occurred in the second half of the S-phase. This suggests that once the megarepfica~:or has been activated, it permits the activation and. firing of "secondary'° c3riginsa and that the replication of the bulk of the satellite DNA takes place from these "secondary" origins during the second half of the ;~-phase. This is supported by the observation that in certain stages of the repticatior~ of the megachromosome, the whole ampiicon can apparently be labelled by a short BrdU pulse.
Megareplicators and secondary replication origins seem to be under strict temporal and spatial control. The first initiation within the megachromosomes usually occurred at the centromere, and shortly afterward all the megareplicators become active. The last segment of the megachromosome to complete replication wa s usually the second segment of the long arm. Results. of control experiments with mouse A9 chromosomes indicate that replication of the heterochromatin of mouse chromosomes corresponds to the replication of this megachromosome amplicons. Therefore, the pre-existing temporal control of replication ire the heterochromatic Mocks is preserved in the megachromosome.
Positive (Hassan et a!. (1994) J. Cell: Sci. 107:425-4.34. and negative [Haase et ai. (1994) Mof. Celf. Biol. 14:2518-2624 correlations 5. between transcriptional activity and initiation of replication have been proposed. fn the megachromosome, transcription of the integrated genes seems to have no effect on the ~riginal timing of the replication origins:
The concerted, precise timing of the megarepiicator initiations in the different amplicons suggests the presence of specific, cis-acting sequences, origins of replication.
Considering that pericentric heterochramatin of mouse chromosomes contains thousands of short, simple repeats spannirig 1-Mb, and the centromere itself may also contain hundreds of kilobases, the existence of a higher-order unit of replication seems probable. The 7 5 observed uncontrolled intrachromos~mal amplification restricted to a replication initiation region of the megachromosome is highly suggestive of a rolling-circle type amplification, and provides additional evidence for 'the presence of a replication origin in this region.
The finding that a specific replication initiation site occurs at the boundaries of amplicons suggests that replication might play a role in the amplification process. These results suggest that each amplicon of the megachromosome can be regarded as a huge megareplicon defined by a primary initiation site [megareplicatorl containing °'secondary°' origins of replication. Fusion of replication bubbles from different origins of bi-directional replication C~ePamphitis 41993) Ann. Rev. Biachem. 62:29-63) within the megareplicon could form a giant replication bubble, which would correspond to the whole megareplicon. In the light of this, the formation of megabase-size amplicons can be accommodated by a reptication-directed amplification mechanism. In H and ~-type -131°
amplifications, intrachromosomal multiplication of the ~mpficons ~nras observed [see; above EXAMPLES], which is consistent with the unequal sister chromatid exchange model. Induced or spontaneous unscheduled replication of a megareplicon in the constitutive heterochromatin may also form new amplicon[s) leading to the expansion of the amplification or to the heterochromatic polymorphism of ',norms!" chromosomes. The "restoration" of the missing segment on the long arm of the megachromosome may well be the result of the re-replication of ~ne amplicon limited to one strand.
Taken together, without being bound by any theory, a replication-directed mechanism is a plausible explanation for the initiation of large- .
scale amplifications in the centromeric regions of mouse chromosomes, as well as for the de navo chromosome formations. if specific [amplifi-cator, i.e., sequences controlling amplification] sequences play a role in promoting the amplification process, sequences at the primary replication initiation site [megarepticator] of the megareplicon are possible candidates.
The presence of rRNA gene sequence at the amplicon borders near the foreign DIVA in the megachromosome suggests that this seguence contrilautes to the primary replication initiation site and participates in large-scale amplification of the pericentric heterochromatin in de novo formation of SATACs. Ribosomal RNA genes have an intrinsic amplification ri~echanism that provides for multiple copies of tandem genes: Thus, for purposes herein, in the construction of SATACs in cells, r~NA will serve as a region for targeted integration, and as components of SATACs constructed in vitro.

-°t 32-Generation of chromosomes with amplified regions derived from moa~se chromosome 1 To show that the events described in EXAMPLES 2-7 are not unique to mouse chromosome 7 and to show that. the EC713 cel! line is not required for formation of the artificial chromosomes, the experiments have been repeated using different initiat Celt lines and DhIA fragments, Any cell or cell tine should be amenable to use or can readily be determined that it is not.
A. Materials The LP11 cell tine was produced by the "scrape-loading "
transfectiors method (Fechheimer et al. (1987! Proc. Netl. ACad. Sci.
U.S.A. 84:8463-8467] using 25 erg plasmid DIVA for 5 x 10s recipient cells. LP11 cells were maintained in F-12 medium containing 3-15 ~tg/ml Puromycin [SIGMA1.
B. Amplification in LP91 cells The large-scale amplification described in the above Examples is not restricted to the transformed EC317 cell line or to the chromosome 7 of mouse. In an independent transforrrr~ation experiment, LMTK' cells were transfected using the calcium phosphate precipitation procedure with a selectable puromyci_n-resistance gene-containing construct desig-hated pPuroTel [see Example 1.E.2. for a descriptiar~ of this ptasmid], to establish cell line LP11. Cell fine LP'11 carries chromosome(sf with amplified chromosome segments of different lengths [ ~ 150-600 Mb].
Cytological analysis of the LP11 cells indicated that the amplification occurred in the pericentric region of the long arm of a submetacentric chromosome formed by Robertsonian translocation. This chromosome arm was identified by G-banding as chromosome' 1. C-banding and ~n seta hybridization with mouse mayor satellite DNA probe showed that an E-type amplification had occurred: the newly formed region was composed of an array of euchromatic chromosome segments containing different amounts of heterochromatin. The size and C-band pattern of the amplified segments were heterogeneous. In several cells, the number 5~ of these amplified units exceeded 50; single-cell subclones of LP11 calf lines, however, carry stable marker chromosomes with 1 (5-15 segments and constant C-band patterns.
Subtines of the thymidine kinase-deficient LP11 cells (e~a., !.P11-15P 1 C517 cell line) established by single-cell cloning of LP11 cells were transfected with a thymidine kinase gene construct. Stable TK+
transfectants were established.
E~CAIVtPLE 10 Isolation of SATArrS ar9d other chror~e~somes with atypical base content and/or size 1: lsolatian of artificial chra~craosomes from endogenous chromos~mes Artificial chromosomes, such as SATACs, n-yy be sorted from endogenous chromosomes using any suitable procedures, and typically invol~re isolating metaphase chromosomes, distinguishing the artificial chromosomes from the endogenous chromosome::, and separating the artificial chromosomes from endogenous chromosomes. Such procedures will generally include the following basic steps: ( 1 ) culture of a sufficient number of ceNs (typically about 2 x 1 C7' mitotic cells) to yietd, preferably on the order of 1 x 106 artificial chromosomes, (2) arrest of the cell cycle of the cells in a stage of mitosis, preferrably metaphase, using a~ mitotic arrest agent.such as colchicine, (~) treatment of the tails, .
particularly by swelling of the cells in hypatonic buffer, to increase susceptibility of the cells to disruption, (4) by application of physical force to disrupt the cells in the presence of isolation buffers for stabilization of the released chromosomes, (51 dispersal of chromosomes in the presence of isolation buffers .for stabilization of free chromosomes, (6) separation of artificial from endogenous chromosomes and (7) storage (and shipping if desired) of the isolated artificial chromosomes in appropriate buffers. Modifications and variations of the general procedure for isolation of artificial chromosomes, for example to accommodate different cell types with differing growth characteristics and requirements and to optimize the duration of mitotic block with arresting agents to obtain the desired balance of chromosome yield and level of debris, may be empirically determined.
Steps 1-5 relate to isolation of metaphase chromosomes. The separation of artificial from endogenous chromoso~-nes (step 6) may be accomplished in a variety of ways. For example, the chromosomes~may be stained with DNA-specific dyes such as Hoeschst 33258 end chromomycin A3 and sorted into artificial and endogenous chromosomes on the basis of dye corstent by employing fluorescence-activated cell .
sorting (FRCS). To facilitate larger scale isotation of the artificial chromosomes, different separation techinietues may be employed such as swinging bucket centrifugation (to effect separation based on chromosome size and density) (see, e.g., Mendelsohn et ~9. (1968) J.
Mot. iol. 3_x:101-1081, zonal rotor centrifugation (to effect separati~n on the basis of chromosome size and density? tsee, e.g., Burki et a!. d19~3) PreQ. Biochem. 3:157-182; Stubfolefield et al. (19?8) Biochem. Bioe~hys.
Res. Commun. 83:1404-1414, velocity sedimentation (ta effect separation on the basis of chromosome size and Shape, (see e.g., Collard efi a1. (1984) Cvtometry 5_:9-19]. Immuno-affinity purification niay also be employed in larger scale artificial chromosome isolation procedures.
In this process, large populations of artificial chromosome-containing cells (asynchronous or mitoticalty enriched? are harvested en masse end the mitotic chromosomes dwhich can be released from the cells using - ~I m~ '~J-standard procedures such as by incubation of the cells in hypotonic buffer andlor detergent treatment of the cells in conjunction with physical disruption of the treated cells) are enriched by binding to antibodies that are bound to solid state matrices (e.g. column resins or magnetic beads9. Antibodies suitable for use in this procedure bind to condensed centromeric proteins or condensed and C~~A-bound histone proteins. For example, autoantibody LU851 (see Hadlaczky et al. ( 1989) Chromosoma 97:282-288), which recognizes mammalian centromeres may be used for Large-scale isolation ~f chromosomes prior to .
10~ . subsequent separation of .artificial from endogenous chromosomes using methods such as FACS- The bound chromosomes would be washed and eventually eluted ,for sorting. Immunoaffinity purification may also be used' directly to separate artificial chromosomes from endagenous chromosomes. For example, SATACs may be generated in or transferred to (e.g., by microinjection or microcell fusion as described herein) a cell line that has chromosomes that contain relatively srr~all amounts of heterochromatin, such as hamster cells (e:g., 'V79 cells o~ CHO-K1 cells,.
'The SATACs, which are predorr~inant9y heterochronnatin, are then separa-tad from the endogenous chromosomes by utilizing anti-heterochromatin 2~ binding protein (Drosophila HP-1) antibody conjugated to a s~lid matrix.
Such matrix preferentially binds SA'i'ACs relative to hamster chromosomes. Unbound hamster chromosomes are washed away from the matrix and the SATACs are eluted by standard techniques.
A. Celt lines arid calf culturing procedure:
. In one isolation pracedure, 1 BS mouse-hamster-human hybrid cells jsee, Figure 4j carrying the megachromosome or the truncated megac~romosome were grown in F-12 medium supplemented with 1 ~Dn/o fetal calf serum, 150 ug/ml hygromycin B and 400 ~rg/ml 6418. GFiB4.2 [a cell line recloned from G3D5 cells] manse-hamster hybrid cells carrying -a 3~-the megachromosome and the minichromosome were also cultured in F-12 medium containing 10% fetal calf serum, 150~g/ml hygromycin B
and 400 Ng/ml 6418. The doubling titres of both cell lines was about 24-40 hours, typically about 32 hours.
6 Typically, cell monolayers are passaged when they reach about 60-80~o confluence and are spot every 48-72 hours. Cells that reach.
greater than 80°10 confluence senesce in culture and are not preferred for chromosome harvesting. Cells may be plated in 100-200 104-mm dishes at about' 50-70% confluency 12-30 hours before mitotic arrest (see, i0 below)...
Other cell lines that may be used as hosts for artificial chromo-somes and from which the artificial chromosomes may be isolated in-elude, but are not limited to, PtK1 {1~8L-3) rvarsupial kidney cells (RTCC
accession no. CCL35), CHO-K1 Chinese hamster ovary cells {ATCC ac-15 cession no. CCL 61 ), i179-4 Chinese hamster fang cells {ATCC accession no. CCL93), Indian muntjac skin cells (ATCC accession no. CCL157), LMTK{-) thymidine kinase deficient marine L cells {ATCC accession no.
CCL1.3), Sf9 fall armyworm (Spodoprera frugiperr~a) ovary cells (ATCC
accession no. CRL 1711 ) 'and any generated heterokaryon (hybrid) eetl 20 fines, such as, for example, the hamster-rnurine hybrid cells described herein, that may be used to construct NIACs, particularly SATACs.
Cell lines may be selected, for example, to enhance efficiency of artificial chromosome production and isolation as may be desired in large-scale production processes. For instance, one consideration in selecting 25 host cells may be the artificial chromosome-to-total chromosome ratio of the celEs. To facilitate separation of artificial chromosomes from endogenous chrornt~osomesB a higher artificial chromosome-to-total chromosome ratio might be desirable. For example, for H1 D3 cells (a murine/hamster heterokaryon; see Figure 4), this ratio is 1;50, i.e-., one -~ 3T
artificial chromosome (the megachromosome) to 50 total chromosomes.
In contrast, Indian muntjac skin cells (ATCC accession no. CCL157~
contain a .smaller total number of chromosomes to diploid number of chromosomes of 7}, as do kangaroo rat cells, (a diploid number of chromosomes of 12) which would provide' for a higher artificial chromosome-to-total Chromosome ratio upon introduction af, or generation of, artificial chromosomes in the cells.
Another consideration in selecting host cells for production and isolation of artificial chromosomes rteay be size of the endogenous chromosomes as compared to that of the artificial chromosomes. Size differences of the chromosomes may be exploited t;a facilitate separation of artificial chromosomes from endogenous chromosomes. For example, because Indian muntjac skin cell chromosomes are consid.erabfy larger than minichromosomes and. truncated megachromosomes, separation of 95 the artificial chromosome from the muntjac chromosomes may possibly be accomplished using univariate tone dye, either Hoechst 33258 or Chromomycin A3p FRCS separation procedures.
Another consideration in selecting host cells for production and isoiatiai~ of artificial chromosomes may be the doubling time of the cells.
Far example, the amount of time required to generate a sufficient number of artificial chromosome-containing cells for use in procedures to isolate artificial chromosomes may be of significance for large-scale production.
Thus, host cells with shorter doubling times may be desirable. Far in-stance, the doubling time of V79 hamster lung cells is about 9-10 hours in comparison to the approximately 32-haour doubling time of. H1~3 cells.
Accordingly, several considerations may go into the selection of host cells for the production and isolation of artificial chromosomes. It may be that the host cell selected as the most desirable for de nova formation of artificial chromosomes is not optimized for large-scale production of the artificial chromosomes generated in the cell line. In such cases, it may be possibta, once the artificial chromosome has been generated in the initial host cell line, to transfer it to a productiori cell line more well suited to efficient; high-leval production and isolation of the 6 artificial chromosome. Such transfer may be accomplished through several methods, for example through rr~icrocell fusion, as described herein, or microinjeetion into the production cell fine of artificial chromosomes purified from the generating cell line using procedures such as described herein. Production cell lines preferably contain two or more copies of'the artificial artificial chromosome per cell.
B. Ch~omosorne isolation In general, cells are typically cultured for two generations at exponential growth prior to mitotic arrest. To accumulate mitotic .1 t33 and GHB42 cells in one particular isolation procedure, 5 ~glml coiehicine 1S was added for 12 hours to the cultures. The mitotic index obtained was 60-80%. The mitotic cells were harvested by selective detachment by gentle pipetting of the medium on the monolayer cells, tt is also possible to utilize mechanical shake-off as a means of releasing the rounded-up (mitotic) cells from the plate. The cells were sedimented by centrifugation at 200 x g for 1 ~ minutes.
Cells (grown on plastic or in suspension) may be arrested in different stages of the cell cycle with chemical agents other than colchicine, such as hydroxyurea, vinblastine, colcemid or aphidicolin.
Chemical agents that arrest the cells in stages other than mitosis, such as hydroxyurea and aphidicolin, are used to synchronize the cycles of all cells in the population 'and then are removed from the cell medium to allow the Celts to proceed, more or Less simultaneously, to mitosis at which time they may be harvested to disperse the chromosomes. IViitotic cells could be enriched for a mechanical shake-off (adherent cells). The -' 39-cell cycles of cells within a population of MAC-containing cells -may also be synchronized by nutrient, growth factor or hormane deprivation which leads.to an accumulation of cells in the Cz or GQ stage; readdition of nutrients or growth factors then allows the quiescent Celts to re-enter the the cell cycle in synchrony for abut one generation. Cell lines that are known to respond to hormone deprivation in this manner, and which are suitable as hosts for artificial chromosomes; include the iVb2 rat lymphoma cell line which is absolutely dependent on prolactin for stimulation of proliferation (see Gout et al. ( 19801 Cancer es. 40:2433-2436). Culturing the cells in prolactin-deficient mer~iurn for 18-24 hours leads to arrest of proliferation, with cells accumulating early in the C, phase of the cell cycle. Upan addition of prolactin, all the cells progress through the cell cycle until IVl phase at which point greater than 90% of the cells would be in mitosis (addition of colchicine could increase the amount of the rraitotic cells to greater than 95%). The time between reestablishing proliferation by protactiri additi~n and harvesting mitotic cells for chromosome separation may be empirically determined.
Alternatively, adherent cells, such as ~P79. cells, may be grown in roller bottles and mitotic cells released from the plastic surface by rotating the roller bottles at 200 rpm or greater ~Shwarchuk et al. (1993) Int. J. a 'at. Biol. -4:801-612). ~ At any given time, approximately 1 %
of the cells in an exponentially growing asynchronous population. is in nlf-phase. Even without the addition of colchicine, 2 x 10' ~~nitotic cells have been harvested from four 1750-cm2 rotler bottles after a 5-min spin at 200 rpm. Addition of colchicine for 2 hours may increase the yield to 6 x 108 mitotic cells.
Several procedures may be used to isolate metaphase chromosomes from these cells, including, but not limited to, one based on a polyamine buffer system (Cram et at. 11990) Methods in Cell Biolocw 33:377-3823, one on a modified hexylene glycol buffer system (Hadlaczky et al. (1982) Chromosome X6:643-65]" one on a magnesium sulfate buffer system (Van den Engh et al. (19885 ~ytometry x:266-270 and Van den Engh et al. (198.) Cvtometrv 5:108], one on an acetic acid fixation buffer system (Stoehr ~t al. (19$2) Histochemistry 74:57-5'6], and one on a technique utilizing hy~aotonic KCI and prszpidium iodide [Cram et al. ('1994) XVII meeting of the International Society for Analytical Cytology, October 16-21 P Tutorial IV Chromosome Analysis and Sortinca with Commerical Flow Cvtometers; Cram et al. ~1990y Methods in Celi Biology X3:376].
1. Polyamine procedure In the pofyamine procedure that was used in isolating artificial chromosomes from either 1 B3 or GHB~.2 cells, about 10' mitotic cells were incubated in 10 mi hypotonic buffer i75 mM KCi, 0.2 mM
spermine, 0.5 mM spermidine) for 10 minutes at room temperature to swell the cells. The cells are swollen in hypotonic buffer to loosen the metaphase chromosomes but not to the point of e:ell lysis. The cells were then centrifuged at 100 x g for 8 minutes, typically at room temperature. The cell pellet was drained carefully and about 10' cells were resuspended in 1 ml polyamine buffer (16 mM Tris-HCI, 20 mM
NaCI, 80 mM KCI, 2 mM EDTA, 0,5 r~tM EGTA, '14 mM a-mercapto-ethanol, 0.1 ~/~ digitonin, 0.2 mM Spermine, 0.5 mM spermidine] for physical dispersal of the metaphase chromosomes. Chromosomes were then released by gently drawing the cell suspension up and expellir'g it through a 22 G needle attached to a 3 ml plastic syringe. The chromosome concentration was about 1-3 x 10g chromosomeslml.
The polyamine buffer isolation protocol is waif suited for obtaining high molecular weight chromosoi~nal DNA [Sillar and Young 41981 ) J.
Histochem. Cytochem. 29:74-78; Van~illa et al. ~ 1988) Biotechnolo4v °141-4:537-552; Bartholdi ~t al. ( 19881 In °'Molecular Genetics of Mammalian Cells" (M.G~ettsman, ed.), Methods in Enzymolocry 151 :252-267.
Academic Press, Orlando. The chromosome stabilizing buffer uses the polyamines spermine and spsrmidine to stabilize chromosome.structure (Blumenthai et al. (1979)x. Cell Biol. 81:255-259; I-elands et af. (1985 Cancer Genet_ Cvto,genet. 23:151-157] and heavy .metals chelators to reduce nuclease activity.
The polyamirie buffer protocol has wide applicability, however, as with other protocols, the following variables must be optimized for each cell type: blocking time, cell concentration, type of hypatonic swelling buffer, swelling time, volume of hypato~ic buffer, and vortexing time.
Chromosomes prepared using this protocol are typically highly .
condensed_ There are several hypotonic buffers that may be used to swell the cells, for example buffers such as the following: 75 mM KCl; 75 m\ol KCi, 0.2 mM spermine, 0.5 mM spermidine; Ohnuki's buffer of 16.2 mM
sodium nitrate, 6.5 mM sodium acetate, 32.4 mM KCI (Ohnuki ( 1965) Nature 2~8:916-917 and Ohnuki (1968) Chromosome 25:402-428i; and a variation of Ohnuki's buffer that additionally contains 0.2 mM spermine and 0.5 mM spermidine. The amount and hypotor~icity of added buffer vary depending on Celt type and cell concentration. Amounts may range from 2.5 - 5.5 ml per 1 O7 cells or .more. Swelling times may vary from 10-90 minutes depending on ce1! type and which swelling buffer is used.
The composition of the ~olyamine isolation buffer may also be varied. For exempts; one modified buffer contains 15 mM Tris-HCi, pH
7.2, 70 mM NaCI, 80 mM KC1~, 2 mM ~DTA, 0.5 mM EGTA, 14 mM
beta-mer.captoethanol, 0.25°lo Triton-7CT"", .02 ~nM ~permine and .m5 rnM
spermidirie.

-1~2-Chromosomal dispersal may also be accomplished by a variety of physical means. For example, cell suspension may be gently drawn up and expelled m a 3-ml syringe fitted with a 22-gauge needle (Cram et al.
( 1990? Methods in Cell Bioloc~y 33:377-3823, cell suspension may be agitated on a bench-top vortex (Cram et g~1. ( 1990) lilllethods in Cell Bio_~_ foav 33:377-382], cell suspension may be disrupted v~ith a homogenizes (Sillar and 'bung (1981) .~. Histochem. Cvtochern. 29:74-' 78; Carrano et al. (1979' Proc. iVatl. Acad. Sci. U.S.A. 75:1382-1384]
and cell suspension may be disrupted with a bench-top ultrasonic bath (Stoehr et al. (1982) Histochemistrv 74:57-61].
2. Hexylene glycol buffer system In the hexylene glycol buffer procedure that was used in isolating artificial chromosomes from either 1 B3 or GHB42 cells, about 8 x 1 ~~
mitotic cells were resuspended in 10 m3 glycine-h~xylene glycol buffer 1100 mM glycine, 1 °~ hexylene glycfll, pH 8.4-8.8 adjusted with saturated Ca-hydroxide solution] and incubated for 10 minutes at 37°C, followed by centrifugation for 10 minutes to pellet the nuclei. The supernatant was centrifuged again at 200 x g for 20 siainutes to pellet the .chromosomes. Chromosomes were resuspended in isolation buffer t~-3x108 chromosomes/ml).
The hexylene glycol buffer composition may also be modified. For example, one modified buffer contains 25 mM Tris-HCI, pH 7.2, 750 mM
hexylene glycol, 0.5 mM CaCl2, 1.0 mM MgCl2 [Carrano et al. (1979) Proc. Natl. Acad. Sci. U.S.A. 76:'( 382-1384].
3. IUlagnesium-sulfate buffer system This buffer system may be used with any of the methods of cell swelling and chromosomal dispersal, such as described above in connection with the polyamine and hexylene glycol buffer systems. In this procedure, mitotic cells are resuspended in the following buffer: 4.8 mM HEPES, pH $.0, 9.8 mM MgSO~, 48 mM K.CI, 2.9 mM dithiothreitol Van den Engh et ai. 419851 Cytometry x:92 and Van den Engh et al.
1984? Cvt m~etrv. 5_:108].
4. Acetic acid fixation buffer system This buffer system may be used, with any of the methods of cell swelling and chromosomal dispersal, such as described above in connection with the polyamine and hexylene glycol buffer systems. tn this procedure, mitotic cells are resuspended in the following buffer: 25 mM Tris-HCI, pH 3.2, 750 mM ( 1,6)-hexandioi, 0.5 mM CaCiZ, 1.0°rr acetic acid EStoehr et ai. (1982) Histochemistrv 74:57-Blj.
5. KCt-propidium iodide buffer system This buffer system may be used with any of the methods of cell swelling and chromosomal dispersal, such as described above in connection with the poiyamine and hexylene glycol buffer systems. In "t 5 this procedure, mitotic calls are resuspended in the following buffer: 25 mM KCI, 50 ,uglml propidlum Iodide, 0.33% Triton X-100, 333 ,uglml RNase ~Cra;n et al. 11990) Methods in Cell Biolocw 33:37fi].
The fluorescent dye propidiurx~ iodide is used and also serves as a chromosome stabilizing agent. Swelling of the cells in the hypotonic medium Iwhich may also contain propidium iodide) may be monitored by placing a smaN drop of the suspension on a microscope slide end observing the cells by phaseJfSuorescent niicroscopy. T>~e cells should exclude the propidium iodide while swelling' but some may lyse prematurely and show chromosome fluorescence. After the cells have been centrifuged and resuspended in the KCl-propidium iodide buffer system, they wilt be lysed due to the presence of the detergent in the buffer. The chcomasomes may then be dispersed and then incubated at 37°C for up to 3~ minutes to permit the RNase to act. The chromosome preparation is then analyzed by flow cytometry. The propidium iodide -144~
fluorescence can be excited at the 488 nm wavelength of an argon laser and detected through an ~G 570 optical filter by a single photomultiplier tube. The single pulse may be integrated and acquired in an univariate histogram. The flow cytometer may be aligned to a CV of 2°~ or less using small ( 9 , 5 ,um diameter) microspheres. The chromosome .
preparation is filtered thraugh 80 pm nylon mesh befare analysis. .
C. Staining of chromosomes with DNA-specific dyes Subsequent to l olation, the chromosome preparation was shined with Hoechst 33258 at 6 pglml and chromomycin A3 at 2Q0 ~g/ml.
Fifteen minutes prior to analysis, 25 mM Na-sulphite and 10 mM Na-citrate were added to the chromosome suspension.
~. Plow sorting of chrornosomes Chromosomes obtained from 1 B3 and GHB42 cells and. maintained were suspended in a polyamine-based sheath buffer (0.5 mM EGTA, 2.0 mM EDTA, 80 mM KCI, 70 mM NaCI, ~ 5 mllii Tris-HCI, pH 7.2, 0.2 mM
spermine and 0.5 mM spermidine) LSsllar and Your~g ~'I 981 ) J.
Histochem. Cv t~ochern. 29:74-78~. The chromosomes were then passed through a dual-laser cell sorter LFACStar Plus or FAXStar Vantage Becton Dickinson Immunocytornetry System; other dual-laser starters may also be used, such as those manufactured by Coulter Electronics (Elite ESP) and Cytomation iMoFlo)3 in which two lasers were set to excite the dyes separately, allowing a bivariate analysis of the chromosome by size and base-pair composition. Because of the difference between the base compositian of the SATACs arid the other chromosomes and the resulting difference in interaction with the dyes, as well as size differences, the SATACs were separated from the other chromosomes.

CA 02429726 20'03-06-09 -1 ~~-E. Storage of the sorted artificial chromosomes Sorted chromosomes may be pelleted by centrifugation and resuspended iri a variety of buffers, and stored at 4°C, f=or exa.mpl.e, the isolated artificial chromosomes rraay be stored in GH buffer (100 mM
glycine, 19'o hexylerte glycol pH 8.4-8.f adjusted with saturated Ca-hydroxide solution) (see, e-o., Hadlaczky et al. (1982) Chromosome 86:643-659) for one day and embedded by centrifuigation into agarose.
The sorted ch.rornosomes were centrifuged into an agarose bed and the plugs are stared in 500 mM EDTA at 4° C, Additional stmrage buffers include CMf3-I/polyamine buffer (17.5 mM Tris-HCI, pH 7.4, 1.1 mM
EDTA, 50 mM epsilon-amino caproic acid, 5 mM benzamide-HCI, 0.40 mM sperrnine, 1.0 mM spermidine, 0.25 mM EGTA, 40 mM KCI; 35 mM
NaCI) and CMB-lllpalyamine buffer (100 mM gfycine, pH 7.5, 78 mM
hexylene glycol, 0.1 mM EDTA, 50 mM epsilon-amino capraic acid, 6 mM benzarnide-HCf, 0.40 mM ~perrnine, 1.0 mM spermidine, 0.25 mM
EGTA, 40 mM KCI, 35 mM NaCI).
When microinjection is the intended use, the sorted chromosomes are stored in 3096 glycerol at -20° C. Sorted chromosomes may also be stored without glycerol for short periods of time (3-6 days) in storage .
buffers at 4.°C. Exemplary buffers for rr'icroinjectior~ include CBM-I
i10 mM Tris-HCI, pH 7.5, 0.1 mM EDTA, 50 mM epsilon-amino caproic acid, 5 mM benzamide-HCI, 0.30 mM spermine, 0.75 mM spermidine), CBM-II
(1-0O mM glycine, pH 7.5, 78 mM hexylene glycol, 0.1 mM EDTA, 50 mM epsilon-amino caproic acid, 5 mM benzamide-HCI, 0.30 mM
spermine, 0.75 mM spermidine?.
For fang-term storage of sorted chromosomes, the above buffers are preferably supplemented with 50% glycerol and stored at -20°C.

- ~ 4~-F. Quality Control 1. Analysis of the purity The purity of the sorted chromosomes was checked by fluorescence in s~ru hybridization (FISH) with a bi~atin-labeled mouse satellite DNA probe (see, Hadlaczky et al. ( 189.1 ) Proc. Nat!. Acad. Sci.
U.S.A. 88:8105-8110. Purity o.f-the isolated chromosomes was about 97-99°/~.
2. Characteristics of the sorted chromosomes Pulsed field gel electrophoresis and Southern hybridization were 1.~ carried out to determine the size distribution of the DNA content of the sorted artificial chromosomes.
G. Functioning of the purified artificial chromosomes To check whether their activity is preserved, the purified artificiaP
chromosomes may be microinjected (using methods such as those described in Example 13) into primary cells, sor~atio cells and stem ceiis which are then analyzed for expression of the heterologous genes carried by,the artificial chromosomes, e.g., such as anaiysis for growth on selective medium and assays of j3-galactcssidase activity.
tf. Sorting of mammalian artificial chrornoson~e-containing microcelis ~l. Micronucleation Cells were grown to 80-90% confluency in ~ T150 flasks.
Coicemid was added to a final concentration of 0.06,~g/ml, and then incubated with the cells at 37°G for 24 hours.
B. Enucleation Ten ~glm! cytochalasin B was added and the resulting microcelis were centrifuged at 15,000 rpm for '~0 minutes at ~8-33° C.
C. Purification of microcells by filtration The microcells were purified using ~~inn~:xT"" filter units and NucleoporeT"" biters (5,um and 3,um], -14~-~. Staining and sorting rnicropelts As above, the cells were stained with Hoechst. and chromomycin A~ dyes. The rr~icroceils were sorted by cell sorter to isolate the microcetls that contain the mammalian artificial chromosomes.
E. Fusion The microcells that contain the artificial chromosome are fused, for example, as described in Example I.A.S., to selected primary cells, somatic cells, embryonic stem cells to generate transgenic (non-human?
animals and for gene therapy purposes, and to other cells to deliver the 90 chromosomes to the cells.
E)CANtPLE 11 Introduction of gnammalian artificial chromosomes Into insect Celts insect cells are useful hosts for MACS, particeaiarly for use in the production of gene products, for a number of reasons, inciuding~
7. A mammalian artificial chromosome provides an extra-genomic specific integration site for introduction of genes encoding proteins of interest (reduced chance of rr~utation in production system.
2. The large size of an artificial chromosot~e permits megabase size DNA integration so that genes encoding an entire pathway leading to a protein or nonprotein of therapeutic value, such as an aikalaid (digitalis, morphine, taxoll can be accomodated by the artificial chromosome.
3. Amplification of genes encoding useful proteins can be accomplished in the artificial ra~ammaiian chromosome to obtain higher protein yields in insect Celts.
4.. Insect delis support required post-translational modifications (glycosylation, phosphorylationl essential for protein biological function.
5. Insect cells do not support mammalian viruses - eliminates cross-contamination of product with human infectious agents.

148' 6. The ability to introduce chromosomes circumvents traditional recombinant baculovirus systems for production of nutritional, industrial or medicinal proteins in insect cell. systems.
7. The low temperature optimum for insect cell growth 128° C~
permiits reduced energy cost of production.

8. Serum free growth medium for insect cells well result in lower production costs.

9. Artificial chromosome-containing cells can be stored indefinitely at low temperature.
1~ 10. Insect larvae will serve as biological factories for the production of nutritional, medicinal or industrial proteins by microinjection of fertilized insect eggs.
~A. ~emonstration that insect cells recognize marr'malian promoters Gene constructs containing a mammalian promoter, such as the CMV promoter, linked to a detectable marker gene [Renilla luciferase gene tsee, e.~., U.S. Patent No: 5,292,658 for a description of DNI~
encoding the Renilla luciferase, and plasmid pTZri_uc'1, which can provide the starting material for construction of such vectors, see also SEQ. ID No. 101 and also including the simian virus 40 (SV401 promoter operably linked to the j-galactosidase gene were introduced into the cells of two species Trichoplcesia ni [cabbage looper] and b'ornbyx mor~ [silk warms.
After transferring the constructs into the insect cell lines either by eiectroporation or by micr~injection, expression of the marker genes was detected in luciferase assays (see e'a., Example ~ 2.C.S? and in ~-galactosidase assays such as iacZ staining assaysf after a 24-h incubation. In each case a positive result was obtained in the samples containing the genes which was absent in samples in which the genes were omitted. In addition; a R. mori ~-actin promoter-Renllla luciferase gene fusion was introduced into the T r~i and 8, mo~i cells which yielded light emission after transfection. Thus, certain mammalian promoters function to direct expression of these marker genes in insect cells.
Therefore, MACs are candidates for expression of heterolagaus genes in insect cells.
B. Construction of vectors for use in insect cells and fusion with maiwmalian cells 1. Transform LiVITK~.ceils with expression vector with:
a. B. mori a-actin promoter- Hyg' selectable marker gene for insect cells, and b. SV4.0 ar CMV promoters controlling a puromycin' selectable marker gene far mammalian cells.
2. Detect expression of the mammalian promoter in LfVITK cells [puromycin' LMTK cells?
3. Use puromycin~ cells in fusion experiments with Bz~mbyx and Tr~choplusia cells, select Hyg~ cells.
C. Insertioni of the MACS into insect cells These experiments are designed to detect expression of a detectable marker gene [such as the ,l3-galactasidase gene. expressed under the control of a mammalian promoter, such as pSV40 ] located on a MAC that has been introduced into an insect cell. Data indicate that ~3-gal was expressed.
Insect cells are fused with mammalian cells containing mammalian a~rtificia! chromosomes, e~a., the minichromosome [EC31~C5] or the mini and the megachramosome [such as GHB42, which is a cell line reeioned from G3D51 or a cel( line that carries only the megachromosome [such as H'1 D3 or a reclone therefrom. Fusion is carried out as follows:

1. mammalian + insect. cells 150/50%) in log phase growth are mixed;
2. calcium/PEG cell fusion: (10 min - 0.5 h);
3. heterokaryons ,( + 7 2 h) are selected.
The following selection conditions to select for insect cells that contain a MAC can be used: [ + = positive selection; - = negative selection]:
'! . growth at 28° C ( + insect cells, - mammalian cells);
2. Grads insect cell medium (SIGMA] -(- mammalian cells);
3. no exogenous COz [- mammalian cells]; andlor 4, antibiotic selection (Hyg or 6418? ( + transformed insect cells).
Immediately following the fusion protocol, many heterokaryons [.fusion events.] are observed between the mammalian and each species of insect cells [up to 90% heterokaryons]. After growth [2+ weeks] on insect medium containing 64.18 andlor hygromycin at selection levels used for selection of transformed mammalian cells, individual colonies are detected growing on the fusion plates. By virtue of selection for the antibiotic resistance conferred by the I~AC and selection for insect cells, these colonies should contain MACS.
The ~, mori,B-actin gene promoter has beers shown to direct expression of the /3-galactosidase gene in 8, mori cells and mammalian cells tea., EC3I?C5 cells). The B: mori (3 actin gene promoter is, thus, particularly useful for inclusion in MACs generated in mammalian cells that will subsequently be transferred into insect cells because the presence of any marker gene linked to the promoter cars be determined in the mammalian and resulting insect cell lines.

Preparation of chrornosorne fragrrsentatiore vectors and other vectors for targeted integration of ~lVA into MACS
Fragmentation of the megachromasome should ultimately result in smaller stable chromosomes that contain about 15 ti~b to 50 lVlb that will be easily manipulated for use as vectors. Vectors to effect such fragmentation should also, aid in determination arid identification of the elements required for preparation of an ire vitro-produced artificial chromosome.
Reduction in the size of the megachromosome can be achieved in a number of different ways including: stress treatment, such as by starvation, or cold or heat treatment; treatment with agents that .
destabilize the genome or nick DtVA~ such as BrdU, coumarin, EMS and others; treatment with ionizing radiation [see, e~a., E3rown E 1992) Curr.
Ooin. Genes Dev. ?:479-486]r and teiomere-directed ire vireo chromosome fragmentation [see, e~a., Farr ~t af, (1995) EMBO J. °i4m5444-54.54).
A. Qreparation of vectors for fragmentation of the artificial chromosome end also fo~° targeted integration of selected gene products 1. Construction of pTEMPLfE~
Plasmid pTEMPUD [see Figure 5l is a mouse homologous recombination "kilter" vector for ire vivo chromosorr°n fragmentation, and also for inducing large-scale amplification via site-specific integration.
With reference to Figure 5, the -~ 3,625-by Satl-Pstl fragment was derived from the pBabe-puro retroviral vector [see, lUforgenstern et at.
( 1990) Nucleic Acids Res. 18:3587-3596]. 'This fragment contains DNA
encoding ampici8tin resistance, the pUC origin of replication, and the puromycin N-acetyl transferase gene under control of the SV40 early promoter. The URA3 gene portion comes from the p'fACS cloning vector CS1GMA]. URA3 was cut out of pYAC5 with 5atl-~hol digestion, cloned into phlEB193 [New England Biolabs], which was then cut with. EcoRi-Sall and ligafed to the Sill site of pBabepuro to produce pPU.
A 1293-by fragment [see SEQ iD iVo. 1 ] encoding the mouse major satellite, was isolated as an EcoRl fragment from a DNA library produced from mouse LMTK- fibroblast cells and inserted into the EcoRl site of pPU
to produce pMPU.
The TK promoter-driven diphtheria toxin gene [DT-A] was derived from pMC1DT-A [see, Maxwell et ate, (1986y Cancer Res. 46:4660-4.666]
by Bglll-Xhol digestion and cloned into the pMC1 neo poly A expression vector [STRATAGENE, La ;lolls, CA] by replacing the neomycin-resistance gene coding sequence, The TK promoter, DT-A gene and poly A sequence were removed from this vector; cohesive ends were filled with Kienow and the resulting fragment blunt end-iigated and ligated into the SnaBl [TACGTA] of pMPU to produce pMPUD.
The I-lutel 2.5-kb fragment [see SEQ ID No.3] was inserted at the.
Pstl site Isee the 6100 Pstl - 3625 Pstl fragment on pTEMPUD] of pMPUD to produce pTEMPUD. This fragment includes a human telomere. It includes a unique ~gl_II sit~ [see nucleotides 1042-104- of SEQ 1D No.3], which will be used as a site for introduction of a synthetic telomere that includes multiple repeats [80] of TTAGGG with Baml-t! and ~gll ends for insertion into the Bglll site which wilt then remain unique, since the BamHl overhang is compatible with the Bglll site. Ligation of a BamHl fragment to a B~II l destroys the B~c III site, so that only a single _Bglll site will remain. Selection for the unique Bdlll site insures that the synthetic telomere will be inserted in the correct orientation. The unique Bglil site is the site at which the vector is iineari~ed.

v ,~13-To generate a synthetic telomere made up of multiple repeats of the sequence TTAGGG, attempts were made to clone or amplify iigation products of 30-mer oligonucleotides containing repeats of the sequence.
Two 30-mer oiigonucleotides, one containing four repeats of TTAGGG
bounded on each end of the complete run of repeats. by half of a repeat and the other containing five repeats of the complement AATCCC, were annealed. The resulting double-standed molecule with 3-by protruding ends, each representing half of a repeat, was expected to ligate with itself to yield concatamers of n x 30 bp. However, this approach was unsuccessful, likely due to formation of quadruplex DNA from the G-rich strand. Similar difficulty has been encountered in attempts to generate long repeats of the pentameric human satellite fl and Ili units. Thus, it appears that, in genera~y any oligomer sequence containing periodically spaced consecutive series of guanine nucleotides is likely to form .
undesired quadruplex formation that hinders construction of tong doubie-stranded DNAs containing the sequence.
Therefore, in another attempt to construct a :synthetic telomere for insertion into the Bglll site of pTEMPUD, the starting material was based on the complementary C-rich repeat sequence {i.e., AATCCC) which ~0 would not be susceptible to quadruplex structure farmation. Two plasmids, designated pTEL280110 and pTe1280111, were constructed as follows to serve as the start)ng materials.
First, a long oligonucleotide containing 9 repeats of the sequence AATCCC ~i.e., the complement of telomere sequence TTA,GGG~ in reverse order bounded on each end of the complete run of repeats by half of a repeat (therefore, in essence, containing 10 repeats), and recognition sites for Pstl and Pacl restriction enzymes was synthesized using standard methods. The oligonucleotide sequende is as follows:
5'-AAACTGCAGGTTAATTAACCCTAACCCTAACCCTAACCCTAACCCTAAC

CCTAACCCTAACCCTAACCCTAACCCGGGAT-3' (SE(1 (D N0. 29) A partially complementary short oligoriucteotide of sequence 3'-TTGGGCCCTAGGCTTAAGG-5' (SEQ 1D NO. 30) was also synthesized. The oligonucleotides were gel-purified, annealed, repaired with Klenow potymerase and digested with EcoRl and i'sti. The resulting EcoRIlPstl fragment v~ias ligated with Ecof~I/Pstl-digested pUCl9. The resulting plasmid was used to firansform ~ coil DH5a competent cells and plasmid DNA, (pTe1~02) from one of the t~ansformants surviving selection on LBfampicillin was digested with 1d Pacl, rendered blunt-ended by Klenow and dNTPs and digested with Hindttl. The resulting 2.7-kb fragment was ge(-purified.
Simultaneously, the same ptasmid was amplified by the . polymerase chain reaction using extended and mode distal 26-mer M't 3 sequencing primers. The amplification product was digested with mat and Hindi!!, the double-stranded B~.-bp fragment containing the fi0-bpw telomeric repeat (plus 24 by of linker sequence) was isolated on a f °fo native polyacrylamide gel, and ligated with the double-digested pTel'! 02 to yield a 7 20-by telomeric sequence. This plasmid was used to transform DHScr cells. t'lasmid DNA from two of the~resulting recombinants that survived selection on ampiciilin ( 100 Wglmt) was sequenced on an ABl DNA sequences using the dye-termination method.
One of the piasmids, designated pTel29, contained a sequence of 20 repeats of the sequence TTAGGG ti.e., 19 successive repeats of TTAGGG bounded on each end of the complete gun of repeats with half of a repeat). The other plasmid, designated pTele'.S, had undergone a deletion of 2 tap (TA) at the junction where the two sequences, each containing, in essence, 'l 0 repeats of the TTAGGG sequence, that had been ligated to yield the plasmid. This resulted in a GGGTGGG motif at the junction,in pTel2>3. This mutation provides a useful tag in telomere-'~~5-directed chromosome fragmentation experiments. Therefore, the pTel~9 insert was amplified by PCR using plaC/IVI13 sequencing primers based on sequence samewhat longer and farther from the potylinker than usual as follows:
5'-GCCAGGGTTTTCCCAGTCACGACGT-3' (~E~2 ID NO. 31) or in some experiments 5'-GCTGCAAGGCGATTAAGTTGGGTAAC-3° (8EC1 ID NO. 32)' as the m13 forward primer, and 5'-TATGTTGTGTGGAATTGTGAGCGGAT-3' (:S~C~ ID NO. 33) as the m13 reverse primer.
The amplification product was digested with Smal and Hindtll. The resulting 144-by fragment v~ias gel-purified on a 6% native potyacrylamide gel and ligated with pTet2.8 that had been digested with Pacl, blunt-ended with Ktenow and dNTP and then digested with Hindll!
to remove linker, The !lgation yielded a pfasmid designated pTel28O1 -containing a telomer.ic sequence of 4t7 repeats of the sequence TTAGGC~
in which one of the repeats (i.e., the 3Qth repeat) tacked two nucleotides (TA), due to the deletion that had occurred in pTet28, to yield a repeat as follows: TGGG.
In the next extension step, pTet2801 was digested with Sma1 and Eimdflt and the 264-by insert fragment was get-purified and tigated with pTe12801 which had been digested with Pacl, bfunt~ended and digested with Hindlll. The resulting plasmid was transformed into DHSa cells and plasmid DNA from 12 of the resulting transformants that Survived selection on ampicillin was examined by restriction enzyme ar~atysis for the presence of a ~.5-kb EcoRllPstl insert fragment. Eleven of the recombinants contained the expected ~.5-kb insert. The inserts of two of the recombinants were sequenced and found to be as expected.
These plasmids were designated pTe128O110 and pTel28~1 l 1 . These -~ FJs-plasmids, which are identical, both contain 80 repeats of the sequence TTAGGG, in which two of the repeats (i.e., the 30th and 70th repeats) lacked two nucleotides (TA7, due to the deletion that had occurred in pTel28, to yield a repeat as follows: TGGG. Thus, in each of the cloning steps (except the first), the length. of the synthetic telomere doubled; that is, it was increasing in size exponentially. !ts length was fi0x2" bp, wherein n is the number of extension cloning steps undertaken.
Therefore, in principle (assuming E. coil, or any other microbial host, e.g., yeast, tolerates long tandem repetitive DNA); it is possible to assemble '10 any desirable size of safe teiomeric repeats.
tn a further extension step, pTe12801 10 was digested with Paci, blunt-ended with Klenow polymerase in the presence of dNTP, 'then digested with Hindlll. Tie resulting 0.5-kb fragment was gel. purified.
Plasmid pTel28~111 was cleaved with Smal and Hindlll and the 3.2-k;b 'i 5 fragment was gel-purified and ligated to the 0.5-kb fragment from pTei2801 10. The resulting plasmid was used to transform DH5n cells.
Plasmid DNA was purified f rom transformants surviving ampiciliin selection. Nine of the selected recombinants were examined by restriction enzyme analysis for the presence of a 1.0-kb 6coR1lPstl 20 fragment. Four of the recombinants (designated p'Tlk2, pTlk6, pTlk7 and pTlk81 were thus found to contain the desired 960 by telomere DNA
insert sequence that included '! 60 repeats of the sequence TTAGGG in which four of the repeats lacked two nucleotides (TA), due to the deletion that had occurred in pTel28, to yield a repeat as follows: TGGG.
25 Partial DNA sequence analysis df the EcoRI/Pstl fragment of two of these plasmids (i.e., pTlk2 and pTlk6), in which approximately 300 by from both ends of the fragrwent were elucidated, confirmed that the sequence was composed of successive repeats of the TTAGGG sequence.

° 157-In order to add Pmel and Bgill sites to the synthetic telomere sequence, pTik2 was digested with Paci and Pstl and the 3.7-kb fragment li.e., 2.7-kb pUCl9 and 1.0-kb repeat sequence) arias gei-purified and ligated at the Psti cohesive end with the foliowing oligonucleotide 5'-GGGTTTAAACAGATCTCTGCA-3' ~SE4 iD N0. 34).
The ligation product was subsequently repaired with iCienow polymerase and dNTP, ligated to itself and transformed into E. caii strain DHScr. A
total of 14 recombinants surviving selection on ampicillin were obtained.
Piasmid DNA from each recombinant was able to be cleaved with Bgili indicating that this added unique restriction site had been retained by each recombinant. Pour of the 14 recombinants contained the complete 1-kb synthetic teiomere insert, whereas the insert of the remaining 1~
recombinants had -undergone deletions of various lengths. The four plasmids in which the ~ -kb synthetic tetomere sequence remained intact 1 a were designated pTik~l2, pTIkV5, pTIkVB an pTIkV ~ .2. Each of these plasmids couid also be digested with Pmel; in addition the presence of both the X11 nad Pmel sites was verified by sequence analysis. Any of these four plasmids can be digested with BamHl and Bgtli to reiease a fragment containing the 1-kb synthetic tefomere sequence which is then 2~ ligated with Bgltl-digested pTEMPUD.
2. Use of pTEPU~ $or ira vivo chromosome fragmentation Linearization of pTEMPUD by B_glfi results in a Linear molecule with a human telomere at one end. Integration of this linear fragment into the chromosorrie, such as the megachromosome in hybrid cells or any mouse 25 chromosome which contains repeats of the mouse major satellite sequence results in integration of the selectable marker puromycin-resistance gene and cleavage of the plasmid by virtue of the telomeric end. The DT gene prevents that entire linear fragment from integrating by random events,, since upon integration and expression it is toxic.

Thus random integration will be toxic, so site-directed integration into the targeted DNA will be selected. Such integration will produce fragmented chromosomes.
The fragmented truncated chromosome with the new telomere will survive, and the other fragment without the centromere will be lost.
Repeated in vivo fragmentations will ultimately result in selection of the smallest functioning artificial chromosome possible. Thus, this vector can be used to produce minichromosomes from mouse chromosomes, or to fragment the rnegachromosome. !n principle, this vectbr. can be used to target any selected t3NA sequence in any ~chromosort~e to achieve fragmentation.
3. Construction. of pVERPUD
A fragmentation/targeting vector analogous to pTEMPUD for ira vivo chromosome fragmentation, and also for inducing barge-scale amplification via site-specific integration but vihich is based on mouse rDNA sequence instead of mouse major satellite DNA has been designated pTEFiPUD. fn this vector, the mouse rt~ajor satellite DNA
sequence of pTEMPUD has been replaced with a 4770-by SamHl fragment of megachromosome clone 161 which contains sequence corresponding to nucleotides 10,232-15,000 in SEQ iD NO. 1 C.
4.. pHASPUD and pl'EMPhu3 Vectors that-specifically target human chromosomes can be constructed from pTEMPUD. These vectors can be used to fragment specific human chromosomes, depending upon the.selected satellite sequence, to produce human minichromosomes, and also to isolate human centrorneres.

° 169-a. pHASPUD
To render pTEMPUD suitable for fragmenting human chromosomes, the mouse major satellite sequence is replaced With human satellite sequences. Unlike mouse chromosomes, each human chromosome has a unique satellite sequence. For example, the mouse major satellite has been replaced with a human hexameric a satellite (or alphoid satellite] DNA sequence. This sequence is an 513-by fragment [nucleotide 232-1044 of SEQ (D No. 2] from clone X512, deposited in the EMBL database under Accession number X60716, isolated from a human colon carcinoma cell line Colo32G (deposited under Accession No.
ATCC CCL 220.11. The 813-by alphoid fragment can. be obtained from the pS12 clone by nucleic acid amplification using synthetic primers, each of which contains an EcoRi site, as follows:
GGGGAATTCAT TGGGATGTTT CAGTTGA forward primer tSEt~ ID No. 4]
'!5 CGAAAGTCCCC CCTAGGAGAT CTTAAGGA reverse primer (SEQ ID No. 5~.
Digestion of the amplified product with EcoRl results in a fragment with EcoRl ends that includes the human a-satellite sequence. This sequence is inserted into pTEMPUD in place of the EcoRl fragrrtent that contains the mouse major satellite to yield pHASPUD, 2a Vector pHASPUD was linearized with._Bgl_Il and used to transform EJ30 (human fibroblast) cells by scrape loading. Twenty-seven puramycin-resistant transformant strains were obtained.
b. pTEMPhc~3 In pTEMPhu3, the mouse major satellite sequence is replaced by 25 the 3kb human chromosome 3-specific ~r-satellite from ~3Z1 (deposited under ATCC Accession No. 85434; seeB also Yrokov (1959) Cvto ec~ net.
Cell Genet. 51:1114j.

-16~-5. Use of the pTEMPt-itJ3 to induce amplification on human chromosome #3 Each human chromosome contains unique chromosome-specific a(phoid sequence. Thus, pTEMPH~l3, which is targeted to the chromosome 3-specific a-satellite, can be introduced into human cells under selective conditions, whereby large-scale amplification of the chromosome 3 centromeric region and production of a de novo chromosome ensues. Such induced large-scale amplification provides a means for inducing de navo chromosome formation and also for in vivo cloning of defined human chromosome fragments erp to megabase size.
For example, the break-point in human chromosome 3 is on the short arm near the centromere. This region is involved in renal cell carc(n~ma formation. By targeting pTEMPhu3 to this region, the induced 9arge-scale amplification may contain this region, which can then be 75 cloned using the bacteria( and ;.east markers in the pTEMPhu3 vector.
The pTEMPhu3 c(~ning vector allows not only selection for homologous recombinants, but also direct cloning of the integration site in YACS. This orector can also be used to target human vhromosome 3, preferably with.a deleted short arm, in a mouse-human mono-2~ chromosomal microcell hybrid line. Homologous recombinants can be screened by nucleic acid amplification ~PCR), and amp(ificatiorE can be screened by ~NA hybridization, Southern hybridization, and In situ hybridization. The amplified region can be cloned into a YAC: This vector and these methods also permit a functional analysis of cloned 25 chrom~some regions by reintroducing the cloned amplified region into mammalian cells.

~~1~7~°
B. Preparation of libraries in YAC vectors for cloning of centromeres and identification of functional chromosomal units Another method that rriay be used to obtain smaller-sized functional mammalian artificial chromosome units and to stone centromeric DNA involves screening of mammalian DNA YAC vector-based libraries and functional analysis of potential positive clones in a transgenic mouse model system. A marr~matian DNA library is prepared in a YAC nectar, such as YRT2 isee Schedl et al. (1393 Nuc. Aci s es.
21:4?83-4?8?fir which contains the murine tyrasinase gene. The library 1~ is screened for hybridization to mammalian telomere and centromere sequence probes. Positive clones are isatated and microin)ected into pronuclei of fertilized oocytes of NMRIIHan mice following standard techniques. The embryos are then transferred into NMRUHan foster mothers. Expression of the tyrosinase gene in transgenic offspring confers an identifiable phenotype tpigmentationy. The clones that give rise to tyrosinase-expressing transgenic mice are thus confirmed as containing functional mammalian artificial chromosome units.
Alternatively, fragments of SATACs may be introduced into the YAC vectors and then introduced into pronuclei of fertilized oacytes of NMRllhian mice following standard techniques as above. The clones that give rise to tyrosinase-expressing transgenic mice ate thus confirmed as containing functional mammalian artificial chromosome units, particularly centromeres.
C, Incorporation of l~eterotogous Genes into Mammalian Artificial Chromosomes through The Use of Homology Targeting Vectors As described above, the use of mammalian artificial chromosomes ' for expression of heteroiogou~ genes obviates certain negative effects that may result from random integration of heterolagous ptasmid DNA
into the recipient cell genome. An essential feature of the mammalian artificial chromosome that makes it a useful tool in avoiding the negative effects of random integration is its preser°~ce as an extra-genomic gene source in recipient cells. Accordingly, methods of specific, targeted incorporation of heterologous genes exclusively into the mammalian artificial chromosome, without extraneous random integration into the genome of recipient cells, are desired for heterologous gene expression from a mammalian artificial chromosome.
One means of achieving site-specific integration of heterologous genes into artificial chromosomes is through the use of homology targeting vectors. The heterologous gene of interest in subcioned-into a 1~ targeting vector which contains nucleic acid seduences that are homologous to nucleotides present in the artificial chromosome. The vector is then introduced into cells containing the artificial chromosome for specific site-directed integration into the artificial chromosome through a recombination event at sites ~f homology between the vector 1 FS and the chromosome. The homology targeting vectors may also contain selectable markers for ease of identifying cells that have incorporated the vectoc pnto the artificial chromosome as welt as lethal selection genes that are expressed only upon extraneous integration of the vector into the recipient cell genome. Two exemplary homology targeting vectorsp 2~ ~1CF-7 and prlCF-7-DTA, are described below.
1. Cor<stmctioro of hector ei~>F-7 hector otCF-7 contains the cystic fibrosis transmembrane conductance regulator ECFTRl gene as an exemplary therapeutic molecule-encoding nucleic acid that may be incorporated into mammalian 25~ artificial chromosomes for use in gene therapy applications.. This vector, which also contains the puromycin-resistance gene as a selectable marker, as well as the JaCCharOmYceS ~ereVlSiaB ~9ra~ gene (orotidine-5-phosphate decarboxylase], was constructed in a series of steps as follows.

a. Construction of piJf~A
Plasmid pURA was prepared by iigating a 2.6-kb Satl/Xhol .' fragment from the yeast artificial chromosome vectoir pY~,CS [Sigma; see also Burke et al. 119871 Science 236:806-812 for a description of YAC
vectors as well as GenBank Accession no. 001086 for the complete sequence of pYACS] containing the ~. cerevisiae ura3 gene with a 3.3-kb SaillSmal fragment of pFlyg [see, e~a., U.S. Paterat,Nos. 4,997,764, 4,686, 7 86 and 5,162,216,. and the description above]. Prior to ligation the Xhol end was treated with Klenow polymerase far blunt end ligation to the Smal end of the 3.3 kb fragment of pHyyg. ~I°hus, pURA contains the S. cerevisiae ura3 gene, and the E. ce~ti ColE1 origin ot~ replication and the ampicillin-resistance gene. The uraE gene is included to provide a means to recover the integrated construct from a mammalian cell as a YAC clone.
b. Constru~ctio~ of pL~F~2 Ptasmid pURA was digested with Satl and ligated to a 1.5-kb Sali fragment of pCEPUR. F'lasmid pCEPUR is produced by tigating the 1.1 kb SnaBl-Nhal fragment of pBabe-puro (Morgenstern ~, al. X1990[
Nucl. Acids Res. 1 x:3587-3596; provided by Clr. L. Szek~ly (tVticrobiotogy and Tumorbiology Center, Karolinska institutet, Stockholm]; see, also Tonghua et al. (1995) Chin. tied. J. (Beijing, Engl.
Ed.) 108:653-659; Couto et ai. (1994.) Infect. lmmun. 62:2375-2378;
Dunckley et al. ( 1992) FEBS Lett. 2,96:128-34; Feench. et al. ( 19958 Anal.
Biochem. 228:354-355; Liu et a1. (1995> Blood 85:1095-1103;
international PCT application Nos. W0 9520044; ~v('~ 9500178, and W~
. 9419456] t~ the Nhel-Nru1 fragment of pCEP4 [lnvitrogen].

-1f4-The resulting plasmid,.pUP2, contains the all the elements of AURA plus the puromycin-resistance gene linked to the Sii40 promoter and polyadenylation signal from pCEPUR.
c. Construction of ptlP-CFTR
The intermediate plasmid pUP-CFTR was generated in order to combine the elements of pUP2 into a plasmid along with the GFTR
gene. First, a 4..5-kb Sal! fragment of pCMV-CFTR that contains the CFTR-encoding ~NA (see, also, Riordan et al. (1983) Science 245:1066-1073, U.S. Patent No. 5,240,846, and Genbank Accession no. M28fi68 for the sequence of the CFTR gene] containing the CF'T'R gene only was ligated to Xhol-digested pCEP4 [lnvitrogen and also described herein] in order to insert the CFTR gene in the multiple cloning site of the Epsteird Barr virus-based ~EBVi sector pCEP4 (lnvitrogen, San Cliego, CA; see also Pates et al. (1985) Nature 313:812-815 see, also U.S. Patent No.
5,468,615) between the CMV promoter and SV40 polyadenylation signal. The resulting plasmid ervas designated pCEP-CFTR. Ptasmid pCEP-CFTR was then digested with Sall and the 5.8-kb fragment containing the CFTR gene franked by the CMV promoter and SV40 poiyadenylation signal was ligated to ,all-digested pUP2 'to generate pUP-CFTR. Thus, pUP-CFTR contains alb elements of pUP2 plus the CFTR gene linked to the CMV promoter and SV40 polyadenylation signal.
d. Construction of eICF-7 Plasmid pUP-CFTR was then linearizeci by partial digestion with EcoRl and the 13 kb fragment containing the CFTR gene, was ligated with EcoRI-digested Charon 4A~1 (see Blattner ~t al~ ( 1977) Science 196:161; Williams and Blattner X1979) J. Virol. 29:555 and Sambrook et a1. (1989) Molecular Cloning, A Laboratory Manual,, Second Ld., Cold Spring Harbor Laboratory Press, Volume 1, Section 2.18, for descriptions of Charon 4.A~t]. The resulting vector, ACF8, contains the Gharon 4A~i -18S°
bacteriophage left arm, the CFTR gene linked to the t:IVIV promoter and SV40 polyadenylation signet, the ura3 gene, the puromycin-resistance gene linked to the SV40 promoter and poiyadenylation signal, the thymidine kinase promoter IT1~), the CoIE 1 origin of repiicaton, the ampiicillan resistance gene and the Charon 4.AA~1 bacteriophage right arm.
The JlCF8 construct was then digested with ;Chol and the resulting 2'7.1 kb was ligated to the 0.4kb Xhot>Eco~?t fragment of pJ8P88 [described below, containing the aV40 polyA signal and the Ec:~RI-digested Charon 4A a right arm. The resulting vector aCF-7 contains the Charon 4A ~l left 1~ arm, the CFTR encoding DNA linked to the CMV promoter and SV4~
polyA signal, the ura3 gene; the puromycin resistance gene linked to the SV4~ promoter and poiyA signal and the Charan 4A ~1 right arm. The ~I DNA fragments provide encode sequences homologous to nucleotides present in the exemplary artificial chrorr~osomes.
1 S The vector is then introduced into cells containing the artificial chromosomes exemplified herein. Accordingly, when the linear ACF-7 vector is introduced into megachrornosome-carrying fusion cell lines, such as described herein, it will be specifically integrated into the megachromosome through recombination betweeh the homologous 2~ bacteriophage a sequences of the vector and the artificial chromosome.
2: Constr~actios~ of '/actor ACF-,?'-~TA
Vector ~ICF-7-DTA also contains alt the eleme~tts contained in aCF-7; but additionafiy contains a lethal selection marker the diptheria to~eir~-A (DT-Ap gene as well as the ampicillin-resistance gene arid an origin of 25 replication. This vector was constructed in a series of steps as f~llows.

v a. Construction of pJBP86 Plasmid pJBP86 was used in the construction of aCF-7, above. A
1.5-kb Sall fragment of pCEPUR containing the puromycin-resistance gene linked to the SV40 promoter and polyadenylation signal wns ligated to Hindlll-digested pJB8 [see, e.~.., 1sh-Horowitz et dal. J1981' ucieic Acids Res. 9_:2989-2998; available from ATCC as Accession Na. 37074a commercially available from Amersham, Arlington Heights, (LJ. Priar to ligatian the Sail ends of the 1.5 kb fragment of pCEPUR and th4 Hindlll linearized pJB8 ends were treated w°ith Klenow polymerise. The 1(9 resulting vector pJBP86 contains the puromycin resistance gene linked to the SV40 promoter and palyA signal, the 1.8 kb C(1S region of Charan 4A.~, the Co(E1 origin of replication and the ampicillin resistance gene.
b. Construction of pMEP-DTA
A 1 .1-kb XhoIISalI fragment of pMC1-DT-A [see, g_.g,,, Maxwell ~t a8. ('! 986) Cancer Res. 46:4660-4656] containing thze diptheria toxin-A
gene was ligated to Xhol-digested pMEP4 (Invitrogen, San Diego, CA] to generate pMEP-DTA. To produce pMC1-DT-A, the coding region of the DTA gene was isolated as a 800 hp PstIHindlll fragment from p2249-1 and inserted into pMC1 neopolyA (pMC1 available from StratagerteJ in place of the neo gene and under the control of the TK promotoer; Tine resulting construct pMC1 DT-A uvas digested with Windlll, the ends filled by Klenow and Sa~ll linkers were ligated to produce a 1 Q61 by TK-DTA .
gene cassette with an Xhol end [5'1 and a S~II end containing the 270 by TK promoter and the -~ 790 by DT-A fragment. This fragment was ligated into Xhol-digested pMEP4. .
Plasmid pMEP-DTA thus contains the DT-A cpne linked to the TK
promoter and SV40, CoiE1 origin of replication and the ampicillin-resistance gene.

-1 ~°
c. Construction of p~11~83-~TA9 Plasmid pJB8 was digested with Hendlll and Clal and ligated with an oGgorZUCleotide [see SECT ID NC3s. ~ and 8 for the sense and antisense strands of the oligonucleotide, respectively] to generate pJB83.
The oligonucleotide that was ligated to CIaI/Hindlll-digested pJ88 contained the recognition sites of Swal, Pacl and Srfl restriction endonucleases. These sites wall permit ready Iinearization of the p~CF-'~-DTA construct.
Next, a 1.~-kb Xhol(Sall fragment of pMlrP-DTA, containing the 1D DT-A gene was ligated to S~II-digested pJB83 to generate pJB83-DTAB.
d. Construction of ~1GF-7-DTA
The 7 2-by overhangs of aCF-7 were removed by Mung bean nuclease and subsequent T4 polymerase treatments. The resulting ~1 .1 kb linear eICF-7 vector was then figated to pFB83-DT'A9 which had been digested with Clal and treated v~rith T4 polymerase. The resulting vector, ~CF-7-DTA, contains all the elements of ACF-'~ as well as the DT-A gene linked to the Tif promoter and the S~l4D polyadenylation signal, the 1.8 kB Char~rn 4A ~i CC)S region, the ampicilin-resistance gene[from pJB83-DTA9] and the Col E1 origin of replication (frorrm pJ883-DT9A(.
2~ ~. Targeting vectors using luciferase markers: P'lasrnid pMCT-RIIC
F'lasmid pMCT-FtIJC El4kbpl vrvas constructed for site-specific targeting of the Renilla luciferase [see, e~ct., tJ.S. Patent Nos. 5,292,858 and 5,4.18,°155 f~c a description of 13~IA encoding guerrilla luciferase, and plasmid pTZrLuc-1, which can provide the starting material for 2a construction of such vectors] .gene to a mammalian artificial chromosome. The relevant features oaf this plass~id are the l3e~rilla iuciferase gene under transcriptianal control of the human cytomegalovirus immediate-early gene enhancer/prosx~ote~~ the hygromycin-resistance gene a, positive selectable marker? under the ° 1 ~~°
transcriptions! control of the thymidine kir~ase promoter. In particular, this plasmid contains plasmid pAG60 [see, e.a., U.S. Patent Nos.
5,118,620, 5.021.344, 5,063,162 and 4,94.6,952; :gee, also Colbert-Garapin et al. ( 1981 ) J. Mol. giol. 150:1-14], which includes ONA ti.e., the neomycin-resistance gene) homologous to the minichromosome, as wel6 as the iierrilla and tiygromyctn-resistance genes,, the I-IS1/-tk gene under control of the tk promoter as a negative selectabke marker for homologous recombination, and a unique Haal site for iinearizing the piasmid.
1~ This construct was introduced, via calcium phosphate transfection, into EC3/7C5 cells [see, LorerEZ ~t al. (1996) J. giolum. Chemilum.
11:31-371. The EC317C5 cells were maintained as a monolayer [see, GIuzW an ( 1981 ~ Cell ~3.-1 ~5-183]. Cells at 5CD~/o confluency in 100 mm Petri dishes were used for calcium phosphate transfection [see, Harper ~t 1a ai. (1981) Chromosome X3:431-4391 using 10,~g of iinearized pMCT-RUC per plate. Colonies originating from single trardsfected cells were isolated and maintained in F-12 medium containing hygromycin (30~
uglmL) and 10°/a fetal bovine serum. Cells were grown in 100 mm Petri dishes prior to the Renilla luciferase assay.
20 The Renilla luciferase assay was performed [see, e.~.; Matthews ~t al. (197'7) Biochemistry 16:85-91). Hygromycin-resistant cell lines obtained after transfection of EC3/~C5 cells with linearized plasmid pMCT-RUC ["B'° cell lines) were grown to 1 ~0% confluency for measure-meets of light emission in viva and. in vitro. Light Emission was 25 measured in viva after about 30 generatians as fotllows: growth medium was removed and replaced by 1 mL RPMI 164~ containing coeienterazine [1 mmol/L final cbncentration]. Light emission frort~ cells was then visuati~ed by placing the Petri dishes in a low light video image analyzer [Hamamatsu Argus-1001, An image was formed after 5 min. of photon accumulation using 100% sensitivity of the photon counting tube. For measuring light emission in vitro, cells were.trypsinized and harvested from one Petri dish, peileted, resuspendad in 1 mL assay buffer [0.5 rr~oIIL.
NaCI, 1 mmollL EDTA, 0.1 moIIL potassium phosphate, pFl 7.~.~ and sonicated on ice for 10 s. Lysates were than assayed in a Turner T~-20e luminometer for 10 s after rapid.injection of 0.5 mL of 1 mrrtollL
coeienterazine, and the average value of light emission was recorded as LU [1 LU = 1.6 x 106 hula for this instrument].
Independent cell lines of EC3I7C5 cells transfected with linearized ' piasmid pMCT-RUC showed different levels of .Renilla luciferase activity.
Similar differences in light emission were observed when measurements were performed on lysates of the same cell lines. This .variation in light emission was probably due to a position,effect resulting from the random integration of plasmid pMCT-RUC into the mouse genome, since enrichment for sits targeting of the luciferase gene was not performed in this experiment.
To obtain transfectant populations enriched in cells in which the luciferase gene had integrated into the minichromosome, transfected cells were grown in the presencd of ganciciovis. This negative selection medium selects against cells in which the added pMCT-RUC plasmid integrated into the host EC3I7C5 genome. This selection thereby enriches the surviving transfectant population with cells containing pMCT-RUC in the minichromosome. The cells s~rrviving this selection were evaluated in luciferase assays which revealed a more uniform levet of luciferase expression. Additior~atly, the results cal ire situ hybridization assays indicated that the Renilla luciferase gene was contained in the minichromosome in these cells, which further indicates successful targeting of pMCT°RUC into the minichromosome.

Plasmid pNEM-1, a variant of pMCT-RUC which also contains J
DNA to provide an extended region of homology to the minichromosome [see, other targeting vectors, below), was also used to transfect EC3/7C5 ce(Is. Site-directed .targeting of the Renlll<~ luciferase gene and the hygromycin-resistance gene in pNEM-~ to the minichromosome in the recipient EC3l7C5 cells was achieved. °~'his wa.s verified by DNA
amplification analysis and by irr situ hybridization. Additionaflyr luciferase gene expression was confirmed in luciferase assays of the trai~sfectants.
1E. E~rotein secretion targeting vects~rs '!~ isolation of heterologous proteins produced intracellufarly in mammalian cell expression systems requires cell disruption under potentially harsh conditions and purification of the recombinant protein from cellular contaminants. The process of pPOtein isolation may be greatly facilitated by secretion of the recombinantly produced protein into °I5 the extracellular medium where there are fewer contaminants to remove during purification. Therefore, secretion targeting vectors have been constructed for use with the mammalian artificial chromosome system.
A useful model vector for demonstrating production and secretion of heterologous protein in mammalian cells contains DNA encoding a 2~ readily detectable reporter protein fused to an effidient secretion signal that directs transport of the protein to the cell membrane and secretion of the protein frorri the cell. Vectors pr NCX-ILRUC and p~NCX-ILRUC~4, described below, are examples of such vectors. 'These vectors contain DIVA encoding an interleukin-2 t1L21 signal peptide-Renilla reniformis 25 fuciferase fusion protein. The IL-2 signal peptide [encoded by the sequence set forth in SEQ lD No. R) directs secrbtion of the iuciferase protein, to which it is linked, from mammalian cells. Upon secretion from the host mammalian cell, the IL-2 signal peptide is cleaved from the fusion protein to deliver mature, active: luciferase protein to the extracellular medium. Successful production and secretion of this heterologous protein can be readily detected by perfi~rming lueiferase assays which measure the light emitted upon exposure of the medium to the bioluminescent luciferin substrate of the luciferase enzyme.
5. Thus, this feature wiN be useful when artificial chromosomes afe used for gene therapy. The presence of a functional artificial chromosome carrying an IL-Ruc fusion with the accompanying therapeutic genes will be readily monitored. Body fluids or tissues can be sampled and tested for luciferase expression by adding iuciferin and appropriate cofactors and observing the bioluminescence.
1. Construction of Rrotein Secretion Vector pLlllCX-ILRUC
Vector pLNCX-1LRUC contains a human iL-2 signal, peptide-R. reniformis fusion gene linked to the human cytomegalovirus (CMV) immediate early promoter for constitutive expression of the gene in mammalian cells. The 1 b construct was prepared as follows.
a. Preparation of the !L-2 signal sequence-encoding GlyA
A 69-by DNA fragment containing DNA encoding the human IL-2 signal peptide was obtained through nucleic acid amplification, using appropriate primers for 1L-2, of an H~K 293 cell line (see, era., ll.S.
Patent No. 4,518,584 for an 1L-2 encoding DNA; see, also SEQ 1D No.
9; the 1L-2 gene and corresponding amino acid sequence is also provided in the Genbank Sequence Database as accession nos. K020b6 and J00264~. The signal peptide includes the first 20 amino acids shown in the translations provided in both of these Genbank entries and in 550. ID
NO. 9. The corresponding nucleotide sequence encoding the first 20 amino acids is also ~provide.d in these entries [see, e.g., nucleotides 2B3-52 of accession no. 1C02~56 and nucleotides 478-537 of accession no.
J00264), as well as in SEQ ID NO. 9. 'The amplification primers included an EcoRl site [GAATTC) for subcloning of the DNA fragment after ligation into pGEMT (Promega]. The forward primer is set forth in SEQ ID
No. 11 and the sequence of the reverse primer is set forth in SEQ ID No.
12.
TTTGAATTCATGTACAGGATGCAACTCCTG forward (SECt ID N~. 11 ]
TTTGAATTCAGTAGGTGCACTGTTTGTGAC revserse [SEQ 1D No. 121 b. Preparation of tlae R. reniformis luciferase-encoding DNA
The initial source of the R. reniformis luciferase gene vvas pfasmid, pLXSN-FiUC. Vector pLXSN [see, B.g_, U.S. Patent Nos.
1~ 5,324,655, 5,470,.730; 5,468,634, 5,358,866 and Miller et a1.
[°1989) Biotechniques 7:980] is a retrovirai vector capable of expressing heterologous DNA under the transcriptional control of the retroviral LTR;
it also contains the neomycin-resistance gene aperatively linked for expression to the SV40 early region promoter. The R. reniformis luciferase gene was obtained from plasrinid pTZrLuc-1 [see, e~cs., U.S.
Patent No. 5r292,658; see also the Genbank Sequence Database accession no. M63501; and see .also Lorenz et al. [ 1991 ) Proc. Natl.
Acad. Sci. U.S,A. 88:4438-4442] and is shown as SECT Its N~. 10. The 0.97 kb EcoRllSmal fragment of pTZrLuc-1 contains the coding region of the Renilla luciferase-encodig DNA. Vector pLXSN was digested with .
and ligated with the luciferase gene contained on a pLXSN-RUC, which contains the luciferase gene located operably linked to the viral LTR and upstream of the SV4D promoter, which directs expression of the , neomycin-resistance gene.
c. Fusion of DNA encoding the IL-2 Signal Peptide arid the ft. reniformis Luciferase Gene to Yield pLXSN-ILRU C
The pGEMT vector containing the IL-2 signal peptide-encoding DNA described in 1.a. above was digested with EcoRl, and the resulting ~0 fragment encoding the signal peptide was ligated to EcoRl-digested pLXSN-RUC. The resulting plasmid, called pLXSN-ILRUC; captains the IL-2 signal peptide-encoding D(~A located immediately upstream of the ~t_ reniformis gene in pLXSN-RUC. Ptasmid pLXSN-1LRUC was then used as a template for nucleic acid amplification of the fusion gene in order to add a Smal site at the 3' end of the fusion gene. The amplification product was. subcioned into linearized [EcoRl/Smat-digested) pGEMT
jPromega~ to generate ILRUC-pGEMT.
ci. Introduction of the Fusion Gene into a i/ector Containing Control Eler~nents far Expression in 1 D Mamrrnafian Cells Pfasmid tLRUC-pGEMT was digested with KSt~! and Smal to release a fragment containing the IL-2 signal peptide-luciferase fusion gene which was ligated to Howl-digested pLNCX. Vector pLNCX [sed, e~g., U.S..Patent Nos. 5,324,655 and 5,467,182; see, also Mifler_ and Rosman (19891 Biotechniaues 7:J80-990[ is a retroviral vector for expressing heterologous aNA under the control of the CMV promoter; it also contains the neomycin-resistance gene under the transcriptional contro6 of a viral promoter. The vector resulting from the ligation reaction was designated pLNCX-ILRUC, ~/ector pLNCX-ILRUC contains the IL-2 signal peptide-luciferase fusion gene located immediately downstream of the CMV promoter and upstream of the viral 3' LTR and poiyadenylation signal in pLNCX. This arrangement provides for expression of the fusion gene under the control of the CMV promoter.
Placement of the heterologous protein-encoding DNA [i.e., the luciferase genel in operative linkage with the IL-2 signal peptide-encoding DNA
provides for expression of the fusion in mammalian cells transfected with the vector such that the heterologous protein is secreted from the host cell into the extracellular medium.

2. Construction of Protein Secretion Targeting Vector pLNCX-ILRUCA
Vector pLNCX-ILRUC may be modified. so that it can be used to introduce the 1L-2 signal peptide-luciferase fusion gene into a marnmaliaw artificial chromosome in a host cell. To facilitate specific incorporation of the pLNCX-ILRUC expression vector into a mammalian artificial chromosome, nucleic acid sequences that are horriologous to nucleotides present in the artificial chromosome are added to the vector to permit site directed recombination.
Exemplary artificial chromosomes described herein contain ei phage DNA. Therefore; protein secretion targeting 'vector pLNCX-ILRUC.i was prepared by addition of ~ phage ~NA [from Charon 4A arms] to produce the secretion vector pLNCX-ILRUC.
3. Expression and Secretion of R. reniforrnis Luciferase from Mammalian Cells a. Expressian of R. reniformis Lueiferase Using pLilfCX-ILRUC
Mammalian calls [LMTlC' from tE:e ATCC] were transiently transfected with vector pLNCX-1LRUC [ -~ 10 dug] by electroporation [B10RA0, performed according to the manufacturer's instructions]. Stable transfectants produced by growth in G4~18 for neo refection (lave also been prepared.
Transfectants were grown and then analyzed for expression of iuciferase. To determine whether active lucifecase was secreted from the transfected cells, culture media were assayed for fuciferase by addition of coeientrazine [see, e.~o., Matthews et al. (1977) Biochemistry 16:85-91].
The results of these assays establish that vector pLNCX-1LRUC is capable of providing constitutive expression of heterologous DNA in mammalian host cells. Furthermore, the results demonstrate that the -~ ~5-human IL-2 signal peptide is capable of directing secretion of proteins fused to the C-terminus of the peptide. Additiona!!y, these data demonstrate that the R. reniformis luciferase protein is a highly effective reporter molecule, which i~ stable in a mammalian ce9l environment, and forms the basis of a sensitive, facile assay for gene expression.
b. Renilfa renfforfs 6uciferase appears to be secreted from LMTK~ cells.
(i~ fienflla luciferase assay of cel6 pellets The following cells were tested:
90 ~ cells with no vector: f_MTK- cells without vector as a negative Control;
cells transfected with pLNCX only;
cells transfected with .RUC-pLNCX jf?en~lla lucife~ase gene in pLNCX Vector);
~15 .cells transfected with pLNCX-ILRUC [vector containing the IL-2 leader sequence -+- Renilla lucifeease fusion gene in pLNCX vest~r].
Forty-eight hours after electroporation, the cells -and culture medium were collected. The cell pellet from 4 plates of cells was resuspended in 1 ml assay buffer and was lysed by sonfcation. Two 20 hundred ,u1 of the resuspended cell pellet was used for each assay for luciferase activity [see, ~, Matthews ~t al. (i877~ Biochemistry 16:85-31 ]. The assay was repeated three times and the average bioluminescence measurement was obtained.
The results showed that there was relatively low background 25 bioluminescence in the cells transformed with pLNCX or the negative , control cells; there was a low level observed in ths; cell pellet from cells containing the vector with the 1L-2 leader sequence-luclferase gene fusion and more than 5000 Rf_lJ in the sample from cells containing RUC=
- pLNCX.

(ii) Renilla luciferase assay of cell medium Forty milliliters of medium from 4 plates of cells were harvested and spun down. Two hundred microliters of medium was used for each luciferase activity assay. The assay was repeated several times and the average bioluminescence measurement was obtained. These results showed that a relatively high level of bioluminescence was detected in the cell medium from cells transformed with pLNCX-1LRUC; about 10-fold lower levels [slightly above the background levels in medium from cells , with no vector or transfected with pLNCX only was detected in the cells transfected with RUC-pLNCX.
(iii) conclusions The results of these experiments demonstrated that Renilla luciferase appears to be secreted from LMTK~ cells under the direction of the 1L-2 signal peptide. The medium from cells transfected with Renilla luciferase-encoding DNA linked to the DNA encoding the IL-2 secretion signal had substantially higher levels of Renilla lu.ciferase activity than controls or cells containing luciferase-encoding DNA without the signal peptide-encoding. DNA. Also, the differences between the controls and cells containing luciferase encoding-DNA demonstrate that the luciferase activity is specifically from luciferase, not from a r~on-specific reaction.
In addition, the results from the medium of RUC-pLNCX transfected cells, which is similar to background, show that the tuciferase activity in the medium, does not come from cell Iysis, but from secreted luciferase.
c. Expression ~f Fi. reniformis Luciferase Using pLNCX-tLRUCa To express the !L-~ signal peptide-R. reniformis fusion gene from an mammalian artificial chromosome; vector pLNCX-1LRUC~t is targeted for site-specific integration into a mammalian artificial chromosome through homotogous.recombination of the rl DNA sequences contained in -the chromosome and the vector. This is accomplished by introduction of pLNCX-lLRUCsI into either a fusion cell line harboring mammalian artificial chromosomes or mammalian host cells that contain mammalian artificial chromosomes. If the vector is introduced into a fusion cell line harboring the artificial chromosomes, for example through microinjection of the, vector or transfection of the fusion cei! line with the vector, the cells are then grown under selective conditions. The artificial chromosomes, which have incorporated vector pLNCX-ILRUC~i, are isolated fr~m the surviving cells, using purification procedures as described above, and 1 ~ then injected into the mammalian host cells.
Alternatively, the mammalian host cells may first be injected with mammalian artificial chrornosomes which have been isolated from a fusion cell sine. The host cells are then transfected with vector pLiVCX-1LRUCeI and grown.
~5 The recombinant host cells are then assayed for iuciferase expression as described above.
F. ~ther targeting vectors These vectors, which are based on vector pMCT-RUC, rely on positive and negative selection to insure insertion and selection for the 20 double recombinants. A single crossover results in incorporation of the DT-A~ which kills the cell, double crossover recombinations delete the DT-1 gene.
1 . Plasmid pNEM~ contains:
DT-A: Diphtheria toxin gene (negative selectable marker) 26 Hyg: I-lygromycin gene (positive selectable marker) ruc: Renilia luciferase gene (non-selectable marker) 1: LTR-MMTV promoter 2: TIC promoter 3: CMV promoter -1'~8~
MMR: Homology region (plasmid pAGfiO) 2, plasmid pNEM-2 and =~ are similar to pNEM 1 except for different negative selectable markers:
pNEM-1: diphtheria toxin gene as '°-'° selectable marker a pNEM-2: hygromycin antisense gene as "-" selectable marker pNEM-3: thymidine kinase EiSV-1 gene as °'-'° selectable marker , 3. Plasmid - e1 t~NA based homology:
pNEM~I-1: base vector pNEM~I-2: base vector containing pS = gene 1; LTR MMTV promoter 2: SV40 promoter 3: CMV promoter 4: ~rTIIA promoter $metallothionein gene promoter) - homology region (plasmid pAG60j 75 ~1 L.A, and ~1 R.A. homology regions for ~l left and right arms (~l gt-WES).

Microinjectiore of mammalian cells with ptasmid DNA
These procedures will be used to microinject MACs into eukaryotic cells, including mammalian and insect cells.
The microinjection technique is based on the use of small glass capillaries as a delivery system itato cells and has been used for.
introduction of ~NA fragments into nuclei [see, e~c~., Chalfie et af. ( 1994) Science 263:802-$a4~. It allows the transfer of almost any type of ~5 molecules, e.a., hormones, proteins, ~NA and RNA, into either the cytoplasm or nuclei -of recipient cells This techniq~ie has no cell type restriction and is more efficient than oxher methods, including Ca2+-mediated gene transfer and liposome-mediated gene transfer.
About 20-30% of the injected cells become successfully transformed.

°~ 79-Microinjection is pdrformed under a phase-contrast microscope. A
gloss microcapi(lary, prefilled with the DNA sample, is directed into a cell to be injected with the aid of a micr.omanipu(ator. An appropriate sample volume [1-1 Q pl] is transferred into the cell by gentle air pressure exerted by a transjecto,r connected to the capillary. Recipient cells are grown on .
glass slides imprinted with numbered sqcaares for convenient localization of the injected cells.
a, Materials and equipment Nuncfon tissue culture dishes 35-x-10 mm, mouse cell line EC317C5 'i~ Plasmid DNA pCH110 [Pharmacia], Purified Creen Florescent Protein (GFP~ [GFPs from Aequ~rea and Renifla have been purified and also ~NA
encoding CiFPs has been. clonedv see, e'4., Prasher ~t al. (1992) Gene 1 1 1 >229-233; International PCT Application No. VtIO 95/0743, °~ 5 _ ~~(SS Axiovert 1'Ot~ microscope, Eppendorf transjecfor 5246, Eppendorf micromanipulator 5171, Fppendorf Ce(loc~$eT""
coverslips, Eppendorf microloaders, Eppendorf femtotips and other standard equipment b. Protocr~l fc~r iryjeoting 20 C1 E Fibrdblast cells are grown in 35 mm tissue culture dishes E37° C, 5n/o COZ] until the cell density reaches 9Q%
conf(uency. The dishes are removed from the incubator and medium is added to about a 5 mm depth.
(2~ The dish is placed onto the dish holder 25 and the cells observed with 10 x objective; the focus is desirably above the cell surface.
[3y Plasmid or chromosomal DNA solution [1 ng/~rl] and GFP protein solution are further purified by centrifuging the DNA sample at a force sufficient to remave any particular debris [typically about 10,000 cpm far 10 minutes in a microcentrifuge~.
(4) °f~wo 2 ,~i of the DNA solution ( 7 ngl~el) is loaded into a microcapillary with an Eppendorf micr~oloader. During loading, the loader is inserted to the tip end of the microcapillary. CFP
6~ mglml) is loaded with the same procedure.
(5) The protecting sheath is rer~aved from the microcapillary and the microcapillary is fixed onto the capillary holder connected .with the micromanipuiator.
$6) The capillary tip is lowered to the surface of the medium and, is focussed on the cells gradualll.y until the tip of the capillary reaches the surface of a cell. 'fhe capillary is lawered further so that the it is inserted into the cell. Various parameters, such as the level of the capillary, the time and pressure, are determined for the particular '15 equipment. Far example, using the fibroblast cel9 line C5 and the above-noted equipment, the best conditions are: injection time 0.4 secondp pressure 80 psi. DNA can then be automatically injected into the nuclei of the cells.
(~9 After injection, the cells are returned to 2~ the incubator, and incubated for about 18-24 hours.
(8) After incubation the number of transformants can be determined by a suitable method, which depends upon the selection marker. For example, if green fluorescent proteirs is used, the assay can be perfori~.ed using UV light source and fluarescent 2.5r filter set at 0-24 hours after injection. If (~-gal-containing DNA, such as DNA-derived from pHCt ~ 0, has been injected, then the transformants can be assayed for (~-gad, -~ ~~1 (c) ~etectian of ~f3-galactosidase irr cells injected with plasrniid DIVA
The medium is removed from the culture plate and the cells are .
fixed by addition of 5 ml of fixation Solution I; (1 °!°
g)utaraldehydes 0.1 M sodium phosphate buffer, pH 7.~; 1 mM nIIgCla), and incubated far 't 5 minutes at 37° C. Fixation Solution I is replaced with 5 ml of X-gal Solution I1: [0.2% X-gal, 10 mM sodium phosphate buffer (pH 7.~), 15~
mM NaCI, 1 mM MgCl2, 3.3 mM K~,Fe(CN~sH2~, 3.3 mNE IC3Fe(CN)61, and the plates are incubated for 30-50 minutes at 37° C. The X-gal solution 1CD is removed and 2 ml of 70% glycerol is added to each dish. Blue stained cells are identified under a light microscope.
This method will be used to introduce a MACP particularly the MAC with the anti-HIV megachramosome, to produce a mouse model for anti-HIV activity.
E3CAMPLE 14.
Transgeriic [non-human) animals Transgenic (non-human) animals can be generated that express heterologous genes which confer desired traits, e~a., disease resistance, in the animals. A transgenic manse is prepared to serve as. a model of a 2~ disease-resistant animal. Genes that encode vaccines or that encode therapeutic molecules can be introduced into embryos or ES cells to produce ahimals that express the gene product and thereby are resistant to or less susceptible to a particular disorder.
The mammalian artificial megachromasome and others of the artificial chromosomes, particularly the SATACs, can be used to generate transgenic (non-human) animals. including mammals arid birds, that stabiy express genes conferring desired traits, such as genes conferring resistance to pathogenic viruses. The artificial chromosomes can also be used to produce-transgenlc (non-humans animals, such as pigs, that can produce immunofogicafly humanized organs for xerootranspfantation.
For example, transgenic mine corttaining a transgene encoding an anti-Ht!/ ribozyme p>"ovide a useful model for the development of stable transgenic !non-human) animals using these methods. The artificial chromosomes can be used to produce transgenic (non-humans animals, particufarty, cows, goats, mice, oxen, camels, pigs and sheep, that produce the proteins of interest in their milk; and to produce transgenic chickens and other egg-producing fowl, that prodc,cce therapeutic proteins or other proteins of interest in their eggs. For example, use of mammary gland-specific promoters for expression of heterologous ~NA in milk is known Esee, e~a. 11.5. Patent No. ~,~7~,37 6l. !n particular, a milk-specific promoter or a promoter, preferably linked to a milk-specific signs! peptide, specifically activated in mammary i:issue is operatively linked to the f~NA of interest, thereby pro~Jiding expression of that DNA
sequence in milk.
1. ~evelopment ~f Contrail Transgenic luiice Expressing Anti-HIV Ribozyme Control transgenic mice are generated in order to compare stability ~~ and amounts of transgene expression in mice developed using transgene DNA carried an a vector (control mice' with expression in mice derreloped using transgenes carried in an artificial megact~romosome.
a. ~e~aelopment of Coratroi 'Trar~sgentc f~lice Expressing ~-gatactosidase .
~ne set of control transgenic mice was generated by microinjectlan of mouse embryos with the ~3-galactosidase gene alone.
The mlcroinjectlori procedure used to introduce the pfasmid 1~NA int~ the mouse embryos is as described in 1~xample 13, but modified for use with embryos [see, ela., I-4ogan ~t al. ('! 99~) Manipulating the Mouse Er»fary~o, 3D .4 :Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring - ~ 8.~-Harbor, NY, see, especially. pages 255-2C4. and Appendix 3I. Fertilized mouse embryos (Strain CB6 obtained from Charles (River Co.) were injected with 1 ng of plasmid pCH110 tPharmacia) which had been linearized by digestion with BamHl. This piasmid contains the ,~-~S galactosidase gene linked to the SV40 late promoter. The ,S-galactosidase gene product provides a readily detectable marker for successful transgene expression. Furthermore, these control mice provide confirmation of the microinjection procedure used to introduce the plasmid into the embryos. Additionally, because the mega-chromosome that is transferred to the mause embryos in the model system (see below) also contains the j3-galactosidase gene, the control transgenic mice that have been generated by injection of pCH1 10 into embryos serve as an analogous system for comparisow of h~terologous gene expression from a piasmid versus from a gene carried on an artifical chromosome.
After injection, the embryos are cultured in modified HTF medium under 5°r6 C02 at 37°C for one day until they divide to form two cells.
The two-cell embryos are then implanted into surrogate mother female mice (for procedures see. Manipulating the Mouse Embryo. A Laboratory Manual (1994) Hogan et al., eds., Cold Spring Harbor Laboratory Press, Cold Spring Harbor. NY, pp. 127 et sec.].
b. Development of Control Transgenic Mice Expressing Anti-HtV Ribozyme One set of anti-HfV ribozyme gene-captaining control transgenic mice was generated by microinjection of mouse embryos with piasmid pCEPUR-i32 which contains three~different genes: (1) ANA encoding an anti-HlV ribozyme, (2) the puromycin-resistance gene and (3) the hygromycin-resistance gene. Plasmid pCEPUR-132 was constructed by ligating portions of piasmid pCEP-132 containing the anti-HiV ribozyme gene (referred to as ribozyme D by Chang et al. ((1990) Clin. Biotech.
2:23-31]; see also U.S. Patent No. 5,144,019 to Rossi et al.., particu-iarly Figure 4 of the patents and the hygromycin-resistance gene with a portion of plasmid pCEPUR containing the puromycin-resistance gene.
Plasmid pCEP-132 was constructed as follows. Vector pCEP4 (invitrogen, San ~iego, CA; see also Yates ~t a1. (1985 Nature 313:812-815~ was digested with Xhoi which cleaves in the multiple cloning site region of the vector. This --10.4-kb vector contains the hygromycin-resistance gene linked to the thyrnidine kinase gene promoter and polyadenytation signal, as well as the ampicillin-resistance gene and ColE1 origin of replication and Ei3NA-1 (Epstein-Barr virus nuclear antigen) genes and OriP. The rn~sltiple cloning site i:y flanked by the cytomegalavirus promoter and SV40 polyadenylation signal.
Xhol-digested pCEP4 was ligated with a fragrnent obtained by digestion of piasmid 132 (see Example 4 for a description of this plasmid, with Xhol and Sall. This Xhol/Satl fragment contains the anti-HIV
ribozyme gene linked at the 3' end to the SV~.O polyadenylation signal.
The plasmid resulting from this ligation was designated pCEP-132. Thus, in effect, pCEP-132 comprises pCEP4 with the anti-HtV ribozyme gene and SV40 pofyadenylation signal inserted in the rnulti~ple cloning site for CMV prom~ter-driven expression of the anti-HIV ribozyme gene.
To generate pCEPUR-132. pCEP-132 was ligated with a fragment of pCEPU,R. pCEPUR was prepared by ligating a 7.7-kb fragment generated upon Nhel/Nrul digestion of pCEP4 with a 1.1-kb NheIlSnaBl fragment of pBabe Isee Morgenstern and Land (19J0) l~ucteic Acids Res.
18:3587-3596 for a description of pBabel that contains the puromycin-resistance gene finked at the 5' end to the SV40 promoter. Thus, pCEPUR is made up of the ampicillin-resistance and EBNA1 genes, as well as the CotE1 and OriP elements from pCEP4 and the puromycin-° ~I $~°
resistance gene from pBabe. The puromycin-resistance gene in pCE~'UR
is flanked by the SV40 promoter (from pBabe) at the 5ro end and' the SV40 poiyadenylation signal (from pCEP~.? at the 3' end.
Plasmid pCEPUR was digested with Xhol and Sall and the a fragment containing the puromycin-resistance gene linked at the a°
end to the SV40 promoter was ligated with Xho!°digested CEP-132 to yield the -12.1-kb pfasmid designated pCEPI~R-132. Thus, pCEPUR-132, in effect, comprises pCEP-132 with puromycin-resistance gene and SV4.0 promoter inserted at the Xhol site. The main elements of pCEPUR-132 are the hygromycin-resistance gene (inked to 'the thymidine kinase promoter and polyadenylation signal, the anti-HIV ribozyme gene linked to the CMV promoter and SV40 polyadenylation signal, and the puromycin-resistance gene linked to the SV40 prorYaoter and polyadenylation signal. The plasmid also contains the ampicillin-resistance and EBNI~'I genes and the ColE1 origin of replication and OriP.
Zygotes were prepared from (C57BL/6JxCBA~/J) F1 female mice [see, ,e.~., Manipulatiwa (rte Mouse Embrvo. A Laboratory Manual X1994) liogan et al., eds., Coid Spring harbor Laboratory Press, Cold Spring Harbor, NY, p. 4.29), which had been previously mated with a dC57BL/BJxCBA/J) F1 male. The male pronuciei of these F2 zygotes were injected [see, Maninulating the Mouse Embrva~A Laboratory Manual (1994) Hogan ~t a1_, eds.~ Cold Spring Narb~r Labaratory Press, Cold Spring Harbor, NY) with pCEPUR-132 (-~3 jrg/ml), which had been linearized by digestion with Nrul. The injected eggs were then implanted 26 in surrogate mother female mice for development into transgenic offspring.
These primary carrier offspring were analyzed !as desca~ibed below) for the presence of the transgene in ~NA isa~ated from tail cells. Seven carrier mice that contained transgenes in their tail cells (but that may not carry the transgene in all their cells, i.e., they may be chimeric) were allowed to mate to produce non-chimeric or germ-line heterozygotes.
The heterozygotes were, in turn, crossed to generate hamozygote transgenic offspring.
2.. ~evelopment of Model Transgenic Mice Using Mammalian Artificial Chromosomes Fertilized mouse embryos are microinjected (as described above) with t'negachromosomes ( 1-1 ~ pL containing 0-1 chrbmosomes/pL) iso-lated from fusion cell line G3D5 or 3-11 D3 (described above). The megachrornasomes-are isolated as described herein. Megachromosomes isolated from either ~celi line carry .tl~e anti-HIV ribozyme (ribozyme D) , gene as well as the hygromycin-resistance and ;~-galactosadase genes.
The injected embryos are then developed into transgenic mice as described above.
Alternatively, the megachromasome-containing cell line G3D5' or H1 D3' is fused with mouse embryonic stern cells [sea, era., U.S. Patent No. 5,453,357, commerically available; sae Maniaulating~the Mouse Embryo, A Laboratory Manual (1994) Hogan et al., eds.; Cold Spring Harbor Laboratory Press, Cofd Spring Harbor, NY, pages X53-2991 2~ following standard procedures see also, e.a., Guide to Techniaues in Mouse Development in Methods in Enzvmologiy Vol. 25, Wassacman and De Pamphilis, eds. ( 1993), pages S~3-9321. (It is also possible to deliver isolated megachromosomes into embryonic stem cells using the Microcell procedure (such as that described above].) The stem cells are cultured in 26 the presence of a fibroblast [e=g_, 5T0 fibroblasts that are resistant to .
hygromycin and puromycin]. Cells of the resultant fusion cell line, which contains megachromosomes carrying the transgenes [i.e., anti-HIV
ribozyme, hygromycin-resistance and (3-galactosidase genes], are then transplanted into mouse blastocysts, which are in turn implanted into a -surrogate mother female mouse where development into a transgenic mouse will occur.
Mice generated by this method are chimeric,- the transgenes wii6 be expressed in only certain areas of the mouse, e.a., the head, and thus 5, may not be expressed in all cells.
3. Analysis of Transgenic Mice fog Transgene ~xp~esston Beginning when the transgenic mice, generated as described above, are three-to-four weeks. old, they can be analyzed for stable expression of the transgenes that were transferred into the embryos [or fertilized eggs] from which they develop. The transgenic mice may be analyzed in several ways as follows.
a, Analysis of Cells Obtained from the Transgenic Mice Cell samples (e~g., spleen, Liver and kidney cells, lymphocytes, tail ce(Is] are ebtained from the transgenic mice. Any cells may be tested for transgene expression. 1f, however, the mice are chimeras generated by microinjection of fertilized eggs or by fusion of embryonic stem cells with rryegachromosome-containing cells, only cells from areas of the mouse that carry the transgene are expected to express the transgene. If the c.eHs survive growth on hygromycin [or hygromycin and puromycin or neomycin, if the cells are obtained frorr~ rvrice generated by transfer of both antibiotic-resistance genes], this is one indication that they are stably expressing the transgenes. RNA isolated from the cells according to standard methods may also be analyzed by northern blot procedures to determine if the cells express transcripts that hybridize to nucleic acid probes based on the antibiotic-resistance genes. Additionally, cells obtained from the transgenic mice may also be analyzed for ,Q-gatactosidase expression using standard assays for this marker enzyme ffor example, by direct staining of the product of a reaction involving p~-galactosidase and the X-gal substrate, see, e~a., Jones (1988) EMB~
5:3133-3142, or by measurement of ~3-gaiact~sidase activity, see, e.g,_, Nfiller (1972) in Experiments in Molecular Genetics pp. 352-355~ Coid Spring Harbor Press). Analysis of ~-gaiactosidase expression is particularly used to evaluate transgene expression in cells obtained from conteol transgenic mice in which the only transgene transferred into the embryo was the ,8-gaiactosidase gene.
Stable expression of the anti-HIV ribozyme gene in cells obtained from the transgenic mice shay be evaVuated in several ways. First, DNA
isolated from the cells according to standard procedures may be subjected to nucleic acid amplification using primers corresponding to the ribozyme gene sequence. If the gene is contained.within the cells, an amplified product of pre-determined size is detected upon hybridization of the reaction mixture to a nucleic acid probe based on the ribozyme gene sequence. Furti-sermore, DNA isolated from the cells may be analyzed using Southern blot methods for hybridization to such a nucleic acid probe. Second, RNA isolated from the cells may be subjected to northern blot hybridization to determine if the cells express RNA that hybridizes to nucleic acid probes based on the ribozyme gene. Third, the cells rrsay be analyzed for the presence of anti-ivilV ribozyme activity as described, for example, in Chang ~t ate. ( 1990) Clin. Biotech. 2:23-31. In this analysis, RNA isolated from the cells is mixed with radioactively labeled HIV aaa target RNA which can b~ obtained by ire vitro transcription of ga_g gene template under reaction conditions favorable to in vitro cleavage of the a~g target, stash as those described. in Chang et ai. (1990) Ciin. Biotech. 2:23-31. After the reaction has been stopped, the mixture is analyzed by gel electraphoresis to determine if clea~rage products smaller in size than the whole template are detected~ presence -of such cleavage fragments is indicative of the presence of stably expressed ribozyme.
b. Analysis of Whole Transgenic Mice Whole transgenic mice that have been generated by transfer of the anti-H!V ribozyme gene (as well as selection and marker genes] into embryos or fertilized eggs can additionally be analyzed for transgene expression by challenging the mice with it,fection with HIV. 6t is possible for mice to be infected with HIV upon intraperitoneal injection with high-producing HIV-infected U937 cells Isee, e.g., Locardi et. al. (1992.) J. Virol. 66:1649-1654j. Successful infection may be confirmed by analysis of DNA isolated from cells, such as peripheral blood mononuclear cells, obtained frond transgenic mice that hava been injected with HlV-infected human cells. Ttee DNA of infected transgenic mice cells will contain HIV-specific saga and env sequences, as demonstrated 16 by, for example, nucleic acid amplification using HIV-specific primers. If the cells also stably express the anti-HIV ribozyme, then analysis of i~NA
extracts of the cells should reveal the smaller g_ag fragments arising by cleavage of the g_ag transcript by the ribc~zyme.
Additionally, the transgenic mice carrying the anti-H1V ribozyme gene can be crossed with transgenic mice expressing human CD4 (i.e., the cellular receptor for HIV) [see Gillespie et al. (1993) Mol. Celt, Biol.
13:2952-2958; Hanna et al. (1994) Mol. Cetl. l3iol.: 14:1084-1C>94; and Yeung et al. (1994) J. Exc. Med. 180:1911-1920, for a description of transgenic mice expressing human .CD4j. The offspring of these crossed transgenic mice expressing both the CG~4 and anti-H1V ribozyme transgenes should be mare resistant to infection [as a result of a reduction in the levels of active H!V in the ceIIsI than mipe expressing CD4 alone [without expressing anti-H1V ribozymel»

1 ~°
4. ~eveloprnent of transgenic chickens using artificial chromosomes' The development of transgenic chickens has many applications in .
the improvement of domestic poultry, an agricultural species of commercial significance, such as disease resistance genes and genes encoding therapeutic proteins. It appears that efforts in the area of chicken transgenesis have been hampered due to difficulty in achieving stable expression of transgenes in chicken cells using conventional methods of gene transfer via random introduction into recipient cells.
Artificial chr~mosomes are, therefore, particularly useful in the development of transgenic chickens because they provide for stable maintenance of transgenes in host cells.
a. Preparation of artificial chromosomes for introduction ~f transgenes into recipient chicken ce66s (i) Mammalian artificial chrom~sor~es Mammalian artificial chromosomes, such as the ~ATACs and minichromosomes described herein, can be modified to incorporate detectable reporter genes andlor transgenes of interest .for use in developing transgenic chickens. Alternatively, chicfcen-specific artifical chromosomes can be constructed using the methods herein. In particular, chicken artificial chromosomes (CAGs] can be prepared using the methods herein for preparing MACs; or, as described above, the chicken librarires can be introduced into MACS provided herein and the resulting MACS introduced into chicken ce(Is and those that are functional in chicken cells selected.
As described in Examples 4 and 7, and elsewhere herein, artificial chromosome-containing mouse LMTK°-derived cell lines, or minichromosome-containing cell lines, as well as hybrids thereof, can be transfected with selected DNA to generate MACs (or CACs] that have -191' integrated the foreign DNA for functional expression of heterologous genes contained within the DNA.
To generate MACs ar. CACs containing transgenes to be expressed in chicken cells, the MAC-containing cell lines may be transfected with DNA that includes ~t DNA and transgenes of interest operably (inked to a promoter that is capable of driving expression of genes in chicken cells.
Alternatively, the minichromosomes or MACS (Br CACsI, produced as described above, can, be isolated and introduced into cells, followed by targeted integration of selected DNA. Vectors for targeted integration are provided herein or can be constructed as described herein.
Promoters of interest include constitutive, inducible and tissue (or celil-specific promoters known to those of skill in the art to promote expression of genes in chicken cells. For example, expression of the IacZ
gene in chicken blastodermat cells and primary chicken fibroblasts has 16 been demonstrated using a mouse heat-shock protein 58 (hsp 68) promoter (phspPTiacZpA; see Brazolot ~t ai. (19913 Mol. Re~rod. Devel.
30:304-3121, a Zn2+-inducible chicken metaltothionein (cMt) promoter [pCBcMtlacZ; see Brazolot et al. (1991) Mol. Reprod. Devel. 30:304-312), the constitutive Rous sarcoma virus and chic(sen Q-actin promoters in tandem [pmiwZ; sae Brazolot et al. ~ 1991 ) Mot. Renrod. Devel, _3Q:304-312) and the constitutive cytomegalovirs~s (CMV9 promoter. ~f particular interest herein are egg-specific promoters that ire derived from genes, such as ovalbumin and lysazyme, that are expressed in eggs.
The choice of promoter wilt depend on a variety of factors, including, for exarnpfe, whether the transgene product is to be expressed throughout the transgenic.chicken or restricted to certain locations, such as the egg. Cell-specific promoters functional in chickens include the steroid-responsive promoter of the egg ovalbumin protein-encoding gene (see Gaub et ai. ( 7 98?) MBO J. x;2313-2320; Tora et gL (1988) I'M.BO

-~ ~2' J. 7:3771-377$; Park et at. (1995) Biochem. Mot. Blot. int. tAustratia?
36:811-816j.
(ii1 Chicken artificial chromosomes Additionally, chicken artificial chramosornes may be generated S using methods described herein. For example, chicken cells, such as primary chicken fibrobtasts [see l3razolot ~t al. 61991 ) I~ol. Rearod.
Devel. 30:304-312], may be transfected With DNA that encodes a selectable marker [such as a protein that confers resistance, to antibiotics) and that includes DNA (such as chicken satellite ~NA) that targets the introduced DNA to the pericentric region of the endogenous chicken chromosomes. Transfectants that survive growth on selection medium are then analyzed, using methods described herein, for the presence of artificial chromosomes~ including minichromosomes, and particularly SATACs. An artificial chromosome-containing transfectant 't 5 Celt tine may then be transfected with DNA encoding the transgene of interest [fused to an appropriate promoter) along with DNA that targets ..
the foreign DNA to the chicken artificial chromosome.
b. introduction of artificial chromosou~es carrying transgenes of interest into recipte~~t chicken cells Cell fines containing artificial chromosomes that harbor transgene(s) of interest (i.e., donor cells) may be fused with recipient chicken cells in order to transfer the chromosomes into the recipient cells, Alternatively, the artificial chromosomes may be isolated from the donor cells, for example, using methods described herein [see, e~a., Example 101, and directly introduced into recipient cells.
Exemplary chicken recipient cell tines include, but are not limited to, stage X biastoderm cells [see, e.~., Brazolot et al. (1991? Mol.
Rearod. Dev. 30:304--312; Etches et al. ( 19931 Poultry Sci. 12:882-889;

Petitte et al. ( 7 99~) Develo~ament 108:185-189] and chick zygotes jsee, era., Love et al. (1994) Biotechnoloav 12:5~-fi3].
For example, microoell fusion is one method far introduction of artificial chromosomes into avian cells (see, e~o., Dieken et ate. (d 199fi) Nature Genet. 12:174=182 for methods of fusing micrc~cells with DT40 chicken pre-B ce(!s]. In this method, microcells are prepart~d jfor example, using procedures described in Example 1.A.5] from the artificial chromosome-containing cell tines and fused with chicken recipient cells.
Isolated arfiificial chromosomes may be directly introduced into chicken recipient cell lines through, for example, lipid-imediated carrier systems, such as lipofection procedures [see, e~ci., Brazolot et al. ( 1991 ) Moi. Reprod. Dev. 30:30--312] or direct r~nicroinjection. Microin~ection is generally preferred for introdluction of the artificial chromosomes into chicken zygotes (see, ae=a., Love et al. 119941 Biotechnofoby 12:60-637.
c. Developrreent of transgenic chickens Transgenic chickens may be developed by injecting recipient Stage X blastoderm cells (which have received the artificial chrorruosomes) into embryos at a similar stage of development [see, e-g., Etches et ai.
41993) oultr Sci. 72:882-889; Petitte et al, 41990) Development 108:185-7 89; and Carsience et al. (1993) Develoament 1 17: 669-675].
The recipient chicken embryos within the shell are candled and allowed to hatch to yield a germline; chimeric chicken that will express the transgene(s) in some of its cells.
Alternatively, the artificial chromosomes may be intr~aduced int~
chick zygotes, for example through direct microinjection [see, e~a., Love et a!. (1994) Biotechnology 12:60-63j, which thereby are incorporated into at feast a portion of the cells in the chicken. Inclusion of a tissue-specific pr~moter, such arc an egg-specific promoter, will ensure appropriate expression of operatively-linked heteroiogous ~l~A.

The ~NA of interest nnay also be introduced into a minichromosome, by methods provided herein. The minichrornosome mey either be one provided herein, or one generated in chicken cells using the methods herein. The heterologous DNA will k>e introduced using a targeting Vector, such,as those provided herein, or constructed as provided herein.
Since rriodifications will be apparent to those of skill in this art, it is intended that this invention be limited only by the scope of t9~e T4J appended claims.

SEQUENCE LISTING
(1) GENERAL INFORMATION
(i) APPLICANT:
(A) NAME: The Biological Research Center of the Hungarian Academy of S<:iences (B) STREET.: Post Office Box 521 (C) CITY: H-6701 Szeged (D) STATE:
( E ) COUNTRY' : Hungary (F) POSTAL CODE {ZIP):
(i) APPLICANT:
{A) NAME: Chromos Molecular Systems, Inc.
(B) STREET: 6660 NW Marine Drive {C) CITY: Vancouver, BC
(D) STATE:
( E ) COUNTR~r : Canada (F) POSTAL CODE (ZIP): V6T 1Z4 (ii) TITLE OF TFiE INVENTION: ARTIFICIAL CHROMOSOMES, USES THEREOF AND
METHODS FOR PREPARING ARTIFICIAL CHROMOSOMES
(iii) NUMBER OF SEQUENCES: 34 (iv) ATTORNEY/A(sENT INFORMATION
(A) ADDRESSEE: Cowling Lafleur Hende:rson LLP
(B) STREET: 2600- 160 Elgin St.
(C) CITY: Ottawa (D) STATE: Ont (E) COUNTRY: Canada (F) ZIP: K1P 1C3 (G) REFERENCE NUMBER: O8-881145CA2 (v) COMPUTER READABLE FORM:
(A) MEDIUM TYPE: Diskette (B) COMPUTER: IBM Compatible (C) OPERATING SYSTEM: DOS
(D) SOFTWARE: FastSEQ Version 1.5 (vi) CURRENT APB?LICATION DATA:
{A) APPLICATION NUMBER: Not Yet Known (B) FILING DA~."E: 1997-04-10 (C) CLASSIFICATION:
(vi) PRIOR APPL==CATION DATA:
(A) APPLICATION NUMBER: 08/695,191 (B) FILING DA''E: 07-AUG-1996 (C) CLASSIFICATION:
(vi) PRIOR APPL?=CATION DATA:
(A) APPLICATION NUMBER: 08/682,080 (B) FILING DATE: 15-JUL-1996 (C) CLASSIFICATION:

(vi) PRIOR APPL7:CATION DATA:
(A) APPLICATION NUMBER: 08/629,822 (B) FILING DATE: 10-APR-1996 (C) CLASSIFICATION:
(2) INFORMATION FOR SEQ TD NO:1:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 1293 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D} TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYPE:
(vi) ORIGINAL SOURCE:
(ix) FEATURE:
(xi) SEQUENCE DESCRIPTION: SEQ ID NO:1:

TTCATGTCCTAAAGTGTATATTTCTCCTTTTCCGCGATTTTCAGTTTTCTCGCCATAT'.CC540 TCCTACAGTGGACATTTCTAAATTTTCCAACTTTTTCAATTTTTC7.'CGACATATTTGACG720 TGCTAAAGTGTGTATTTCTTATTTTCCGTGATTTTCAGTTTTCTCGCCATATTCCAGG.eC780 ACATTTCTAAATTATCCACCTTTTTCAGTTTTTCATCGGC.A.CATTTCACGTCCTAAAG7.'G1200 TGTATTTCTAATTTTCAGTGATTTTCAGTTTTCTCGCCAT.ATTCCAGGACCTACAGTG'7.'G1260 (2) INFORMATION FOR SEQ ID N0:2:
(i} SEQUENCE CHARACTERISTICS:
(A) LENGTH: 1044 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO

-19i-(iv) ANTISENSE: NO
(v) FRAGMENT TYPE:
(vi) ORIGINAL SOURCE:
(ix) FEATURE:
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:2:

AATATCTTCC

GTATCTACTCAGCTAACAGAGTTGAACCTTCCTTTGAGAGAGCAG'.CTTTGAAACACTCTT480 AACTAGACAGAAGCATTCTCAGAAACTTATTTGTGATGTGCGCCC'.CCAACTAACAGTGTT780 (2) INFORMATION FOR SEQ ID N0:3:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 2492 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYPE:
(vi) ORIGINAL SOURCE:
(ix) FEATURE:
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:3:
CTGCAGCTGGGGGTCTCCAATCAGGCAGGGGCCCCTTACT.ACTCAGATGGGGTGGCCGAG60 ATTACAATGGACACAGGAGGTTGGGACACCTGGAGTCACC.AAACAAAACCATGCCAAGAG420 AGGGCCCCTGCTGCCACCTAGTGGCTGATGGCATCCACATGACCC'I'GGGCCACACGCGTT540 TTTCCACCTATTCGAAACAATCACATAAAATCCATCCTGG.AAAAAGCCTGGGGGATGGC'.A660 AGAGTTCTTGTTTTTCCTTCAGCAATTTGTCATTTTTAAAAGAGT'CTAGCAATTCTAACA960 CATTTCTTGNNTTTNGGCTGTTTAACTTATTGTTTAGTTTTAATAATTTTTTATATAT'rT1140 TTTTGTGTATATCTACCTTTTGTGTCATTTGTTAAAATTCATTACC~AAACCCAAAGGCAG1320 GGCAAGTTGGGGAGCTAAGGCAGTAGCAGGAAAACCAGACAAAGAA.AACAGGTGGAGACT2040 TGAGACAGAGGCAGGAATGTGAAGAAATCCAAAATAAAATTCCCTCiCACAGGACTCTTAG2100 CCTCGACACTGACAGCAATAGGGTCCGGCAGTGTCCCCAGCTGCCAGCAGGGGGCGTA(~G2460 (2) INFORMATION FOR SEQ ID N0:4:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 28 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYPE:
(vi) ORIGINAL SOURCE:
(ix) FEATURE:
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:4:

(2) INFORMATION FOR SEQ ID N0:5:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 29 base pairs (B) TYPE: nucleic acid (C) STRANDEDNE;SS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYPE:
(vi) ORIGINAL SGURCE:
(ix) FEATURE:
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:5:

(2) INFORMATION FOR SEQ ID N0:6:
(i) SEQUENCE CHP.RACTERISTICS:
(A) LENGTH: 47 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: RNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYFE:
(vi) ORIGINAL SOURCE:
(ix) FEATURE:
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:6:

(2) INFORMATION FOR SEQ ID N0:7:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 25 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY. linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYPE:
(vi) ORIGINAL SOURCE:
(ix) FEATURE:
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:7:

(2) INFORMATION FOR SEQ ID N0:8:
(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 27 base pairs (B) TYPE: nucleic acid (C) STRANDEDNE;SS: single {D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETIC'.AL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYPE:
(vi) ORIGINAL SOURCE:
(ix) FEATURE:
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:8:
TAAATTTAAT TAATTCGGGC C'CGTCGA 27 (2) INFORMATION FOR SEQ ID N0:9:
(i) SEQUENCE CHR,RACTERISTICS:
(A) LENGTH: 69 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(D) OTHER INFORMATION IL-2 signal sequence (xi) SEQUENCE DESCRIPTION: SEQ ID N0:9:
ATG TAC AGG ATG CAA C'fC CTG TCT TGC ATT GCA CTA AGT CTT GCA CTT 48 Met Tyr Arg Met Gln Leu Leu Ser Cys Ile Ala Leu Ser Leu Ala Leu Val Thr Asn Ser Ala Pro Thr (2) INFORMATION FOR SEQ ID N0:10:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 945 base pairs {B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: CDNA
(vi) ORIGINAL SOURCE:
(ix) FEATURE:
(A) NAME/KEY: Coding Sequence (B) LOCATION: 1...942 {D) OTHER INFORMATION: Renilla Reinformis Luciferase (x) PUBLICATION INFORMATION:

PATENT NO.: 5,418,155 (xi) SEQUENCE DESCRIPTION: SEQ ID N0:10:

Ser Leu Lys Met Thr Se.r Lys Val Tyr Asp Pro Glu G~wn Arg Lys Arg Met Ile Thr Gly Pro Gln Trp Trp Ala Arg Cys Lys Gln Met Asn Val CTT GAT TCA TTT ATT AA.T TAT TAT GAT TCA GAA P.AA CAT GCA GAA AAT 144 Leu Asp Ser Phe Ile Asn Tyr Tyr Asp Ser Glu Lys H~_s Ala Glu Asn GCT GTT ATT TTT TTA CA.T GGT AAC GCG GCC TCT TCT TAT TTA TGG CGA 192 Ala Val Ile Phe Leu His Gly Asn Ala Ala Ser Ser Tyr Leu Trp Arg CAT GTT GTG CCA CAT AT'T GAG CCA GTA GCG CGG TGT ATT ATA CCA GAT 240 His Val Val Pro His Ile Glu Pro Val Ala Arg Cys I7.e Ile Pro Asp Leu Ile G1y Met Gly Lys Ser Gly Lys Ser Gly Asn Gly Ser Tyr Arg Leu Leu Asp His Tyr Lys Tyr Leu Thr Ala Trp Leu Asn Phe Leu Ile Tyr Gln Arg Arg Ser Phe Phe val Gly His Asp Trp Gly Ala Cys Leu GCA TTT CAT TAT AGC TAT GAG CAT CAA GAT AAG .ATC AAA GCA ATA GTT 432 Ala Phe His Tyr Ser Tyr G1u His Gln Asp Lys Ile Lys Ala Ile Val His Ala Glu Ser Val Val Asp Val Ile Glu Ser Trp Asp Glu Trp Pro Asp Ile Glu Glu Asp Ile Ala Leu Ile Lys Ser Glu Gl.u Gly Glu Lys ATG GTT TTG GAG AAT AAC TTC TTC GTG GAA ACC .ATG TTG CCA TCA AAA 576 Met Val Leu Glu Asn Asn Phe Phe Val Glu Thr :Met Leu Pro Ser Lys Ile Met Arg Lys Leu Glu Pro Glu Glu Phe Ala .Ala Tyr Leu Glu Pro TTC AAA GAG AAA GGT GA.A GTT CGT CGT CCA ACA TTA TCA TGG CCT CGT 672 Phe Lys Glu Lys Gly Glu Val Arg Arg Pro Thr Leu Ser Trp Pro Arg GAA ATC CCG TTA GTA AP,A GGT GGT AAA CCT GAC GTT GTA CAA ATT GTT 720 Glu Ile Pro Leu Val Lys Gly Gly Lys Pro Asp Val Val Gln Ile Val AGG AAT TAT AAT GCT TA,T CTA CGT GCA AGT GAT GAT TTA CCA AAA ATG 768 Arg Asn Tyr Asn Ala Tyr Leu Arg A1a Ser Asp Asp Leu Pro Lys Met TTT ATT GAA TCG GAT CC'A GGA TTC TTT TCC AAT GCT ATT GTT GAA GGC 816 Phe Ile Glu Ser Asp Pro Gly Phe Phe Ser Asn Ala Ile Val Glu Gly GCC AAG AAG TTT CCT AA,T ACT GAA TTT GTC AAA GTA AAA GGT CTT CAT 864 Ala Lys Lys Phe Pro Asn Thr Glu Phe Val Lys Val Lys Gly Leu His TTT TCG CAA GAA GAT GC'A CCT GAT GAA ATG GGA AAA TAT ATC AAA TCG 912 Phe Ser Gln Glu Asp Ala Pro Asp Glu Met Gly Lys Tyr Ile Lys Ser Phe Val Glu Arg Val Leu Lys Asn Glu Gln (2) INFORMATION FOR SEQ ID NO:11:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 30 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYPE:
(vi) ORIGINAL SOURCE:
(ix) FEATURE:
(xi) SEQUENCE DESCRIPTION: SEQ ID NO:11:

(2) INFORMATION FOR SEQ ID N0:12:
(i} SEQUENCE CHARACTERISTICS:
(A) LENGTH: 30 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYPE:
(vi) ORIGINAL SOURCE:
(ix) FEATURE:
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:12:

(2) INFORMATION FOR SEQ ID N0:13:
(i) SEQUENCE CHP,RACTERISTICS:
(A) LENGTH: 2434 base pairs (B) TYPE: nucleic acid (C) STRANDEDNE;SS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYF~E:
(vi) ORIGINAL SOURCE:
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:13:

AATCATTATC

AGGTCCAGACGACTGACACCATTAACACTTTGTCAGCCTCAGTGACTACAGTCATAGA'CG240 TACCCAATTGGATCTCCTCAGCATTTTCTTTCTTTAAAAAATGGG~.'GGGATTAATATTAT600 TTGGAGATACACTTTGCTGTGGATTAGTGTTGCTTCTTTGATTGG7.'CTGTAAGCTTAAGG660 CCCAAACTAGGAGAGACAAGGTGGTTATTGCCCAGGCGCTTGCAGC~ACTAGAACATGGAG720 CTTCCCCTGATATATGGTTATCTATGCTTAGGCAATAGGTCGCTGCiCCACTCAGCTCTTA780 TATCCCACGAGGCTAGTCTCATTGTACGGGATAGAGTGAGTGTGC'3.'TCAGCAGCCCGAGA840 (2) INFORMATION FOR SEQ ID N0:14:
(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 1400 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYPE:
(vi) ORIGINAL SOURCE:
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:14:

AGGTCCAGATACAACTAGATGTATTATGACAAATAACTCAGCAGGC~ATGTGAACAAAAGT240 CCTTTACCTACACACTGGGGATTTGACCTCTATCTCCACTCTCAT'.CAATATGGGTGGCCT960 TGAGAACGCGTCTAATAACAATTGGTGCCGAAACCCGGGTGATAA'.CGATTATCATCTACA1380 (2) INFORMATION FOR SEQ ID N0:15:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 13,69 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYPE:
(vi) ORIGINAL SC>URCE:
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:15:

W

ACCTAATAAAAATTAAATTAAGAAGGTGTGAATATACTACAGTAGGTAAATTATTTCA'.CT420 ATTAAACATCAGTCCCAAATTACAAACTTCAATAAAAGATTTGAC'.CCTCCAGTGGTGGCA1320 (2) INFORMATION FOR SEQ ID N0:16:
(i) SEQUENCE CHARACTERISTICS:
{A) LENGTH: 22118 base pairs (B) TYPE: nucleic acid (C) STRANDEDNE;SS: single {D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYFE:
(vi) ORIGINAL SOURCE:
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:16:

TGTCTTGAAGACACTTTGTAGGCCTCAATCCTGTAAGAGCCTTCC'.CCTGCTTTTCAAATG660 AATATAAAAT AAAA.ATTTTA AAGAATTTTA AAAAACTACA GAAATC:AAAC ATAAGCCCAC 1440 GAATCATATG TCTGAAAATA AAAGCCAGAA CCTTTTCTGC 'CTTTGTTTTC TTTTGCCCCA 1620 AATTGCCTCA GCTCTGCTCT AATTCTCTTT AF~AAAAAAAC AAACAAAAAA AAAACCAAAA 2460 CTACACAGAA AAACCATATC TCAGAAAAA.A AA.AAAGTTCC AAACACACAC ACACACACAC 2700 ACACACACAC ACACACACAC ACACACACAC ACACACACAG CGCGCC:GCGG CGATGAGGGG 2760 TTTTTTTTTT TTTTTTTCTC CAGAAGCCTT GTCTGTCGCT GTCACC:GGGG GCGCTGTACT 4080 CTTCCAAGCCGATGTGGCCCGGCCAGCTGGAGCTTCGGGTCTTTTi'TTTTTTTTTTCCTC4680 GGCTTCCAGGCCGATGTGGCCCGGTCAGCTGGAGCTTTGGATCTT'.CTTTTTTTTTTTTCT5340 TGTCAGGGTCGACCAGTTGTTCCTTTGAGGTCCGGTTCTTTTCGTTATGGGGTCATTT'T'T5520 GGGCCACCTCCCCAGGTATGACTTCCAGGCGTCGTTGCTCGCCTG'.CCACTTTCCTCCCTG5580 TCTTTTCTCTTCCCGGTCTTTCTTCCACATGCCTCCCGAGTGCAT'.CTCTTTTTGTTTTTT6060 CGTTGTGTTCTCTTGTTCTGTGTCTGCCCGTATCAGTAACTGTCT'.CGCCCCGCGTGTAAG6240 GAGGTGCTCCTGGAGCGTTCCAGGTTTGTCTCCTAGGTGCCTGCT'.CCTGAGCTGGTGGTG6720 CCGTTGCTGCGGAGCATGTGGCTCGGCTTGTGTGGTTGGTGGCTGGGGAGAGGGCTCCcsT7440 -20g-TCCGTCGCGT GCGTCCCTCT CGCTCGCGTC CACGACTTTG GCCGC'I°CCCG

CTGCGCCGCG CGTGGTGCGT GCTGTGTGCT TCTCGGGCTG TGTGG'rTGTG TCGCCTCGCC 7680 CCTCGGGGTC GAGAGGGTCC GTGTCTGGCG TTGATTGATC TCGCTr_TCGG GGACGGGACC 7800 ACCCGTGGCC GTGCTGTCGG ACCCCCCGCA TGGGGGCGGC CGGGCACGTA CGCG'I'CCGGG 8280 CATCTCTCGC GCAATGGCGC CGCCCGAGTT CACGGTGGGT TCGTCCTCCG CCTCCGCT'PC 8820 TCGCCGGGGG CTGGCCGCTG TCCGGTCTCT CCTGCCCGAC CCCCG'.CTGGC GTGGTCTTCT 8880 CTCGCCGGCT TCGCGGACTC CTGGCTTCGC CCGGAGGGTC AGGGGGCTTC CCGGTTCCC_C 8940 GAGCCCCTGC CGCACCCGCC GGTGTGCGGT TTCGCGCCGC GGTCAGTTGG GCCCTGGC(~T 9060 TCGTTGGTGT GGGGAGTGAA TGGTGCTACC GGTCATTCCC TCCCGCGTGG TTTGACTG'.CC 9240 CGGCCCGGTG CGGTCGACGT TCCGGCTCTC CCGATGCCGA GGGGT~.'CGGG ATTTGTGCCG 9360 GTTGGCTTTG CCGCGTGCGT GTGCTCGCGG ACGGGTTTTG TCGGACCCCG ACGGGGTCCsG 9480 TCCGGCCGCA TGCACTCTCC CGTTCCGCGC GAGCGCCCGC CCGGCTCACC CCCGGTTT<~T 9540 CCTCCCGCGA GGCTCTCCGC CGCCGCCGCC TCCTCCTCCT CTCTCCiCGCT CTCTGTCCCG 9600 CCTGGTCCTG TCCCACCCCC GACGCTCCGC TCGCGCTTCC TTACCTGGTT GATCCTGCC~A 9660 AGTGAAACTG CGAATGGCTC ATTAAATCAG TTATGGTTCC 'rTTGGTCGCT CGCTCCTCTC 9780 CTACTTGGAT AACTGTGGTA ATTCTAGAGC TAATACATGC CGACGCdGCGC TGACCCCCCT 9840 TCCCGGGGGG GGATGCGTGC ATTTATCAGA TCAAAACCAA CCCGG'T.'GAGC TCCCTCCCGG 9900 CTCCGGCCGG GGGTCGGGCG CCGGCGGCTT GGTGACTCTA GATAACCTCG GGCCGATCC~C 9960 TCGCCGTGCC TACCATGGTG ACCACGGGTG ACGGGGAATC .AGGGTTCGAT TCCGGAGAGG 10080 GAGCCTGAGA AACGGCTACC ACATCCAAGG AAGGCAGCAG GCGCGCAAAT TACCCACTC:C 10140 CGACCCGGGG AGGTAGTGAC GAAAAATAAC AATACAGGAC TCTTTC."GAGG CCCTGTAA7.'T 10200 CAGCCGCGGT AATTCCAGCT CCAATAGCGT ATATTAAAGT TGCTGC,'AGTT AAAAAGCTCG 10320 TAGTTGGATC TTGGGAGCGG GCGGGCGGTC CGCCGCGAGG CGAGTC'ACCG CCCGTCCCCG 10380 CCCCTTGCCT CTCGGCGCCC CCTCGATGCT CTTAGCTGAG TGTCCC'GCGG GGCCCGAAGC 10440 GTTTACTTTG AAAAAATTAG AGTGTTCAAA GCAGGCCCGA ~:~CCGCCTGGA TACCGCAGCT 10500 AGGAATAATG GAATAGGACC GCGGTTCTAT TTTGTTGGTT 'rTCGGAACTG AGGCCATGAT 10560 TAAGAGGGAC GGCCGGGGGC ATTCGTATTG CGCCGCTAGA ~:~GTGAAATTC TTGGACCGGC 10620 GCAAGACGGA CCAGAGCGAA AGCATTTGCC AAGAATGTTT 'TCATTAATCA AGAACGAAAG 10680 TCGGAGGTTC GAAGACGATC AGATACCGTC GTAGTTCCGA C:CATAAACGA TGCCGACTGG 10740 CGATGCGGCG GCGTTATTCC CATGACCCGC CGGGCAGCTT CCGGGAAACC AAAGTCTT7.'G 10800 GCGTTCAGCC ACCCGAGATT GAGCAATAAC AGGTCTGTGA TGCCC'rTAGA TGTCCGGGGC 11160 AACCCGTTGA ACCCCATTCG TGATGGGGAT CGGGGATTGC AATTA'rTCCC CATGAACGAG 11280 GGTCGGCCCA CGGCCCTGGC GGAGCGCTGA GAAGACGGTC GAACT'TGACT ATCTAGAGGA 11460 CGCGTGCGTC CCGGGTCCCG TCGCCCGCGT GTGGAGCGAG GTGTC'ZGGAG TGAGGTGAGA 11640 GGTTTTTGAC CCGTCCCGGG GGCGTTCGGT CGTCGGGGCG CGCGC'rTTGC TCTCCCGGCA 11880 CCCATCCCCG CCGCGGCTCT GGCTTTTCTA CGTTGGCTGG GGCGG'rTGTC GCGTGTGGGG 11940 CCCGACCCGC GCCGCCGGCT TGCCCGATTT CCGCGGGTCG GTCCT(JTCGG TGCCGGTCGT 12300 TGCGTCGATG AAGAACGCA.G CTAGCTGCGA GAATTAATGT GAATTGCAGG ACACATTGAT 12600 CGCGCTCGCG GCTTCTTCCC GCTCCGCCGT TCCCGCCCTC GCCCG'~'GCAC CCCGGTCCTG 12900 GTTTGGGTCT TGCGCTGGGG GAGGCGGGGT CGACCGCTCG CGGGG'7.'TGGC GCGGTCGCCC 13200 GGGAGGGAGA GGGCCTCGCT GACCCGTTGC GTCCCGGCTT CCCTGGGGGG GACCCGGCC~T 13560 ATTAGTCAGC GGAGGAAAAG AAACTAACCA GGATTCCCTC .AGTAACGGCG AGTGAACACiG 13860 GTTGCTTGGG AATGCAGCCC AAAGCGGGTG GTAAACTCCA 'TCTAAGGCTA AATACCGGCA 14100 CGAGACCGAT AGTCAACAAG TACCGTAAGG GAAAGTTGAA .zIAGAACTTTG AAGAGAGAGT 14160 GTCACGCGTC TCCCGACGAA GCCGAGCGCA CGGGGTCGGC GGCGA'TGTCG GCTACCCACC 14880 CGACCCGTCT TGAAACACGG ACCAAGGAGT C'rAACGCGTG CGCGAGTCAG GGGCTCGTCC 14940 AGGTGGAGCA CGAGCGTACG CGTTAGGACC CGAAAGATGG TGAAC'rATGC TTGGGCAGGG 15120 CGAAGCCAGA GGAAACTCTG GTGGAGGTCC GTAGCGGTCC TGACG'rGCAA ATCGGTCGTC 15180 AGAAGCCCGG CTCGCTGGCG TGGAGCCGGG CGTGGAATGC GAGTGCCTAG TGGGCCAC'TT 15420 TGGAGCCGCC GCAGGTGCAG ATCTTGGTGG TAGTAGCAAA TATTCAAACG AGAACTTT(sA 15900 GGGGAGAGGG TGTAAATCTC GCGCCGGGCC GTACCCATAT CCGCAGCAGG TCTCCAAG(~T 16320 CGCGGCGCCC CCGCCTCGGC CGGCGCCTAG CAGCCGACTT .AGAACTGGTG CGGACCAGGG 17040 CGCGCATGAA TGGATGAACG AGATTCCCAC TGTCCCTACC 'TACTATCCAG CGAAACCACA 17280 CCGCCGGGCG TCGGGACCGG GGTCCGGTGC GGAGAGCCGT TCGTC'TTGGG AAACGGGGTG 18300 GCTCCCTCGC TGCGATCTAT TGAAAGTCAG CCCTCGACAC AAGGG'TTTGT CTCTGCGGGC 18480 GTTGGAGGGG CGGGAGGGGT TTTTCCCGTG AACGCCGCGT TCGGCCsCCAG GCCTCTGGCG 18900 TGGAACCTTA AGGTCGACCA GTTGTCCGTC TTTCACTCAT TCATA7.'AGGT CGACCGGTGG 19680 TACACAAACA TGCACTTTTT TTAAAATAAA TTTTTAAAAT AAATGCGAAA ATCGACCAF~T 20100 GCAGACTTCT GAGTTCGAGG CCAGCCTGGT CTACAGAGGA .ACCCTGTCTC GAAAAATGAA 20340 AATAGATAGA TGGATAGAGT GATACAAATA TAGGTTTTTT 'TTTCAGTAAA TATGAGGTTG 20520 ATTAACCACT TTTCCCTTTT TAGGTTTTTT TTTTTTTCCC ~~TGTCCATGT GGTTGCTGGG 20580 ATTTGAACTC AGGACCCTGG CAGGTCAACT GGAAAACGTG fiTTTCTATAT ATATAAATAG 20640 TGGTCTGTCT GCTGTTTGTT TGTTTGCTTG CTTGCTTGCT 'TGCTTGCTTG CTTGCTTGCT 20700 CAATTTTGGA GTAAAGGTGT GCTACACCAC TGCCTGGCAT TATTATCATT ATCATTA'I'TA 20880 GATTTTTGTA AAGATTACTT TTCTTAGTCT GAGGAAAA.AA TAAAATAATA TTGGGCTA.CG 21120 TTCCCAGACG GCCTTTTGAG AATAAAATGG GAGGCCAGAA CCAAAGTCTT TTGAAT_~1AAG' 21540 (2) INFORMAT7:ON FOR SEQ ID N0:17:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 4;999 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
{v) FRAGMENT TYFE:
(vi) ORIGINAL SOURCE:
{xi) SEQUENCE DESCRIPTION: SEQ ID N0:17:
GCTGACACGCTGTCCTCTGGCGACCTGTCGTCGGAGAGGTTGGGCCTCCGGATGCGCGC:G60 GCGTCGGCTCCGCCTGGGCCCTTGCGGTGCTCCTGGAGCGCTCCGG'GTTGTCCCTCAGGT 840 GTGTGACCCACCCTCGGTGAGAAAAGCCTTCTCTAGCGATCTGAG.AGGCGTGCCTTGGGG1440 CGCGTGACCCCCTCCGTCCGCGAGTCGGCTCTCCGCCCGCTCCCG'TGCCGAGTCGTGACC1800 CCCGGCGTCCGCGTCCCCCGGCGCGCGCCTTGGGGACCGGGTCGG'TGGCGCGCCGCGTGG2400 TCGCCGAGGGCCGGTCGGCCGCCCCGGGTGCCCCGCGGTGCCGCC(zGCGGCGGTGAGGCC2580 TCGGCCGGGCCCCGGGCCCTCGACCGGACCGGCTGCGCGGGCGCTGCGGCCGCACGCC(aC3180 TGGTTGATCCTGCCAGTAGCATATGCTTGTCTCAAAGATT.AAGCCATGCATGTCTAAGTA3720 CGCTCGCTCCTCTCCTACTTGGATAACTGTGGTAATTCTA~;~AGCTAATACATGCCGACGG3840 TCGGGCCGATCGCACGCCCCCCGTGGCGGCGACGACCCAT'TCGAACGTCTGCCCTATCAA4020 GATTCCGGAGAGGGAGCCTG.AGAAACGGCTACCACATCCAAGGAAGGCAGCAGGCGCGC:A4140 AATTACCCACTCCCGACCCGGGGAGGTAGTGACGAAAAAT,AACAATACAGGACTCTTTCG4200 AATCCTTTAA

ACCCGGCCCGGACACGGACAGGATTGACAGATTGATAGCTCTTTC'rCGATTCCGTGGGTG4980 GTGGTGCATGGCCGTTCTTAGTT.GGTGGAGCGATTTGTCTGGTTAATTCCGATAACGAAC5040 CGGCAGGCGCGGGTAACCCGTTGAACCCCAT'~'CGTGATGGGGATCGGGGATTGCAATTAT5280 CCCGCGCCTCCACCGCGGACTCCGCTCCCCGGCCGGGGCCGCGCCCiCCGCCGCCGCCGCG6300 GCGCCCCGCGCCGTGGGGGCGGGAACCCCCGGGCGCCTGTGGGGTGGTGTCCGCGCTC<~C6420 GGGTGGCGGGGGGGAGAGGGGGGCGCGCCCGGCTGAGAGA~GACGGGGAGGGCGGCGCCGC7020 GTCGGCGGTGGGGGCGTGTTGCGTGCGGTGTGGTGGTGGG~3GAGGAGGAAGGCGGGTCCG7260 CTCGGACCCG TCCCCCCGAC CTCCGCGGGG GAGACGCGCC GGGGCGTGCG GCGCCCGTCC '7740 GCGAGGGGGG TCTCCCCCGC GGGGGCGCGC CGGCGTCTCC '1'CGTG(iGGGG GCCGGGCCAC 9000 GGATAGCTGG CGCTCTCGCA GACCCGACGC ACCCCCGCCA CGCAG7.'TTTA TCCGGTAAAG 9660 TAAGAAGCCC GGCTCGCTGG CGTGGAGCCG GGCGTGGAAT GCGAG2'GCCT AGTGGGCCAC 9780 TTTTGGTAAG CAGAACTGGC GCTGCGGGAT GAACCGAACG CCGGG'TTAAG GCGCCCGATG 9840 CCGACGCTCA TCAGACCCCA GAAAAGGTGT TGGTTGATAT AGACAt~CAGG ACGGTGGCCA 9900 AAAATGGATG GCGCTGGAGC GTCGGGCCCA TACCCGGCCG '.CCGCCGGCAG TCGAGAGTGG 10020 ACGGGAGCGG CGGGGGCGGC GCGCGCGCGC GCGCGTGTGG TGTGCGTCGG AGGGCGGCCiG 10080 GAGAAGGGTT CCATGTGAAC AGCAGTTGAA CATGGGTCAG 'TCGGTCCTGA GAGATGGGCG 10380 CTGGCATGTT GGAACAATGT .AGGTAAGGGA AGTCGGCAAG CCGGATCCGT AACTTCGGGA 10800 -21 ~-CGCGGCGGGT GTTGACGCGA TGTGATTTCT GCCCAGTGCT CTGAA'rGTCA AAGTGAAGAA 11640 ACTGGCTTGT GGCGGCCAAG CGTTCATAGC GACGTCGCTT 'TTTGA'i'CCTT CGATGTCGGC 12360 CCGTCCTTCC GTTCGTCTTC CTCCCTCCCG GCCTCTCCCG CCGACCGCGG CGTGGTGG~CG 13080 CTCGGCCCGC GGTGGAGCTG GGACCACGCG GAACTCCCTC 'TCCCACATTT TTTTCAGCCC 13800 CTGCCTCCTC CTTTTTCGCT TTTAGGTTTT GCTTGCCTTT 'TTTTTTTTTT
TTT°TTTTTTT 13920 TTTTTTCTTT CTTTCTTTCT TTCTTTCTTT CTTTCTTTCT 'rTCTTTCTTT CGCTTGTCTT 13980 CTCTCGCTCT CGCCCTCTCT CTCTTCTCTC TCTCTCTCTC 'rCTCTCTCTG TCTCTCGCTC 14100 TTTCTTTCTT TCTTTCTTTC TTTCTTTCTT TCTTTCTTTC TCTCTCTCTC TCTCTCTC'I'C 15480 TCCTTCCTTT TTTCAATCTT ATTTTCTGAA CGCTGCCGTG TATGA~-1CATA CATCTACACA 15780 AACTATGTAA ATGATATTTC CATAATTAAT ACGTTTATAT TATGTTACTT TTAATGGA'rG 15900 AATATGTATC GAAGCCCCAT TTCATTTACA TACACGTGTA TGTATATCCT TCCTCCCT'.PC 15960 CGACCAAACG GTCGTTCTGC CTCTGATCCC TCCCATCCCC ATTACC:TGAG ACTACAGGCG 16500 CGCACCACCA C..ACCGGCTGA CTTTTATGTT GTTTCTCATG TTTTCCGTAG GTAGGTATGT 16560 CCTGCCTGCC TGCCTGCCTA TCAATCGTCT TCTTTTTAGT .ACGGATGTCG TCTCGCTT'7."A 16860 CGGCCTCCCG GAGTGCTGTG ATGACACGCG TGGGCACGGT .F.1CGCTCTGGT CGTGTTTGTC 17040 TGTCGCCCAG GGTGGAGTAC GATGGCGGCT CTCGGCTCAC ~~GCACCCTCC GCCTCCCAGG 17220 TTCGGTGCCG AAACCTCCCG AGGGCCTCCT TCCCTCTCCC CCTTG'TCCCC GCTTCTCCGC 17820 CCCAGCATTG TAAAGGGTGC GTGGGTATGG AAATGTCACC TAGGA'rGCCC TCCTTCCCTT 18240 CAGATCAAAC ACTATTTCCG GGTCCTCGTG GTGGGATTGG TCTCTCTCTC TCTCTCTC'rC 18480 CGCCCCACCC TCCACCCGTT GGCTGACGAA ACCCCTTCTC TACAATTGAT GAAAAAGA'TG 18660 GTGGATCGCT TGGGGCCGGG AGTTCGAGAC CAGGCTGGCC GACGTGGCGA AACCCCGTCT 1$780 CTCTGAAAAA TAGAACGATT AGCCGGGCCT GGTGGCGTGG GCTTG(iAATC ACGACCGCTC 18840 TTCTTTCTTT CTTTCTTTCT TTTTCTTTTT CTCTCTTCCC CTCTTTCTTT CCTGCCTTr_C 19260 TCGATTTAGT GTCATGCCTC TTTCACCACC ACCACCACCA CCGAAC;ATGA CAGCAAGGAT 20640 CTGTGGCCCT TACGCTCAGA ATGACGTGTC CTCTCTGCCG TAGGT'rGACT CCTTGAGTCC 21480 CCTAGGCCAT TGCACTGTAG CCTGGGCAGC AAGAGCCAAA CTCCG~1NCCC CCACCTCCTC 21540 GGCCAACGTG GTGAAACCCC GTCTCTACTG AAAATACGAA .ATGGAciTCAG GCGCCGTGGG 21780 CTCTCCCTCC CTGTTTGTTT CTCTCTCTCC CTCCCTGTCT GTTTC'CCTCT CTCTCTTTCT 22320 GTCTGTTTCT GTCTCTCTCT GTCTGTCTAT GTCTTTCTCT GTCTG'.CCTCT TTCTCTGTCT 22380 TCTGTCTGTC TCTCCCTCCC TTTCTGTTTC TCTCTCTCTC TCTCTC_TCTC TCCCCCTCTC 22560 TCTCTCTCTC TCTCTCTCTC TCCCTGTCTG TCTGTTTCTC TCTATCTCTC GCTGTCCA'.CC 22740 GAGGCCGGGT CCCCGCTTGG ATGCGAGGGG CATTTTCAGA CTTTTC:TCTC GGTCACGTGT 22980 CGGCTAAATA CCGCGTGTTC TCATCTAGAA GTGGGAACTT .ACAGATGACA GTTCTTGCAT 23160 GGGCAGAACG AGGGGCzACCG GG1VACGCGGA AGCCTGCTTG AGGGRGGAGG GGYGGAAGGA 23220 GGTGATGAAA TCATCTGCAC ACTGAACACC CCCGTCACAA ~;~TTTACCTAT GTCACAGTCT 23340 TGCTCATGTA TGCTTGAACG ACAAATAAAA GTTCGGGGGG ~,GAGAACIAGAG GAGAGAGAGA 23400 TTCTGGCCTT TTGGGAGAAC GTTCAGCGAC AATGCAGTAT 'TTGGGCCCGT TCTTTTTTTC 23580 GTCCTCTCTG CCATAGGTTG ACTCCTTGAG TCCCCTAGGC CATTGC'ACTG TAGCCTGGGC 24000 TGTCTGTTTC TCTCTGTTCG TCTCTGTCTT TCTCTCTGTG TCTCT'rTCTC TGTCTGTCTG 24600 CTCTCTCTCT CTCTNNNCCC TCCCTGTCTG TTTCTCTCTG TCTCCCTCTC TTTCTGTC'rG 25020 TTTCTCACTG TCTCTCTCTG TCTGTCTGTT TCATTCTCTC 'I°GTCTCTGTC

TGTCTCTCTC TGTCCGTCTC TGTCTTTTTC TGTCTGTCTG TCTCTCTCTT TCTTTCTG°rC

TTTCTCTCTG TCTCTCTGTC CATCTCTGTC TTTCTATGTC TGTCTCTCTC TTTCTCTC'.CG 26160 CCCTCTCTCT CCCTCCCTTT CTGTTTCTCT CTCTCTCTCT TTCTGTCTGT TTCTCTCT~~T 26340 TCTGTTTCTC TCTCTGCCTC TCTCTCTCTC TGTCTGTCTC TTTCTC'.TGTC TGTCTGTC7.'C 26580 C.ACTGTGTCT GTCTTCTGTC TTACTCTCTT TCTCTTGCCT GCCTCTCTGT CTGTCTGTCT 26880 CTCTCCCTCC ATGTCTCTCT CTCTCTCTCA CTCACTCTCT CTCCGTCTCT CTCTCTTTC.'T 26940 GTCTGTTTCT CTCTCTGTCT GTCTCTCTCC CTCCATGTCT ~(:TCTCTCTCT CTCTCACTC".A 27000 GTTTCTTTGT CTGTCTGTCT GTCTGTCTGT CTGTCTCTCT ~~TCTCTCTCT CTCTCTCTCT 27180 CTCTCTGTTT GTCTTTCTCC CTCCCTGTCT GTCTGTCTGT C:TCTCTCTCT CTGTCTCTGT 27240 CCGTACTTCT CCTATTTCCC CGATAAGTCT CCTCGACTTC AACAT.AAACT GTTAAGGCCG 27600 GACGCCAACA CGGCGAAACC CCGTCTCTAC TAAAAATACA AAGCTGAGTC CzGGAGCGGTG 27660 GGGC.AGGCCC TGTAATGCCA GCTCCTCGGG AGGCTGAGGC GGGAGAATCG CTTGAACCAG 27720 GGAAGCGGAG GCTGCAGGGA GCCGAGATCG CGCCACTGCA CTACGGCCCA GGCTGTAG.AG 27780 TGCTGACGGA CATTTGC.AGG C.AGGCATCGG TTGTCTTCGG GCATCACCTA GCGGCCACTG 27900 TTATTGAAAG TCGACGTTGA CACGGAGCzGA GGTCTCGCCG ACTTC.ACCGA GCCTGGGGCA 27960 CTCGCCTAGG GAACCTCCGC CCTGGGGGCC CTATTGTTCT TTGATCGGCG CTTTACTT'TT 28080 C.AAGTTGCCC CCCGGCTCCC CCCACTACCC ACGTCCCTTC ACCTTAATTT AGTGAGNCGG 28200 CGATCTCATT CTTGCCAGGC TGAC.ATTTGC ATCGGTGCsGC GTCAGGCCTC ACTCGGGGGC 28380 CACCGTTTTT GAAGATGGGG GCGGC:A.CGGT CCCACTTCCC CGGAGGCAGC TTGGGCCGAT 28440 GGCATAGCCC CTTGACCCGC GTGGGCAAGC GGGCGGGTCT GCAGT'.CGTGA GGCTTTTCCC 28500 CTGAAAACTA ATAACTTTNC TCACTTAAGA TTTCCAGGGA CGGCGCCTTG GCCCGTGT'rT 28740 TTTCTTTTC.A GGTGAAGTAG AAATCCCCAG TTTTCAGGAA GACGTCTATT TTCCCCAAGA 28860 CCCCTCTCTC TGTCTCTCTG TCTGTCTCTG TCTCTCTCTT TCTCTCiTCTG TCTTCTCTCT 29220 CTCTCTCTCT CTGTGTCTCT CTCTCTCTGC CTGTCTGTTT CTCTC'7:'CTCT GCCTCTCTCT 29280 TCTCTGCCTG CCTGTCTCTC TC..ACTCTCTC TCTCTGTGTG TCTCTCTCTC TCTTTCTGTT 29580 TGTCTTTCCT TCTCTCTGTC TCTGTCTCTC TCACTGTGTC 'TGTCTTCTGT CTTAGTCTCT 29700 CTCTCTCTCT CTCCCTGTCT GTCTGTCTCT CTCTCTCTCT CCCCCTGTCT GTTTCTCTC:T 29760 TCTCTGTCTG TCTCTCTCTC TCTCTCCCCC TGTCGGCTGT 'TTCTCTGTCT CTGTCTGTGT 29940 CTCTCTTTCT GTCTGTTTCT CTCTGTCTGT CTTTCTCTCT C'.TGTCTCTTT CTCTCTGTCT 30000 GTGTATGTGT GTGTGTGTGT ~,~TGTGTGTGT CTGCCTTCTG TCTTACTCTC TTTCTCTGCC 30120 TGTCTGTCTG CCTGTCTGTT "TGTCTCTCTC TCTCTGCCTG 'TCTCTCTCCC TTCCTGTCT'G 30180 TTTCTCTCTC TTTCTGTTTC "TCTCTGTCTC TGTCCATCTC 'TGTCTT'TCTC CGTCTGTCTC 30240 TTTATCTGTC TCTCTCCGTC 'TGTCTCTTTA TCTGTCTCTC TCTCTCTTTC TGTCTTTCTC 30300 TCTCTGTGTA TCGTTGTCTC 'TCTCTGTCTG TCTCTGTCTC TGTCTCTCTG TCTCTCTCTC 30360 TCTCTCTCTC TCTCTGTCTG 'TCTGTCCGTC TGTCTGTCTC GGTCTCTGCG TCTCGCTATC 30420 ACTGGCGAGT TGATTTCTGG :~CTTGGATAC CTCATAGAAA CTACATATGA ATAAAGATCC 30600 TGTCCCACCG AGGTCAAATG GATACCTCTG CATTGGCCCG AGGCC'rCCGA AGTACATCAC 31020 TCTCAGCGCC ACCATGGCCG GCTCATTTTT TTTTTTTTTT TTTTTGGTAG ACACGGGG'rT 31860 CTGTGCTAAT GATAGTGAAA GTGAAGACAA AAGAAAGGCT ATCTA'.CTTTG TGGTTAGAAT 32100 GGACTACAGG TGCCCGCCAC CACGCCCAGC TAATCTTTAT ACTTTTAATA GAGACGGGC~T 32880 GAAGTAGGAC CACACTTTTT CCTATCTTAT TCAGTTGATA .ACAATATGAC CTAGGTAGTA 33120 ATTTCCTATG TGCCTACTTA TACACGAGTA CAAAAGAGTA .AAACAGAGAG ACTGCTAAAT 33180 TAAAGGGTAC GTGAAGTTCT TCATAGTAAC TCCGTAAACT GGAACACTGT CAAAA.AGCAG 33240 ATGAGTTCAC TTCAGAGTTT GTTCAAGACA TACGTTTCGT :AAGGAP,ACAT CTTAGTTAGA 33540 AGTTATTCAG CAGTAGGTAC CATCCCTAAG TATTTTTCAC CAAATC"CGTG ACAATAAAGA 33600 TTCATTAGCA CTTACCATGC CTTACAATGT CTAGGATTGA CCCTGP.TAGC ATTTCGAAAA 33720 CAAGCTAATG CTTTGTCCAG 'TTCTTCAGTG AAGACAACTC ACGCCCTAAT GCGCTATAGG 33780 CATAAGCATC ATTTGGATCC .ACTTCGAGAG TTCTCTGGAA GAATTGAATC GCAATATCGT 33840 TCTGTCTTGC AATATACATG TCCCGACGAT GGAAGGGGAA .AGCGAGCTGA ATCACCAAAC 35280 CCGTGCCCGG CTAACTTTTT GTATT'T'TGAG TAGAGATGGG GTTTCACTGT GGTAGCCAGG 37140 -zz~.-GATGACAGGC GCGAGCCTAC CGCGCCCGGA CCCCCCCTTT CCCCT'TCCCC CGCTTGTCTT 37440 CTGCGTCCCC CCAGGAGCCC TGGTCGATTA G'TTGTGGGGA TCGCCTTGGA GGGCGCGGTG 37860 AGTGAGCCGA GATTGCGCCA TCGCACTCCA G'TCTGAGCAA CAAGAGCGAA ACTCCGTCTC 38640 AGAACAACCC CACCGTGACA TACACGTACG CTTCTCGCCT TTCGAGGCCT CAAAC.ACGTT 39180 TCCTGCCTCA GCCTCCTGAT TAGCTGGGAT TACAGGCATG GGCCACCGTG CTGGC'TGATG 39720 TTGCTTGCTT GCTTGCTTTC GTGCTTTCTT GCTTTCCTGT TTTCTTTCTT TCTTTC~'TTC 40020 TCTTTCTTTT GTTTCTTTCT TGCTTGCTTT CTTGCTTGCT TGCTTTCGTG CTGTC'TTGTT 40320 CCATGTTGCT CAGGCTGGTC TCCAACTCCC G.~CCTCCTGT GATGCGCCCA CCTCGGCCTC 40980 TTTATTTCTT TCGTTTCCAC GCGTTTACTT A'I°ATGTATTA ATGTAAACGT

TTATATGCAA ACAACGACAA CGTGTATCTC TGCATTGAAT ACTCTTGCGT ATGGT'APATA 41160 CGTATCGGTT GTATGGAAAT AGACTTCTGT A'rGATAGATG TAGGTGTCTG TGTTA.TACAA 41220 TCTTCCTCTC CTTCGTGTTT TTCTTCCTTC C'rTTCTTCCT TTCTCTCCTT CTTTAGGTTT 41400 (2) INFORMATION FOR SEQ ID N0:18:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 175 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYPE:

(vi) ORIGINAL SOURCE:
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:18:

(2) INFORMATION FOR SEQ ID N0:19:
(i) SEQUENCE CHAR.ACTERISTTCS:
(A} LENGTH: 755 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic 77NA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYPE:
(vi) ORIGINAL SOURCE;
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:19:

(2} INFORMATION FOR SEQ ID N0:20;
(i} SEQUENCE CHARACTERISTICS;.
(A) LENGTH: 463 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii} MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYPE:
(vi) ORIGINAL SOURCE:
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:20:

-zz~-(2) INFORMATION FOR SEQ ID N0:21:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 378 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYPE:
(vi) ORIGINAL SOURCE:
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:21:

TTTAAATCCTTTAAGCAG 37g (2) INFORMATION FOR SEQ ID N0:22:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 378 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYPE:
(vi) ORIGINAL SOURCE:
(xi) SEQUENCE DESCRIPTION: SBQ ID N0:22:
GATCCATTGGAGGGCAAGTCTGGTGCCAGCAGCCGCGGTAA'rTCCAGCTCCAATAGCGTA60 GCCGCGAGGCGAGTCACCGCCCGTCCCCGCCCCTTGCCTCTCGGCGCCCCCTCGA'rGCTC180 AAGCAGGCCCGAGCCGCCTGGATACCGCCAGCTAGGAAAT.AATGGAATAGGACCGCGGTT300 CCTTATTGCGCCCCCCTA 37g (2) INFORMATION FOR SEQ ID N0:23:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 719 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYPE:
(vi) ORIGINAL SOURCE:
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:23:

(2) INFORMATION FOR SEQ ID N0:24:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 685 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYPE:
(vi) ORIGINAL SOURCE:
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:24:

(2) INFORMATION FOR SEQ ID N0:25:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 33 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYPE:
(vi) ORIGINAL SOURCE:
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:25:

(2) INFORMATION FOR SEQ ID N0:26:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 35 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYPE:
(vi) ORIGINAL SOURCE:
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:26:

(2) INFORMATION FOR SEQ ID N0:27:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 33 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYPE:
(vi) ORIGINAL SOURCE:
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:27:

(2) INFORMATION FOR SEQ ID N0:28:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 34 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYPE:
(vi) ORIGINAL SOURCE:
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:28:

(2) INFORMATION FOR SEQ ID N0:29:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 80 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYPE:
(vi) ORIGINAL SOURCE:
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:29:

(2) INFORMATION FOR SEQ ID N0:30:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 19 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYPE:
(vi) ORIGINAL SOURCE:
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:30:

(2) INFORMATION FOR SEQ ID N0:31:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 25 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO

(v) FRAGMENT TYPE:
(vi) ORIGINAL SOURCE:
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:31:

(2) INFORMATION FOR SEQ ID N0:32:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 26 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYPE:
(vi) ORIGINAL SOURCE:
(xi) SEQUENCE DESCRIPTTON: SEQ ID N0:32:

(2) INFORMATION FOR SEQ ID N0:33:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 26 base pairs (B} TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYPE:
(vi) ORIGINAL SOURCE:
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:33:

(2) INFORMATION FOR SEQ ID N0:34:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 21 base pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D} TOPOLOGY: linear (ii) MOLECULE TYPE: Genomic DNA
(iii) HYPOTHETICAL: NO
(iv) ANTISENSE: NO
(v) FRAGMENT TYPE:
(vi) ORIGINAL SOURCE:
(xi) SEQUENCE DESCRIPTION: SEQ ID N0:34:

Claims

1. A method for amplifying nucleic acid, comprising:
introducing a nucleic acid molecule into a cell, wherein the nucleic acid !molecule includes a sequence of nucleotides that target it to or, amplifiable region of a chromosome in the cell;
growing the cell; and identifying from among the resulting cells those that include a chromosome with a portion that has undergone amplification.

2. The method of claim 1, wherein the nucleic acid includes sequence from a ribosomal DNA unit.

3. A method for amplifying nucleic acid, comprising:
introducing a DNA fragment into a cell, wherein the DNA
fragment comprises a selectable marker;
growing the cell under selective conditions to produce cells that have incorporated the DNA fragment or a portion thereof that comprises the selectable marker into a chromosome; and identifying from among the resulting cells those that include a chromosomes or fragment thereof with a portion that has undergone amplification.

4. The method of claim 3, wherein the DNA fragment comprises DNA
from a ribosomal DNA unit.

5. The method of claim 1 or claim 3, wherein the cell is an animal cell or plant cell.

6. A method for amplifying a nucleic acid, comprising:
introducing a nucleic acid fragment comprising a sequence of nucleotides targeted to an amplifiable region of a chromosome into a cell under conditions whereby the fragment integrates into the chromosome.

7. The method of claim 6, further comprising replicating the cell.

8. The method of claim 6 or claim 7, wherein tine targeting sequence of nucleotides is selected from among those that target the molecule to the pericentric heterochromatic region of a chromosome.

9. The method of claim 6 or claim 7, wherein tree targeting sequence comprises DNA from a ribosomal DNA unit.

10. The method of claim 6 or claim 7, wherein the targeting sequence comprises an origin of replication or an amplification promoting sequence (APS).

11. A nucleic acid molecule, comprising:
nucleic acid encoding a gene product or gene products;
a selectable marker; and a sequence of nucleotides targeted to an amplifiable region of a chromosome in a cell.

12. The nucleic acid molecule of claim 11, wherein the targeting sequence of nucleotides is selected from among those that target the molecule to the pericentric heterochromatic region of a chromosome.

13. The nucleic acid molecule of claim 11, wherein the targeting sequence comprises DNA from a ribosomal DNA unit.

14. The nucleic acid molecule of claim 11, wherein the targeting sequence comprises an origin of replication or an amplification promoting sequence (APS).

15. The nucleic acid molecule of claim 11 that is a plasmid.

16. The nucleic acid molecule of claim 15, wherein the cell is an animal cell.

17. A method for amplifying a nucleic acid, comprising:
introducing a nucleic acid fragment that comprises DNA from a ribosomal DNA unit into a cell under conditions that produce cells that have incorporated the nucleic acid fragment or a portion thereof that comprises the DNA from the ribosomal DNA
unit into a chromosome of the cell.

18. A method for amplifying a nucleic acid, comprising:
introducing a nucleic acid fragment that comprises an origin of replication or an amplification promoting sequence into a cell under conditions to produce cells that have incorporated the nucleic acid fragment or a portion thereof that comprises the origin of replication or an amplification promoting sequence into a chromosome of the cell.

19. The method of claim 17 or claim 18, wherein the cell is an animal cell or plant cell.

20. A plasmid comprising a nucleic acid molecule consisting essentially of one or more of the nucleotide sequences set forth in SEQ ID
Nos. 18, 19, 20, 21, 22, 23 and 24.

21. A plasmid comprising a nucleic acid molecule comprising one or more of the nucleotide sequences set forth in SEQ ID Nos. 13, 14 and 15.

22. The method of claim 1, wherein the nucleic acid comprises an origin of replication or an amplification promoting sequence (APS).

23. The method of claim 1, wherein the nucleic acid comprises a sequence of nucleotides that target it to a pericentric heterochromatic region of a chromosome.

24. The method of claim 6, wherein the cell is an animal cell or a plant cell.

25. The method of claim 6, wherein the cell is a mammalian cell.