CA1312027C - Methods and compositions for expression of competent eukaryotic gene products - Google Patents

Abstract

A B S T R A C T

Heat-shock gene control elements from different eukaryotic cells are utilized to provide expression of competent gene products in the same or similar eukaryotic as well as in procaryotic host cells. The control elements can be joined to a suitable eukaryotic replication system to form an expression vector, or may be joined to the gene of interest and introduced directly into the host genome.

Description

1 3 1 ~027 The industrial production of proteins coded by specific genes such as those encoding hormones, clotting factors, virus proteins, insulin, interferon and other~, involves selecting and isolating gene sequences from viral, eukaryotic and other RNA or DNA, splicing of the sequences in the form o~ DNA into DNA vectors to provide recombinant DNA, and introducing said recombinant DNA into host cells capable of expressing these genes. The vectors may impart to the transformed host cells a phenotypic trait used for isolation and cloning purposes. The gene products are isolated from cell cultures by usual techni~ues.
Up to now, efforts in this field have usually been based on the adaptation of microorganisms as host cells for expressing genes of interest. Indeed, bacteria are often the organisms of choice since they can be grown rapidly, in large quantiti~s and at low cost. Foreign DNA can be introduced easily into bact~rial cells by using vectors such as plasmids, cosmids, viruses and the like.
Use of bacterial hosts, however, is not always ~ufficient to obtain expression of mature eukaryotic proteins. Many important eukaryotic proteins are modified by glycosylation, acetylation, phosphorylation, specific proteolytic cleavage and other forms of processing. In many instances post-transcriptional and post-translational modifications are crucial in determining the final biological , `'~ .L~, - 2 - 13~2027 properties of protein products. Non-proteolytic post-translational mcdifications of certain proteins such as glycosylation, acetylation, phosphorylation and others may not occur correctly, if at all, in b~cteria or cell types significantly different from the cell type in which the gene product of interest is normally produced in the organism. The correct form of these modifications may prove criti-cal in the synthesis of fully conpetent gene products b~ molecular cloning techniques. For example, when the gene product is a glyco-protein, such as the haemagglutinin of influenza virus, the precise nature of glycosylation may influence the efficiency of antikodies raised against this synthetic protein to protect human beings or ani-mals against influenza virus infections. Moreover, when glycosyla-tion, acetylation or phosphorylation or other difications are requir-ed to stabilize, activate, or mediate intracellular transport or ex-cretion of a protein, the precision of these moaifications may be critical to the utility of these genetically engineered protein pro-ducts in practice.
Because of the above-mentioned shortccmings of using bacteria for ~he synthesis of complex gene products, there have been a numr ber of attempts to introduce D~ encoding specific eukaryotic pro-teins into eukar~otic cells. DNA can be intrcduced by co-transfor-mation into suit.~ble mutant cells which are deficient in the produc-tion of a pQrticular enzyme, such as thymidine kinase or hypoxanthi-ne phosphoribosyl transferase. Then by culturing the cells in a selec-tive medium deficient in the enzyme product, trar.sformed cells may be selected for by their abilit~ to grow, i.e., their ability to pro-duce the enzyme. Alternatively, cells can be co-transfected in a posi-tive sense to add a gene that will transduce cells to become selec-tively resistant to a drug or selective medium such methotrexate, neomycin, or others. Although m~ny of these atte~pts have enjoyed some measure of success, in most cases the yield of mature gene pro-ducts has been quite limited.
In order to maximize the yields of expression of fully compe-tent gene products of interest, it would be desirable that the host expression unit system employed consists of an expression vector deriv-ed from one cell type or organism that is capable of synthesizing the said fully co~petent gene products, and that this expression vec-1 31 ~027 tor directs the synthesis of said gene products in the same or simi-lar cell type or organism as host.

The heat-shock phencmenon has been studied most extensively in rroscphila melanoqaster. Fbr a review, see Ashburner and Bonner ~1979) Cell 17: 241-254. When Drosge~__a cells or organs, normally at about 25~C, are exposed to a heat treatment at 35-37C, a family of heat-shock genes is activated and most of the genes active at 25C are no longer transcribed. Seven genes ccde for polypeptides with mole-cular weights between 22,000 and 84,000 daltons. ~uring heat treat-ment these heat-shock pol~peptides are synthesized aLmost exclusi-vely and after 8 hrs, represent 10~ of the total cellular protein ~Arrigo, P. (1979) Ph. D. Thesis, University of Geneva). During heat treatment of Drosophila cells, much of the polysome bound m~NA codes for heat-shock proteins (McRen~ie et al. (1975) 72: 1117-1121; Mi-rault et al. (1978) Cold Spring Harbor Symp. Quant. BiolO 42: 819 - 827).
All seven Drosophila heat-shcck protein genes have been cloned.
See, Livak et al. (1978) PNAS 75: 5613-5617; Schedl et al. (197B) Cell 14: 921-929; Craig et al. (1979) Cell 16: 575-588; ~olmgren et al. (1979) Cell 18: 1359-1370; ~dsworth et al. (1980) PN~S 77: 2134 - 2137; Corces et al. (1980) PNAS 77: 5390-5393; Voellm~ et al. (1981) ~ell 23: 261-270. A number of the genes have been se~uenced. See ~arch and Tbrok (1980) Nucleic Acid Res. 8: 3105-3123, and Ingola and Craig (1982~ PN~S 79: 2 50-2364.
All eukaryoti~ organisms appear to possess heat shock genes.
See Relly and Schlesinger (1978) Cell 15: 1277-1288. Many of the heat shock genes appear to be conserved throughout widely diverse species, and Drosq~hila heat shock genes have been shown to be transcribed in mouse cells (Corces et al. (1981) PNAS 78: 703~-7042), frog cells ~Voellmy and ~ungger (1982) PNAS 79: 1776-1780), and monkey cells (Pelham (1982) Cell 30: 517-528). Fusion genes consisting vf Droso-phila heat-shock gene regions and ~erpes Simplex virus thymidine kina~
se gene regions are also transcribed in these heterologous cell sys-tems. No evidence has been presented which would su~gest that the 1 3 1 2~2~

protein products of these genes are formedO Pelham, H. and Bienz, M. (1982) p. 43-48 in Heat Shock from Bacteria to Man.
~d. Schlesinger, Ashburner and Tissières. Cold Spring Harbour Press. Corces et al. tl982) p. 27-34 in Heat Shock from Bacteria to Man. Ed. Schlesinger, Ashburner and Tissières. Cold Spring Harbour Press.

DNA constructions are provided comprising control elements derived from heat-shock genes associated with genes of interest, which permit the expression of the gene of interest in both procaryotic and eucaryotic cells. In particular, for the efficient synthesis of fully competent gene products, a homologous combination of expression unit and host cell for expression is used and the nature of this homologous system is that of the same or similar cell type or organism that normally produces the said gene product.

More particularly, in one aspect, the invention provides a host cell transformed with a gene expression unit, comprising: (a) an expression control region of about 650 bp from a eucaryotic HSP70-gene carrying promoters, operators, activators, cap signals, ribosomal binding signals and leader sequences from an eucaryotic heat-shock protein gene; and (b) a structural gene of interest under the transcriptional and translational control of said expression control region (a);
and wherein the host cell comprises cells which are either from bacterial origin or from at least one species similar to those from which the said expression control region was isolated and in which control region (a) linked to the said structural gene (b) of interest is present in order to produce the product of said structural yene.

In preferred embodiments of this aspect, the invention provides:

1 3 1 2û27 - 4~ -The above host cell, wherein the structural gene of interest is inserted into the gene expression control region.

The above host cell, wherein the heat-shock gene control region is derived from the genomic DNA of Drosophila melanogaster; and wherein the heat-shock gene control region is derived from any one of the genes encoding the 70 kilodalton heat-shock proteins of Drosophila melanogaster.

The above host cell, wherein at least part of the functional heat-shock gene control region is of synthetic origin.

The above host cell, in which said expression control region ~urther comprises at least one selectable marker linked to the expression control xegion which allows for selection of transformed hosts.

The above host cell, in which said expression control region further comprises a procaryotic replication system which allows propagation of the expression control region, and at least one antibiotic resistance gene.

The above host cell, wherein said expression control region is linked to a fragment containing an eucaryotic extrachromosomal replication system.

The above host cell, wherein the expression control region is linked ko a cellular or viral transcription enhancer element to allow for constitutive expression of the gene of interest.

1 ~ 1 21~

~ 4b -The above host cell, wherein a heat-shock promoter element is linked to the complete RNA coding region of the structural gene of interest.

The above host cell, wherein an eucaryotic heat-shock gene promoter-RNA leader segment is linked to the protein-coding region of the structural gene of interest.

The above host cell, wherein an eucaryotic heat-shock gene control element is linked to at least part of the DNA-coding region of the structural gene of interest.

A method for preparing the above host cell, comprising introducing the expression unit into the host cells by cotransformation with a selectable marker.

The above method, wherein the expression unit is introduced into the host cells with an amplifiable gene.

The above method, wherain the expression unit is introduced into the host cells by transformation or transfection.

In a further aspect, the invention provides plasmids pRV15 and P~8~.

In a still further aspect, the invention provides an expression vector comprising: an eucaryotic extrachromosomal replication system; an expression control region of about 650 bp from an eucaryotic HSP70-yene carrying promoters, operators, activators, cap signals, ribosomal binding signals and leader sequences from an eucaryotic heat-shock protein gene; and at least one inserted protein encodiny sequence '' ' "

1312~7 -- ~Lc --gene under the transcriptional control of the expression control region.

In preferred embodiments of this aspect, the invention provides:

The above~ expression vector, said vector being free of a translational start codon between the expression control region and the insertion sit~.

The above expression vector, having a translational start codon upstream from the insertion site.

The above expression vector, wherein the heat-shock gene control region is derived from the genomic DNA of Drosophila melanogaster; and wherein the heat-shock gene control region is derived from the heat-shock gene which encodes the 70 kilodalton heat-shock protein of Drosophila melanogaster.

The above e~pression vector, further comprising a prokaryotic replication system which allows stable maintenance in prokaryo-tic hosts~

The above expression vector, further comprising at least one selectable marker which allows for selection of transformed hosts.

The invention also provides an expression vector comprising: an eukaryotic extrachromosomal replication system; a heat-shock gene control region substantially homologous to a 650 base pair se~uence from an eucaryotic HSP70-gene carrying promoters, operators, activators, cap signals, ribosomal binding signals and leader sequences which controls transcription of the 70 kilodalton heat-shock - 4d -protein in Drosophila melanoqaster; and at least one inserted protein encoding sequence gene under the transcriptional control o~ the heat-shock gene control region.

Further, the invention provides a DNA construct o* less than 15 kbp which includes an expression control region of about 650 bp from an eucaryotic ~ISP70-gene carrying promoters, operators, activators, cap signals, ribosomal binding signals and leader sequences from an eukaryotic heat-shock gene and at least one inserted protein encoding sequence gene under the transcriptional control of the heat-shock gene control region.

Still further the invention provides a method for producing competent gene products, said method comprising:
joining a structural gene to an expression control region of about 650 bp from an eucaryotic HSP70-gene carrying promoters, operators, activators, cap signals, ribosomal binding signals and leader sequences from an eukaryotic heat-shock gane; introducing said structural gene and control region into a eukaryotic host cell suitable for expression of the structural gene; and growing said host, whereby said product is produced.

In preferred emhodiments, the invention provides:

Tha above method, wherein the structural gene and control region are joined together in an expression vector having an eukaryotic extrachromosomal replication system.

The above method, wherein the structural gene and control region are introduced to the host by co-transformation with a selectable marker.

1~12~27 - 4e -The ahove method, wherein the expression control region is derived from a heat-shock gene which encodes a 70 kilodalton heat-shock protein in Drosophila melanogaster.

The above method, wherein the heat-shock gene is derived from a different eukaryotic species from the host.

The above method, wherein the structural gene is a mammalian gene.

The above method, wherein the host is a mammalian cell culture.

The above method, wherein the structural gene is a human gene.

The above method, wherein the host is a mammalian cell culture.

The above method, wherein ths host is a COS I cell culture.

Fig. la is a schematic representation of a partial restriction map of part of plasmid 132E3 containing two 70 kdal heat-shock protein (hsp) genes (heavy lines) and their orientation (arrows). The lower part of Fig. la represents the relevant partial restriction map of plasmid 51 for comparison purposes.

Fig. lb represents on an enlarged scale a more detailed restriction map of part of plasmid 51 with additional restriction sites, including the two characteristic Sau3A
sites.

- 4f -Fig. 2 represents the complete DNA sequence of one strand of the DNA preceding the 70 kdal hsp gene in plasmid 51 except for the white sequences on a black background, which are present in another 70 kdal gene sequence, as determined by the procedure of Maxam and Gilbert (19773, PNAS
74: 560), and in addition some important restriction sequences as well as the site of the start of transcription and translation (arrows).

Figs 3a-3d represent a diagrammatic representation of the procedure for the construction of plasmid pRV15.

In particular, Fig. 3a shows a restriction map of plasmid 51 _ 5 _ 1 31 2~27 which after digestion with Sau3A gives rise to an approximately 650 bp fragment referred to henceforth as the 650 bp fragment, one end of which is indicated in detail in Fig. 3c (left part).
Fig. 3b is a restriction map of plasmid p~C1403 which, after restriction with BamHI, leaves one end with the structure indicat-ed in detail in FigO 3c (right part).
Fig. 3c represents details of the sequence of the ends produced by the restriction digests of plasmids 51 and pMC1403.
Fig. 3d represents the structure of plasmid p~V15 showing the linkage of the Sau3A fragment of plasmid 51 in the BamHI site of plas-mid pMC1403. The position of the ~ -galactosidase gene and of the Dro, scphila DNA are indicated in the plasmid by heavy lines. S~me res-triction sites are also indicated.
Fig. 4a is a photograph showing characterization of the struc-ture of plasmid p~V15 indicated in Fig. 3d by the colony hybridiza-tion assay of Grunstein and ~ogness (1975) PNAS 72. 3961~ This was perform2d using radioactive probes prepared by nick translation (Ma-niatis et al~, (1975) PN~S 72: 1184) of either a 2 kbp XbaI fragment excised from plasmid 132E3 (see Fig. la) or a labelled 650 bp Sau3A
fragment from plasmid 51 (~ig. 4b)o A number of transformants pro-duced using the construction scheme of Fig. 3 were tested as were some colonies of FMC1403 as a control; these colonies are indicat-ed by arrcws in Fig. 4a.
Fig. 4b represents a similar colony hybridization assay but using the Sau3A 650 kbp fragment from plasmid 51 as a probe. Control colo-nies of pMC1403 are indicated by arrcws.
Fig. 5 represents the characterization, by restriction analy-sis, of the structure for plasmid pRV15 indicated in Fig. 3d. Frag-ment sized in bps are indicated. In particular:
Fig. 5a is a photograph showing the migration of various D~
fragments under electrophoresis. DN~ was prepared from mini-plasmid preparations (~avis et al., (1980), Methods in Enzymology, Grossmann and Moldave, eds. 65: 404-414) for plasmid pMK1403, pRV15 and two similar isolates, p~V25 and pRV5. These plasmids were digested with restriction enzymes as indicated below and the fragments of DNA pro-duced were electrophoresed on 0.85~ agarose gels at 150 v for 4 hours after which band were visualized using ethidium bromide staining `` I ~ 1 2027 and U.V. photography. Identi~ication of ~he lanes is as follows:

Lane 1 pMC1403 - Sal~A/XhoI
2 ~ SalI/XbaI

3 " " - S I/EcoRI

4 pF~25 - SalI/XhoI
" " - S lI/XbaI
6 " " - S lI/EcoFI
7 pRV15 - SalI/XhoI
8 " " - SalI/XbaI
9 " " - S ll/EcoRI
pRV5 - SalI/XhoI
11 " " - SalI/XbaI
12 " " - SalI/EcoRI

Fig. 5b is similar to Fig. 5a for a saries of digests and analy-ses performed identically. m e lanes are as follcws:

lane 1 p~V15 - XhoI
2 " " - _ I
3 " " - SalI/XhoI
4 " " ~ lI/XbaI
" " - _ I
6 pRV5 - XhoI

8 " " - SalI/XhoI
9 " " - 5 I/XbaI
" " - _ I

Fig~ 5c is similar to Figs 5a and 5b and concerns restriction digests and analyses performed for fragments separated on gels of 5% polyacrylamide. Electrophoresis was at 40 v for 14 hours.

lane 1 51 - Sau3A/X
2 51 Sau3A/XhoI

4 pMC1403 - Sau3A

p~N15 - Sau3A/XbaI

6 pRV15 _ 3A~XhoI

7 pRV15 ~ Sau3A

8 Diagrammatic representation of 51 - Sau3A

9 Diagrammatic representation of pBR322 ~ Sau~

Figs 6a-6b refer to an analysis of the orientation of the Droso-. _ .
phila 70 kdal hsp gene ~au3A fragment in plasmid p~V15. In particular-Fig. 6a is a diagrammatic representation of the sizes of pre-dicted restriction fragments that would result if the desired orien-tation of oscphila elements and the E. coll Q -galactosidase gene is indeed the case in plasmid pRV15.
Fig. 6b is a photograph of an acrylamide gel (5%) analysis of the results of restriction of plaæmid p~N15 and pBR322 to test the prediction presented in the diagram of Fig. 6a. Anal~ses were per-formed as describ2d for the cases represented by Fig. 5c.
lane 1 pBR322 - ~infI
pR~715 - i~;cbRI
3 pR~15 - XbaI/E ~RI
4 pfi~q5 - XhoI/Eco~I
pRV15 - XhoI

m~ positicn of bands of sizes 517, 220 and 154 bps are indi-cated by arrows and are derived from the kncwn sizes of fragments deriv~d b~ restriction of plasmid pBR322 by ~mfI.
Figs 7a-7e, which show plasmids or fragments of plasmids after digestion with restriction enzymes, represent schematicall~ the cons-tructlon of plasmld 520, which consists of incorporating the SmaI
- SalI frayment of pRV15 containing the .lac operon fragment and the Droscehila 650 bp control ele~ent into plasmid ~SVod using the B~nHI
- SalI restriction site. In particular:
Fig. 7a represents pRV15, Fig. 7c represents p~Vod; Fig. 7b repre-sents a SmaI - SalI fragment of pRV15; Fig. 7d represents pSVod after excision of a BamHI - SalI fragment; and Fig. 7e represents p520.
Fig. 8a is a restriction map for plasmid vector 520 indicating the position of selected restriction sites in the structure of the 1~20~7 ~ 8 --plasmid.
Fig. 8b is a photograph of an agarose gel analysis of restric-tion enzyme digestions of plasmid 520 performed as described in the case of pRV15 (see Fig. _).
lane 1 - HindIII
2 - XhoI/EcoRI
3 - ~indIII/E oRI
4 - PstI
5 - HindIII/PstI
6 - PstI/hindIII/E cRI
7 - P tI/EcoRI
Fig. 8c is a photograph also referring to the characterization of plasmid 520. It shcws agarose gel analysis of various restriction digest of plasmids pSVod, p~V15 and 520-like plasmids, i.e., plas-mids 639, 520, 519, X, Y, Z.
lane 1 639 - S lI/EcoRI
2 520 - " "

4 S16 - " "
5 X - " "
6 y . .... Il 7 z _ .... "
8 pSVod - 11 "
9 pRV15 - " "
p6Vcd --XhoI/EccR
11 p3~V15 ~
Note that the digests in lanes 8 and 9 are inco~plete.
Fig. 8d is a photograph showing further restriction digests of plasmid 520 by ccnparison with digests of pLasmids pSVod and pRN15.
Lane 1 p6Vod - PstI/EcoRl 2 pSVod - HindIII/SalI
3 pRN15 - PstI/EcoRI
4 pRV15 - HindIII/SalI
pRV15 - SalI~XbaI
6 520 - P I/EcoRI
7 520 - HmdIII/SalI
~ 520 - SalI/ _ I

g Nbte that the digests in lanes 2 and 7 are incomplete.
Figs 9a-9c refer to the construction of plasmid PR81. In parti-cular:
Fig. 9a represents the Mbunt Sinai A/PR 8/ms/4.76 pLasmid from which a fragment (ref. no. 20) was excised with BamHI and H mdIII;
Fig. 9b represents pSVod less a HindIII - SalI fragment; and Fig.
9c represents plasmid PR~l, i.e., the ligation product of the BamHI
- HLndIII fragment 20 with a SalI - HindIII digest.
Figs 9d and 9e are photographs of a Grunstein co~ony hybridiza-tion analysis of PR81-like plasmids, per~ormed as described for pRW15 (see ~ig. 4). m e nick-translated probe used was the BanHI/HindIII
fragment for A/PR 8/ms/4.76. As a control some p6Vod containing colo-nies are indicated by the arrows.
Fig. 9f is a photograph of a restriction analysis of PRBl-like plasmids sho~n by calony hybridization ~lalysis in Figs 9d and 9e.

lane 1 filter 21, Sample 30 - EccRI
2 " 21, Sample 44 - " "
3 " 22, Sa~ple 11 - " "
4 " 22, Sample 24 - " "
5 " 23, Sample 10 - " "
6 " 23, 5amp1e 31 - " "
7 " 24, Sampl~
8 p6Vod - " "
9 P~ 8/ms/4.76 ll ll 10 filter 21, Sample 30 - HindIII/SalI

11 " 21, Sample 44 -12 " 22y Sample 11 - " "

13 " 22, Sample 24 - " "

14 " 23, Sample 10 - " "

15 " 23, Sample 31 - " "

16 " 24, Sample 8 - " "

17 pSVod - " "

18 PR 8/ms/4.76 - " "

Figs lOa-lOe refer to the construction scheme of plasmid PR84.
m is plasmid was constructed as shown in the figure and consists of - lO - 1 3 1 2027 substituting the haemagglutinin (HA) gene 20 present in plasmid PR81 for the~-galactosidase gene in plasmid 520. In parti~ular:
Fig. lOa represents plasmid 520; Fig. lOc represents pla3mid PR81; Fig. lOb represen~s an NcoI - PvuII fragment excised from 520;
Fig. lOd represents PR81 after digestion with HindIII and NcoI; Fig.
lOe represents schematically PR84.
Fig. lOf is a more detailed representation of PRB4 with one strand of base-pairs including the sequence originating from PR 8/ms;
only the crucial joining region between the Drosophila HS control element and the HA start sequence is detailed.
Fig. 11 refers to checking the results from the construction of PR84. Fig. lla is a photograph referring to the restriction diges-tion of plasmids PR81, PR 8/ms/34, pSVod and 520.

lane 1 PR81 - HindIII/EcoRI
2 PR81 - XhoI/E cRX
3 P~81 - NcoI
4 PR81 - N I/XhoI
PR 8/ms/34 - ~IindIII/BcoRI
6 PR 8/ms/34 - ~hoI/EcoRx 7 PR 8/ms/34 - N~oI
8 PR 8/ms/34 - NcoI/XhoI
9 pSVod - NcoI/SalI
10 p6Vod - PvuII/SalI
11 E~;Vod - HindIII/E cRI
12 pSVod - HinfI STANnARD
13 520 - PvuII
14 520 - P I~/XbaI
15 520 - PvuII/XhoI
16 520 - P II/EcoRI
17 520 - PvuII/NcoI
18 520 - NcoI

19 520 - NcoI/XbaI

20 520 - NcoI/XhoI

: Figs llb(l) and llb(2) refer to a Grunstein colony hybridiza-tion analysis of PR 84-like plasmids. A nick translated radioactive probe (approximately 108 cpm/~g) of the plasmid 51 Sau 3A/Xho I Dro-scphila control element fragment was used in these assays~
Fig. llc refers to a 5% polyacrylamide yel analysis of the XhoI
digestion products of PR84-like plasmids selected from the positive colonies identified in Fig. llb.

lane 1 filter 3, Sample 12 ~ _ I
2 " 3, " 15 (PR~2) - "
3 " 3, " 22 " - 1' 4 " 3, " 42 (PR84) - "
5 " 3, "45 " - "
6 " 4, "19 " - "
7 " ~, "40 " - "
8 " 5, "16 " - "
9 " 5, "28 " _ 1.
10 " 5, "39 " - "
11 pSVod " - ~infI STANDARD

Fig. lld refers to further restriction analysis of plasmids giv--ing the XhoI digestion pattern shown in lanes 2, 4, 6 and 8 of Fig.
llc. Analysis was performed as de~crib~d for Fig. llc.

lane 1 ~ X 174 RF - HaeIII STANDARD
2 filter 5, Sample 16 - aI~NcoI
3 " 5, " 16 - XhoI/~
4 " 4, " 19 - XbaI/NcoI
" 4, " 19 - XhoI/NcoI
6 " 3, " 42 - X aI/hcoI
7 " 3, " 42 - XhoI/ _ I
8 " 3, " 15 - XbaI/~coI
9 " 3, " 15 - X I/NcoI

Fig. 12 describes the characterization of plasmid F629~
Fig. 12a refers to a 5% polyacrylamide gel analysis of scme canr didate 629 series plasmids.
lane 1 629/8 BglII/E oRI

3 629~6 "
4 62g/4 "
DN~ s~andard Fig. 12b refers to another 5% polyacrylamide gel analysis of all eight candidate 629 series plasmids.
Pla lane 1 629/1 XhoI
2 629/2 n 3 629/3 n 4 629/4 "
629/5 n Fig. 12c shows a Grunstein colony hybridiæation assay using the nick translated Xho~ II fragment frcm plas~id p622b Fig. 12d shows a Grunstein colony hybridization assay using the nick translated HindIII~BamHI~A gene fragment frcm plasmid A/PR
8/mx/3~.

Mbthods and cc~positions are prcvided ~or ~he con~rolled expres-sion of a gene in a eukaryotic host. The present invention is par-tic~larly useful for expressing mammQlian genes in mamnalian hosts, ~ypically in a host cell culture si~ilar or identical to the host in which the gene is naturally expressed. Xn this way~ post-trans-criptional and post-translational mcdifications of the gene products will be achieved~ thus assuring the co~petence of such resulting gene products.
DN~ seguences which comprise an inducible expression control region derived from a eukaryotic heat-shock protein (hsp) gene are employed. Such control regions are inducible for example by elevat-ed temperatures and include sequences responsible for both transcriE-, ~.L., . ..

tional and translational control of the hsp gene activity, e.g., the promoter region (RNA polymerase reco~nition and binding sites), and possibly also heat-shock protein gene control se~uences which have been modified in such a way that they allow the constitutive expres-sion of downstream coding se~uences. The c~ntrol region may also in clude other portions of the translated and untranslated regions of the hsp gene which participate in transcriptional and/or translatio-nal regulation including 3'untranslated sequences~ F~lrthermoxe, sequen-ces derived from the 3' untranslated region flanking the structural gene (i.e. theaoncoding portion of the gene of interest) may be in-cluded. abnveniently, the entire hsp gene including the structural region as well as the 5l and 3' flanking regions may be employed.
In the latter case, the gene of interest may be inserted internal to the structural region to produce a fused protein upon translation.
m e protein of interest may then be recovered by conventional means.
The term heat-shock protein (hsp) gene will be enployed in the specification and the claims to denote the entire gene locus respon-sible for the inducible expressicn of heat-shcck proteins, specifi-cally including the 5' regulatory se~uences upstream of the struc-tural region, ~he structural region which encodes the mRNA transcr~pt (includiny untranslated leader sequences) and the 3~ flanking region dcwnstream of the structural gene.
Heat-shock protein genes may be obtained frQm most higher eukaryo~
tic organismsO Particular heat-shock proteins have been identified in fruit flies (Drosce_ila), mice, fro~s, monkeys, and man. Heat-shock genes suitable as a source for the control regions of the present invention may be obtained from any of these or other eukaryotic orga-nisms. Although heat-shcck genes derived from one species can be ex~
pressed in other species and even in procaryotes, it is generally desirable that in order to optimize the expression advantages of heat-shock control elements at both the transcriptional and the transla-tional levels, these elements should be derived from the same or simi-lar organism or cell type that will be used for expression of the gene of comnercial interest. For instance, constructions including genes under control of Droso~hlla heat-shock elements can be most advantageously expressed in Drosophila cells in culture.
In the Experimental section hereinafter, an inducible expres-sion control region obtained from a heat-~hock protein gene which encodes for a 70 kdal heat-shcck protein found in ~ was used to express E. coli ~-galactosidase gene in E~ coli, CQS I Monkey cells, and in Xenopus occytes. The same region was used to control the expression of the influenza haemagglutinin gene in CCS I cells.
The inducible expression control region of the present inven-tion may be combined with an extrachromosomal replication system for a predetermined host to provide an expression vector for that host.
Such vectors will include DN~ sequences having restriction site(s) for insertion of gene(s) 3' to said control regions to provide the regulated transcription and translation of the inserted genes. m e vector can also include markers for selection in bacteria and in eu-karyotic host cells, a prokaryotic replication system allowing clon-ing of the vector in a prokaryotic host, and other DNA regions of interest.
Alternatively~ the expression control region can be joined to a desired structural gene and the resulting DNA constructs introduced directly into the host cells. Methods for such direct transfer in-clude injection of the DNA into nuclei ~Ca~pechi (1980) cell 220 479 - 4~8) and co-transformation by calcium phosphate precipitation ~Wi-gler et al. (1979) Cell 16 777~785) or nEAE dextran (McCutchan, J.H.
and Pagano J.S. (1968) J. Nat. Can~er Inst. 41: 351-357).
As stated above, the DN~ constructs of the present invention ma~ include differing portions of ~he hsp gene. m e constructs will include at least the 5' regulatory region which carries the promo-ter, regulatory sequences such as operators, activators, cap signals, signals enhancing riboscmal binding, and other sequences as well as additional DN~ leader se~uences responsible for transcriptional and translational control. The DNA constructs may also include part of the protein coding sequence of the hsp gene, resulting in the pro~
duction of fused proteins when the foreign gene to be expressed is inserted downstream of the hsp sequences. If a fused protein is not desired, it will be necessary to remove the hsp coding sequences from the isolated hsp D~A by usual methods such as restriction by enzy-mes and exonuclease digestion.
It may be desirable to leave at least a portion of the hsp struc-tural gene dcwnstream from the natural translational initiation codon.

- 15 - 1312~27 In this way, a fused pro~ein including the amino-terminal amino acid sequences of the heat-shock protein is provided~ When such fused prcr teins are produced, it may be desirable to introduce selective clea-vage sites so tha~ the desired protein can be separated from the pre-cursor protein. See U.S. Patent No. 4,366,246 to Riggs which teaches how such cleavage sites may be introduoed.
m e expression control region of the present invention will usual-ly be co~bined with a terminator for complete transcriptional con-trol of the inserted structural gene. aonveniently, the terminator can be derived from the heat-shock gene itselft although the inserted structural gene may carry its own or any other suitable terminator se~uence.
Extrachromosomal replication systems may also be used. Suitable replication systens include autonomously replicating sequences as described by Struhl et al. ~1979) PNAS 73: 1471-1475 and the 2 ~um plasmid for replication in yeast. M~mmalian replication sys~ems could be derived frcm pa~ovaviruses, such as simian virus 40 and bovine papilloma virus; adenoviruses; avian retroviruses, such as avian sar-coma virus; and mammalian retroviruses such as l~oloney leuke~ia virus.
In addition to the optional eukaryotic replication system, it is advantageous to provide a prokaryotic replication system to al-low for cloning of ~he vector in a bacterial host. This allcw~ large quantities d the vector to be grown in well characterized bacterial systens prior to transfsrming a eukaryotic host. Suitable prokaryo-tic replication systems are well known and include plasmids such as pBR322, pRK2~0, ColE1, and bactericphages, e.g. i~dv. The prokaryo-tic replication systems will necessarily include an origin of repli-cation recognizable by a prokaryotic host, and will usually include one or more markers for the selection of transformants in the pro-karyotic host. Such markers include biocide resistance, toxin resis-tance, and the like. Alternatively, co~plementation allowing the growth of an auxotrophic host in a selective medi~ may be employ-ed. Such techniques are well known in the art and need not be des-cribed further.
Usually, the markers employed will be different for selection in prokaryotic and eukaryotic hosts. Various dominan~ly acting mar-kers are useful in selecting for transformed mammalian cell lines.

- 16 - ~312027 They usually comprise a specific gene whose expression confers a new drug-resistant phenotype to the mammalian cells in an appropriate selective medium. Specific markers include the bacterial xanthine-gua-nine phosphoribosyl transferase gene which can be selected in medium containing mvcophenolic acid and xanthine (Mulligan et al. (1981) PNAS 78: 2072-2076~; transformants with vectors carryin~ a mouse cDNA
fragment coding for dihydrofolate reductase may be selected for using medium containing aminopterin (Subramani et al. (1981) Mol. Cell Biol.
1: 854-861); and a bacterial plasmid gene specifying an amino-gly-coside phosphotransferase that inactivates the anti-bacterial action of neomycin-kanamycin derivatives may ~e selected for using medium containing G418 a neo~ucin derivative toxic for most mammElian cell lines (Colbere-Garapin et al. (1981) JO Mol. Biol. 150: 1-14).
It is evident that the number of copies of gene expression units introduced into a host cell may vary and that by proximal combina-tion of one of the amplifiable genes such as the mouse dihydrofolate reductase gene the number of copies of integrated gene expression units may likewise be amplified by induction of the amplification of the asscciated amplifiable gene and the proximal DN~. See for exam-ple the co-~mplification of dihydrofolate reductase c~NA and the E. coli XGPRT (xanthine-guanine phorphoribosyl trans~erase) g~ne in Chinese ovary cells (Rungold, G. et al., (1981). J. Mol. and Appl.
Genetics, 1, 165-175).
I~ the exemplary method for preparing the vectors of the sub-ject invention, both the expression control region of the heat-shock protein gene and the eukaryotic replication system are inserted in-to a suitable prokaryotic plasmid. The manner and order of the in-sertion are not critical, and it is necessary only that the result-ing vector retains viable replication systems for prokaryotic and when necessary eukaryotic hosts.
m e D~ constructs containing Drosoe~ heat-shock gene con-trol segments functionally linked to a structural gene of interest can be used to synthesize products of said gene in Drosophila cells or cultured cells closely related to Drosophila. Since the structu-ral gene is under the transcription and if necessary translation con~
trol of the heat-shcck control element, its expression can be enhanc-ed by for example increasing the ambient te~perature of the cells ~ ~7 - ~312~27 for instance in the ran~e 37-42C. By such induction a large frac-tion of the newly made mRN~ will be deri~ed from said structural gene.
Now if the struc~ural gene is a human gene encoding a protein r~uir-ing complex processing, the expression vector preferably conprises, in addition to said structural gene, replication elements and mar-ker genes, a heat-shock control element derived from a human heat-shock gene. This construct will be introduced advantageously into cultured human cells by procedures described above to produce pro-ducts of sai~ h~man gene. As necessary the choice of recipient human cells will be made on the basis of their competence in correctly ex-pressing the fully processed gene product.
A wide variety of structural genes may be introduced into the subject vectors to permit the production of various gene products including polypeptides, such as enzymes, proteins, hormones, novel protein structures and the like.

--E~er-~e~a~

The following exanples are offered by way of illustration~

1. The construction of plasmid ~RV15 which contains the E. coli ~-galactosldase ~ene under the control of a Drosc~hila heat shock control element.
-A 650 base pair (bp) DN~ fragment from one of the two ~ 3~70 kdal heat-shock protein genes, containing a heat-shock gene trans cription contr~l element, a complete RNA leader sequence, a trans-lation initiation siynal and a sequence coding for the first few ami-no acids oE the Drosophila 70 kdal heat-shock protein, was obtain-ed from a sub-clone of part of plasmid 132E3 (Schedl et al. (1978) Cell 14: 921-929; see Fig. 1). Plasmid 132E3 contains two complete genes for the 70 kdal heat-shock protein, and the aforementioned iso~
lated sub,clone, plasmid 51 (Karch et al. (1981) J. Mol Biol. 148:
219-230), contains a fragment of the first heat-shock gene in plas-mid 132$3. This fragment was isolated by digestion of plasmid p51 with the restriction endonucleases ~ II and BanEII.
m e upper part oE Fig. la is a representa~ion of a portion of ~ 18 - 1312027 plasmid 132E3 indicating the two genes encoding 70 kdal heat-shock proteins (thick line segments) a~d some of the identified restric-tion sites. m e lower part of Fig. la represen~s a portion of p61 containing the ~ Bam~I segnent containing the aforesaid 650 bp fragment, bounded by the Sau3A cleavage sites.
Fig. lb is a more detailed representation of the portion of in-terest contained in p51 with additional restriction sites indicat-ed, and al.so showing the position of the 650 bp fragment between two Sau3A sites.
Fig. 2 provides the DN~ sequence data for one strand contained in the aforementioned 650 base pair fragmentO The limits of this se-quence are indicated by the position of the Sau3A restriction sites (GATC). Also, XbaI, XhoI and PstI recognition sites for restriction are indicated, and the position of the transcription and translation start sequence are indicated by arrows la and lb, respectively. m e 650 bp frasment was functionally inserted into plasmid pMC1403 in a position to control the expression o test genes in this plasmid.
Plasmid FMC1403 ~Casadaban et al. (1980) J. Bacteriol. 143: 971-980) is a derivative of pla~mid pBR322 (Boli~ar et alO (1977) Gene 2: 95 - 113) containing the entire E. coli lac operon with the exception of the sequences coding for the first seven amino acids of ~ -galac-tosidase and all sequences 5' to the ~-galactosidase protein coding region (promoter, ribosomal binding site, translation initiation co-don~. Consequently lac strains of E._col such as MC 1061 (Casada-ban et al. (1980) J. M~l. Biol 138: 179-207) carrying plasmid pMK1403 do not produce ~ -galactosidase. A polylinker (EccRI, 5maI, BamHI) at the 5' end of the lac sequences in EMC1403 permits the introduc-tion of foreign DN~ s~quences upstream from the ~ -galactosidase cod-ing region. Insertion of a segment containing a functional promoter and the RN~ leader se~uences will result in the production of~ -galac-tosidase activity. It should be noted that the amino-terminal end of ~-galactosidase is not essential for its enzymatic activit~ (Mul-ler-Hill B. and Kania, J. (1974) Nature 249: 561-563). A11 that is therefore re~uired is that the inserted promoter/RNA leader sequen-ce permits reading in the correct frame with the incomplete ~-galac-tosidase coding region. Bence the 650 bp segment was ligated into the BamHI site of plasmid pMC1403, in front of the incomplete ~-galac--19 131:20.~

tosidase gene. ~he new recombinant plasmid thus obtained (see Fig.
3) was designated p~N15. The procedure of this plasmid construction is shcwn in a diagrammatic form in Fig. 3. In this figure, ~he up-per left part (Fig. 3a) represents a section of the aforementioned plasmid 51 con~aining the 650 bp fragment 2 to be cleaved out with Sau3A and to be inserted into the Ban~I site of plasmid pMC1403 shown ~n Fig. 3b. m e truncated lac cperon of the plasmid p~C1403 is repr~-sented by the heavy line with numeral 3 on the diagram of Fig. 3b.
It also comprises a structural gene for ampicillin resistance~ Plas-mid p51 is digested with Sau3A to excise the aEoresaid 650 bp frag-ment represented on the left of Fig~ 3c by numeral 2; this fragment was purified from polyacrylamide gels (see Fig. 5c) by electroelu-tion. This fragment was inserted into the BamHI site of pMC1403 in either orientation. The ligation mixture was then used to transform the lac strain of E. coli MC 1061. Transformants were plated on media containing ampicillin and Xgal (Xgal is 5-bromc,4-chlorot3-indolyl-N-D-galactoside) which is a subs-trate for ~ -galactosidase, and which after cleavage by the enzyme produces an identifiable colored product (see Miller, J. ExperLments in Molecular Genetics, pp. 47-55, Cold Spring Harbor, ~.Y~ 1972).
Plasmid pMC1403 did not produce ~ -galactosidase ac~ivi~y in ~his as-say while plasmid pRV15 and a large numker of other transformants, prepared in the aforesaid manner, were fo~nd to produce substantial an~unts of ~ -galactosidase as determined by ~he color change prcduced on Xgal plates.
In addition, ~ -galactosidase-coding plasmid constructions such as pRV15 were analyzed further by the colony hybridization assay of Grunstein (Grunstein and Hogness (1975) PNAS 72: 3961-3966) using either radioackively labelled 650 bp 5au3A fragments from p51 (see Fig. 4b) or the 2 kbp XbaI gene fragnent (see Figs la, 4a) from plas-mid 132E3 as hybridization probes. These fragments were radioacti-vely labelled by the process oE nick translation (Maniatis et al.
(1975) PNAS 72: 1184). As seen in Figs 4a and 4b, all selected trans-formants hybridized to both radioactive probe DNAs demonstrating the presence of both DN~ sequences in reccmbinant plasmids such as pRNq5.
The presence of the 650 bp fragment containing the Drosophila 70 kdal heat-shock gene control element was confirmed by restriction - 20 - 1 3 1 ~27 analysis, see Figs 5a, b and c. Plasmid pMC14U3 contain5 unique res-triction sites for EccRI and SalI, but no XhoI or XbaI sites. The Sau3A 650 bp fragment from plasmid p51 however contains unique sites for restriction by XhoI and XbaI but has no sites for restriction enzymes such as EcQRI and Sall. In contrast, reco~binant plasmids such as p~V15 should contain unique sites for all four aforemention-ed enzymes (see Fig. 3). Restriction of recombinant DNA from plas-mids such as p~Vq5 with XhoI and SalI or with XbaI and SalI should produce two fragments of about 4 and 6.5-7 kbp if the structure shown for pRVlS in Fig. 3 is correct. Plasmid EMC1403 hcwe~er should only be linearized. FigO 5a shows a photograph of DN~ frayments produced by the aforementioned restriction enzymes after electrqphoresis on agarose gels, and confirms the structure of plasmid pRV15 shown in Fig 3. m ese results are presented again in Fig. 5b where it is also demonstrated that digestion with SalI, XhoI or XbaI individually only l mearize pRV15, further confirming the structure shcwn in Fig. 3.
Further, if pRV15 contains the 650 bp Sau3A fragment of plas-mid p51 inserted into the Bam~I site of pMC1403, then it should be possible to recover the 650 bp Sau3A fragment by digestion of recom-binant DN~s such as pRV15 with the restriction en2yme Sau3A. That this is the case is shown in Fig. 5c lane 7 by the arrows. In addi-tion, the identity of the 650 bp excised fragment is confirm~d by its restriction b~ XbaI or by XhcI. See Fig. lanes 5 and 6 m e demonstration that the Sau3A 650 bp fragment and the p-galac-tosidase coding 3equences are in the orientation shcwn for pRV15 in ~ig. 3 is presented in Figs 6a and b. In Fig. 6a, the correct orien-tation is shown in a diagrammatic form from which the respective double digestions with EccRI and XbiaI or by EcoRI and XhoI would pre dict the excision of fragments from the 650 bp fragment, correctly oriented with respect to the ~ galactosidase gene 12, of respecti-vely 150 and 220 bp. In Fig. 6a, the arrcw 11 indicates the trans-criptional direction of the heat-shock control elements. The above prediction is demonstrated by the restriction gel analysis shown in Fig. _ where EcoRI and XbaI double-digestion libera~e an approxi-mately 150 bp fragment, and EcoRI, XhoI double-digestion, a 220 bp fragment as evidenced by electropheresis on 5% acrylamide gels. These experiments confirm that plasmid p~V15 has the desired orientation of control and coding elenents shown in Fig. 3. m e position of bands of sizes 515, 220 and 154 bp are indicated by arrGws marked with the corresponding numbering.
The hybrid pLasmid pRV15 and other identical isolates ~onstruct-ed as descriked in the preceding section, produce substantial amounts of ~ -galactosidase as determined by examining the color chan~es on Xgal plates. The presence of ~ -galactosidase protein produced in E.
coli under the control of a _roso~h_la heat-shock control element has also been clearly demonstrated by imn~ne precipitation of pro-tein extracts of E. coli containing plasmid pRV15 after radioactive labelling a newly synthesized protein with 35S-methionine. A poly-peptide of molecular weight appro~imately 1~0,000 daltons was preci-pitated frcm such protein extracts by immune sera directed against authentic E. coli ~ galactosi~ase.
m e details of the construction and exp~ession of pRV15 are now set forth.

A. Construction of Pla$mid pRV15 10 ~ g of plasmid p51 (Fig. 1 , lower part) were digested for 4 hours at 37C with 10 units of 5au3A (incubation buffers for this, and all other digestion described belGw as suggested by the supplier of the restri~tion enzymes: N~w England Biolabs Cat. 1982). m e conr c~ntration of nNA during digestion with Sau3A was 40 ~ g/ml. m e diges-tion products were electrcphoresed on a non-denaturing 5~ polyacry-lamide gel in 1 x TBE (10.9 g Txis base, 5.5 g boric acid, 0.93 g Na2EDTA per liter H20) buffer. A Sau3A digest of pBR322 wa~ electrcr phoresed in parallel and served to identify the 650 bp 51 Sau3A pro-moter fragment (fig. 5c) o DNA fragments were visualized with ethi-dium bromide (EtBr). m e region containing the prcmoter fra~ment, indicated by the arrows in Fig. 5c, was cut out of the gel and the fragment was electroeluted into a dialysis bag. Elec~roelution was carried out for 5 hours at 200v in 1 x TBE buffer. The eluate was collected in a 15 ml siliconized Corex tube and ethanol-precipatat-ed overnight at -20C. The precipitate was collected by centrifuga-tion (Sorvall, 30 min., 10,000 rpm, SS34 rotor), dried in a lycphi-lizer and resuspended in 200 ~ 1 of TE buffer (10 mM Tris. HCl, pH

~ 22 - 1 3 1 202-I

7.5, 1 mM Na2EDrA)o The D~ was then extracted twice with TE-satu-rated phenol and twice wi~h ether. Xhe solution was then passed through a Sephadex G75 mini-column (in a Pasteur pipet). DM~ in the column eluate was ethanol-precipitated twice, dried and resuspend-ed in 20 ~ of TE buffer. A 10 ~ portion was incubated with several units of XbaI in the appropriate digestion buffer for 1 hour at 37C.
The partially digested DN~ was electrophoresed on a 5% polyacryla-mide gel. This experiment (Fig. 5c) demonstrated that the isolated segment was the 650 bp Sau3A fragment, since the promoter fragment includes the only XbaI site of plasmid p51.
Cne ~g of plasmid pMC1403 was digested with 4 units oE BamHI
for 30 min. at 37C. m e digested DNA was extracted 3 times with phe-nol and then 3 times with ether. The DN~ was further purified by 2 subsequent ethanol precipitations. m e pelleted DNA was dried and then resuspended in 20 ~1 of TE buffer.
Ten ~1 aliquots oE the solutions containing the 650 bp promo-ter fragment and digested pMfl403 were combined and then incubated overnight at 14C with an excess of T4 D~ ligase in a total volume of 25 ~ (incubation buffer as recommended by New England Biolabs 19~2 Cat.). m e ligation mix~ure was then incuba~ed for several hours with 8 units of BanHI at 14C. Aliquots (1, 3 and 7 ~ ) were then used to transform 0.5 ml aliquots of CaC12-treated E. coli MC 1061.
Aliquots (0.1 ml) of the transformation su~pension were then pla~-ed on LB agar containing 10 pg/ml ampicillin and 40 ~ug/ml of Xgal.
~lue colonies were observed a~ter overnight incubation of the pla-tes at 37C. Approximately 80 ~ ~galactosidase-producing colonies were isolated. The presence of the p51 Sau3A prom~ter fragment in the re-combinants was established by Grunstein colony hybridization (Fig.
4b). m e hybridization probes were prepared as follows: the p51 Sau3A
promoter fragment was isolated as described above. 50 JUg of 13 23 were digested with 50 units of XbaI for 2 hours at 37C. m e diges-tion products were separated on a 0.9% agarose gel. The 2 kbp 70 kdal hea~-shock protein gene fragment was eluted electrophoretically from the gel and was purified as described above. m e latter fra~ment and the p51 Sau3A fragment were then 32P-labelled by nick translation to a specific radioactivity of 2X108 cpm/~g. The two probes were then denatured and hybridized to two nitrocellulose filters containing - ~3 - l 31 2027 DN~ of the 80 selected transformants and of pMC1403 (Figs 4a, b).
Both probes strongly hybridized to DNAs of all 80 recombinants~ sug-gesting that all reconbinan~s contained the 650 bp promoter fragment.
Clones 5, 15, ~5, 35, 45 and 55 were selected ~sr further studies.
Small quantities of DM~ were prepared from each o~ these clones (Davis et al. (1980) Methods in Enzymology, Grossmann and M~ldave, eds. 65:
404-414). m ese DNAs and p~C1403 DN~ were compared further by res-triction digestion and electrophoresis on 0.9% agarcse gels (Figs 5a, b). The results indicate that restriction of the recombinant plas-mids such as p~V15 with XhoI and SalI or with XbaI and SalI produce two fragments of 4 and 6.5-7 kbp, thus confirming the structure of plasmid p~N15 as shown in FigO 3d.

B. Expression of the Drosophila Heat-Shock - E coli ~alactcr sldase Fusion Gene in E. coli Bacteria containing the hybrid plasmid pRV15 and other isolates constructed as described in the preceding section, produce substan-tial amounts of p-galactosidase as determined by examining the color changes on X~al plates. The presence of a ~-galactosidase protein produced in E. coli under the control of a Drosophila heat-shock conr trol element was also demonstrated by immune precipitation (Bromley et al. (1979) J. Virol. 31: 86-93) of protein extracts of E. coli containing plasmid pRV15 after radioactive labelling of newly synthe-sized proteins with 35S-methionine~ A polypeptide of molecular weight about 120,000 daltons was clearly precipitated from such protein ex-tracts by immune sera directed against authentic E. coli ~-galacto-sidase.

C. Expression of a Heat-Shock ~ -galactosidase Hxbrid Gene in Xenoe~s Oocy~es Plasmid pRV15 (50 - 100 ~g) was digested with an excess of SalI
and Ecc~l. The two resulting restriction fragments were then sepa-rated on 0.85% agarose gels. The 7kb fragment containing the hybrid gene but no vec~or sequences was purified by electroelution and gel filtration on Sephadex ~ G75. This fragment was then incubated with an excess of T4 DNA ligase in a total volume of 25Jul as described above, to permit circle formation. The ligated fragmen~s were inject-ed into oocytes as de~cribed by VoeLl~y and Rungger (1982) PN~S 79:
1776-1780. Following a 6-20 hours preincuba~ion at 20C, the occytes were injected a second time with ~-32~-Grp an~i either heat-treated for 2 hours at 37C or kept at 20C for the same period. Total RNA
prepared from these occytes or frcm oocytes ~hat did not contain fo-reign D~ was hybridized to Southern blots of Sal I/Eco Rl digests of plasmid pMC1403. Oocytes that did not contain the hybrid gene did not produce measurable am~unts of RMA complementary to the ~ -galac~
tosidase gene. ~oth heat treated and untreated oocytes containing the p~l5 fragment were found to have made ~-galactosidase ~NA in approximately similar qu~ntities.
In another series of experiments, unlabelled RNA was isolated from oocytes ~ontaininy the hybrid gene which had been either heat-treated for 2 hours or incuba-ted at 20C for the same time in th~
presence or absence of low concentration of ~ -Amanitin (~oellmy and Rhngger (1982) PNAS 79: 1776-1780). The oocyte ~N~s were labelled by reverse transcription as described by Bromley et al. (1979) J.
Virol. 31: 86-93 and hybridized to pMC1403 Southern blots. A~ain -galactosidase transcripts were present in RN~s for heat~treated and untreated oocytes.o~-~manitin did st~p transcription of the hy-brid gene at both temperatures indicating that RN~ polymerase B (i.e.
II) was responsible for all observed transcriptional ac~ivity.

~ . The Construction and use in Eukaryotic C lls of Pla.smld PR84 which contains a Hunan Infl~lenza Virus Haemagglutinin Gene under the Control of a Droscphila Eeat-shock abntrol element A plasmid denoted 520 was constructed from plasmid p~l5 and plasmid pSVod (Mellon et al. (1981) Gell 27: 279-288) which is capable of replicating either in procaryotic cells or in certain eukaryotic cells. Plasmid 520 allows rapid analysis of transcriptional control since it includes the SV40 virus origin of replica~ion and is able to replicate efficien~ly in SV~0 mutant-transformd, transformation antigen positive COS cells (Gluzman (1981) Cell 23: 175-182).

- 25 ~ 1312027 The schematic orm of the construction of plasmid 520 is shown in Fig. 7, Plasmid pRV15 (Fig. 7a, representation in 1 mear form) was digested with SmaI, subsequently digested with SalI and the re-sultiny 7 kbp fragment (Fig. 7c) containing the 650 bp control element 2 and the lac gene fragmen~ 3 originating from EMC1403 present in pR~15 was isolated by electrophoretic elution from a pre-parative agarose gel.
Plasmid pSVcd (Fig. 7c) was digested with BamHI, the cohesive ends were filled in with DNA polymerase Klencw fragment and the DN~
finally digested with SalI. In Fig. 7c, numeral 10 indicates the site of the deletion of the lkb pBR322 se~uence which has been claimed to be inhibitory for replication in eukaryotic cells. The truncat-ed pSVod ~ragment (Fig. 7d) so formed was ligated to the 7 kbp frag-ment isolated from pRV15, and the ligation mixture was used to trans-form E. coli MK1061. m e structure of plasmid 520 is shown in Fig.
7e in which a re detailed representation of the 650 bp segment also appears with the XbaI, XhoI and Ps~I restriction sites. Another res-triction site tPVUII) i5 also indicated and corresponds to the CAGCIG
seguence between 60 and 70 bp, as denoted in Fig. 2. The significance of the PvuII site will be mentioned hereina~ter. Plasmid 520 is shGwn in circularized form in Fig. 8a where the segments a, y and z repre-sent genes of the lac operon. Numeral 11 designates the fragment conr taining the SV40 origin.
Transformants were plated out on Xgal-Ampicillin plates as des-cribed previously, and blue transformants suspec~ed of containing plasmid 520 were isolated. From the structure shown in Fig. 8a, plas-mid 520 should provide the following size fragments when digested with the nam~d restriction enzyme combinations.

Eragment sized in kbp Enzymes O.75/1.1/8.5 PstI/EcoRI
2.6/7.2 ~lindIII/SalI
3.1/6.7 SalI/XbaI

The 520 plasmids indeed produced frayments of the correct sizes when digested as evidenced in ~he gel analyses of Figs 8b, c and d.
O~her fragments produced by various combinations of enzyme digestions serve as controls for the interpretio~ of the digestions noted above~
Plasmid p81 places a gene which encodes for a eukaryotic pro-tein in the correct reading frame with the 650 bp heat-shock control element of p520. As start mg material for this construc~ion, plas-mid A/PR 8/ms/34 (Mbunt Sinai) Clone No. 4~76 containing a human in-fluenza haemagglutinin gene (1775 bp, numeral 20) inserted into the PvuII site of vector PAT 153/Pvu II/8 was obtained frcm Dr. G. Brcwn-lee (University of Oxford, ~epartment of Pathology U.K.). The plas-mid containing the haemagglutinin (HA) gene (insert 20, see Fig. 9) was digested with BamHIr and the ends were filled in with DNA poly-merase Klenow fragment. Following digestion with HindIII, the haemag-glutinin gene fragment was purified on agarose gelsO The insert 20 contains a conplete HA protein coding sequence, an RNA leader seg-ment, and 40 bp of 3' nontranslated sequence. The HA gene-contain-ing fragment was inserted into plasmid vector pSVod to produce plas-mid PR81 as follows: pSVod DWA was digested with SalI, ends were fil-led in as described above, and the DN~ was further digested with HindIII (see structure in Fig. 9b). m e HA fragment 20 and the SalI
/ HindIII digest of p~Vod were ligated together using T4 DN~ ase in the presence of BamHI and SalI to give PR81 (see Fig. 9c), and the ligation mixture was used to transforn E. coli C 600.
Transformants were analyzed for the presence of PRBl-like plas~
mids (for predicted structure, see Fig. 9c) as follows: Grunstein colony hybridiæation wa~ performed in a manner similar to that des-cribed earlier in connection with p~l5 (see Fig. 4)O A nick trans-lated proke of the HA gene (BamHI/~ dIII fragment from pAfPR 8/34fms /4.76 kb) was prepared and hybridized to the PR~l-like transformants (see Figs 9d and 9e). Figs 9d and 9e show the autoradiograms of eight colonies. It can be seen frcm Figs 9d and 9e that this DNA probe hybri-dizes to most of the transformants. Nbgative controls are provided by the pSVod containing colonies indicated in the figure by arrows~
PRBl~like plasmids were further characterized by restriction analysis, as shown in Fig. 9f. The structure indicated for PR81 in Fig. 9c is verified by the formation of DMA fragments of the follow-ing sizes after digestion with the restriction enzymes indicated below:

- 27 - ~12027 Plasmids Si~es of frasments kbp Enzymes used PR~l 3.2, 1.5 - 1.6 EcoRI
PR 8/34 3.8, 1 3 ~cbRI
ESVod 3~3 EcoRI
PR81 4.5 HindIII/SalI
PR 8/34 3.3, 1.7 - 1.8 Hind~II/SalI
pSVod 2.7, 0.6 - 0.7 HindIII/SalI

The results provided in Fig. 9f indicate the presence of the above fragments.
The construction scheme for plasmid PR84 i5 presented in Figs lOa to lOe. Briefly, plasmid PR~4 places the HA gene under the con-trol of expression control region of the Drosophila heat-shock gene described heretofore. m e hybrid gene itself is placed under the repli-cation control of pla$mid pSVod. Digestion of plasmid 520 (Fig. lOa) with P uII and NcoI results in ~he formation of two fragments of a~wt 1 kbp in length, one of which (Fig. lOb) contains part of the SV40 origin of rep~ication sequence and part of the 650 bp heat-shock con-trol element, including 400 bp of 5'-nontranscribed sequence and 60 bp of the RMA leader sequence. m e two 1 kbp fragments were isolat-ed by electrophoresis o~ preparative gels of 1.7% low melting agarose.
Plasmid PR81 (Fig. lOc~ was digested with HindIII the ends were Eilled up with DNA polymerase Klenow fragment, and the ~N~ was further digested with NcoI which has a unique cutting site in this pla$mid situated within the SV40 ori~in of replication sequence (see Fig.
lOd). The resulting PR~l fragments were ligated with the 1 kbp frag~
ments isola~ed from plasmid 520 and the ligation mixture was used to transform E. coli C 600. The resulting transformants include plas-mids such as PR~4 whose predicted structure is shown in Fig. lOe.
A more detailed structure for PR~4 including partial DNA sequence predictions are presented in Fig. lOf in the Eorm of a single strand representation of plasmid sequences. As shcwn in Fig. lOf, PR~4 com-prises from the 5' end, successively, a 250 bp origin of replication segment of SV40 origin, a 346 bp sequence from p~R322, a 400 bp Dro-5' nontranscribed sequence (p51; pl3 æ3, see Karch et al.
(1981) J. Mol. Biol 148: 219-230) originally included in the 650 bp heat-shock con~rol sequence, a Drosophila 65 bp hsp gene R~ leader fragment (base pairs numbered 1 ~o 653, a linker segment of 8 kase-pairs, a haemagglutinin (~I~) R~ leader segment (base pairs number-ed 1-32), a HA signal peptide coding segment (base pairs numbered 33-83) and a HA protein coding segment (bp 84 onwards).
m e rather complicated construction scheme chosen for plasmid PR84 was checked by restriction digest analysis of the parent plas-mids used, the results of which are shown in Fig. lla. The follow-ing restriction si~es are predicted for the pla~mids:
PR81: 1 NcoI site, 1 HindIII site, the NcoI site in the SV 40 origin of replication sequenceO
PR 8/ms/34: 1 NcoI site, no site in HA gene.
520: 1 NcoI site, several PvuII sites in the lac operon a pair of 1 kbp NcoI/PvuII fragments, one of them containing the origin -hsp 70 promoter fragment.
That these predictions fit with the structures of these plas-mids illustrated in the construction schemes is shGwn by the results of the restriction analyses indicated by the gel pa~terns in Fig.
lla. The t~o 1 kbp NcoI/~ II fragments are clearly seen and are in-dicated in Fig. 11 (lane 17) by an arrow.
Further charac~erization of PR84-like plasmids was performed using the Grunstein colony hybridization procedure described before (see Fig. 4). In this case, a nick translated probe (appro~imately 108 cp~/Jug) of the u3A/XhoI fragment from plasmid p51 (containing the Drosclphila hsp con~rol element) was employed, and the results are presented în Fig. llb) A number of the positive colonies identified in this way were grcwn up for the preparation of small amcunts of each plasmid as des-cribed previously, and a series of restxiction analyses were perform-ed. The gel patterns so obtained are shown in Figs llc and d. The expected pattern of digestion products formed by XhoI digestion as predicted f.rom the structure of PR~4 presented in Fig. 10 are: one XhoI fragment of 405 hps and one very large fragment. Fig. llc shows the results of this analysis and the plasmid clones in lanes 2, 4, 6 and 8 have the expected pattern as co~pared to the HinfI standard digest pattern of pfiVod. These plasmids were selected for further analysis where they were digested either with XhoI~ I or with XbaI/NcoI. m e results obtained are shown in Fig. lld. m e expect-ed sizes of fragments produced from the map shGwn in Fig. 10 would be one fragment oE about 700 bp and one of 405 bp with XhoI/NcoI and one large fra~ment, and with XhaI/NcoI, one fragm~n~ of 650 bp and one large fragment. Fig. lld shows the validity of these predic~ions for all of ~he plasmids anal~zed.
DN~ frcm plasmid PR84 was used to transfect eukaryotic cells.
As suitable cells for bringing about such experiments COS cells (Gluz-man (1981) Cell 23: 175-182) were selected because of their useful-ness for the rapid expression analysis of gene constructions (Geth-ing and Sambrook (19813 ~ature 293: 620-625j. Af~er transfection, the cells were incubated at either 37C or 42C in the p~esence of labelled methionine (35S). Then the contents of the cells were ana-lyzed by imn~ne precipitation using rabbit antiserum raised against PR 8/ms influenza virus (a gift of Dr. J. Skehel, M.R.C. Laborato-ries, Mill Hill, L~ndon U.K.). m e complexes purified on protein A~Sepharose~ were finall~ subjected to electrophoresis on polyacry-lamide gels and identified by fluorography. m e analytical results indicated the presence in good yield of a protein (Mr = 75,000) in distinguishable from authentic glycosylated haemagglutinin (Elder et al. (1979) Virology 95: 343-350). Synthesis of haemagglutinin o~-curred bo~h at 37C and to a lesser extent at 42C. No such speci-fic ~olypeptides w~re o~served using any other transfecting DNA than of PRK4 nor in inmune ca~plexes other than with anti-PR 8 antiseraO
m e details of the construction and expression of the eukaryotic expression vectors are ncw set forth.

A. C~nstruction of plasmid 520 Fifty Jug of plasmid pRV15 were digested first with 50 units of SmaI for 2 halrs at 37C and subsequently with 50 units of SalI for 2 hours at 37C (buffers as reccmmended by New England Biolabs). The digest was then electrophoresed on a 0.9~ agarose gel. The region containing the 7 kb lac fragment was cut out of the gel, and the DNA
was recovered by electroelution (200 v, 6 hours).
DN~ in ~he eluate was collected by ethanol precipitation. The fra$ment was then dried and resuspended in 200~u1 of TE buffer, phe-nol- and ether-extracted twice, and then passed through a small Sepha-~ 30 - 1312027 dex ~ ~75 column as described before. The DN~ in ~he column eluate was collected by ethanol precipitation.
Ten JUg of plasmid pSVod were digested with 10 units of Ba~HI
for 2 hours at 37~C. The digested DN~ was purified by 3 phenol-ex-tractions, 3 ether-extractions and 2 ethanol precipitations. The dried DNA was then resuspended in TE buffer and incu~ated with several units of DNA polymerase Klenow fragment in the presence of 0.5 mM deoxya-denosine triphosphate (dATP), deoxycytidine triphospha~e (dCTP), deo-xyguanosine triphosphate (dGrP) and deoxythymidine triphosphate (dTTP) at 6 - 7~C for one hour. m e DNA was purified again by repeated phe-nol- and ether-extraction and ethanol precipitation. m e purifi~d DNA was then digested with lO units of SalI ~or 2 hours at 37C. The digestion products were purified again as described above. The '7 kbp pkV15 SmaI/SalI ~ragment and the ab~ve described pSVod fragments were then incubated overnight (in a lar ratio of lO:l) with an excess of T4 DNA ligase at 14C. This ligation mixtur~ was used to trans-form E. coli MC 1061. SmaI and BamHI were included in th~ ligation mixture. Transformants were plated on L3 plates containing 10 Jug/~l ampicillin and 40 Jug/ml Xgal. Blue transformants were isolatedO DN~
from such reco~binants was prepared as described by Schedl et al.
(1978) Cell 14: 921-929. Rbcombinants were identified by restriction analyses.

B. Construction of plasmid PR~l Fifty ~9 of plasmid ~/PR 8/34 Clone No. 4.76 obtained from Prof.
G. Brownlee, University of Oxford, U.K. were digested with 30 units of HindIII and Ban~II for 2 hours at 37C. The digest was electropho-resed on a 0.85~ agarose gelO The HindIII-BamHI haenkagglutinin gene fragment was purified by electroelution as described above. .~liquots of this DW~ were nick translated to serve as probes for identifying PR81 and similar recombinants by Grunstein colony hybridization (spe-cific radioactivity of probes about 108 cp~/~ug).
Fifty ~g of A/PR 8/34/4.76 were digested with 50 ~mits of BamHI
for 2 hours at 37C. The DN~ was phenol-extracted three times, ether-ex~racted and ethanol-precipitated. The dried DNA was then re-suspended in 25 ~ of TE buffer and incubated with several units of - 31 - ~ 3 1 2027 DNA polymerase Klen~w fra~ment and 0.3 n~1 of all Eour deoxynucleo-tide triphosphates in a total volume of lOOJul for 90 minutes at 6-7C. The D~A was then purified by phenol- and ether-ex~ractions as akove. Following ethanol precipitation, the DN~ was dried and then resuspended in 25 ~ oE T~ buffer. It was digested subs~quently with 54 units ~f HindIII for one hour at 37C. The digested DN~ was re-purified by phenol-extraction.
Ten ug of plasmid pSVod (Mellon et al. (1981) Cell 27: 279-288) were digested with 30 units of SalI for one hour at 37C. The D~
was purified by phenol-extraction as above and was then incubated with several units of Klenow fragment in 100 ~ul in the presence of O.5 mM, dATP, dCTP, dGrP and dTTP for one hour at 37C. Fo~lowing repurification by phenol-extraction, the DN~ was digested further with 25 units of HindIII for one hour at 37C and repurified by phe-nol extraction.
An A/PR 8/34 digest (1.5~ug) and a pSVod digest (2 pg) were incu-bated overnight at 14C wi~h an excess of T4 D~ ligase in a total volume of 30 ~ . Several units of BamHI and SalI were included in the ligation mixture. The ligated DN~ was purified by phenol extrac-tion as described above. It was then digested with 6 units of BamHI
for one hour at 37C, followed by i~cubation with 12 units of SalI
for one hour at 37C. Aliquots of this reaction mixture were used to transfonm CaC12-treated E. coli C 600. Transformants were isolat-ed on LB plates containing lO~ug/ml ampicillin. About 350 clones were purified and analyzed by colony hybridization using the nick trans-lated HindIII - Bam~I haemagglutinin gene fragment described above as probe. Approximately 80 - 90~ of the clones were fo~d to contain a haemagglutinin gene insert.
Seven of the positive clones were grown up, and DNA was prepared from them. These DNAs were co~pared wi~h the original clones A/PR
8/34 and pSVod by restriction digestion and electrophoresis on a 0.9%
agarose gel.

Predicted fra~m~nt sizes Digestion with EcoRI
Reco~binants 3.2 kbp, 1.5 - 1.6 kbp A/PR 8/3d~ 3.B kbp, 1.3 k~p pSVod 3.3 kbp Digestion with H~ndIII and SalI
~ecombinants 4.S kbp A~PR 8/34 3.3 kbp, 1.7 - 1.8 kbp pSVod 2.7 kbpl 0~6 - 0.7 kbp Five out cf the seven reccmbinants that were analyzed showed the expected dige~tion pattern. PR81 is one of these five clones.
PR81 DN~ was prepared as described by the method of Schedl et al. (197a) Cell 14: 921-929 and was analyzed further by restriction enzyme digestion.

C. Cbnstruction of E~asmid PR84 p520 DN~ (75 ~g) was digested with 25 units of PvuII for one hour at 37C and subsequently with 20 units of NcoI for 2 hours at 37C. The digest was electrophoresed on a 1.2% low melting agarose gel con~aining 0.05 fug/ml EtBr. The region including the tw~ 1 kbp ~coI~uII fragments (one o them is a hsp 70 gene promoter fra~ment) was cut out of the gel. The agarose piece and ln volumes of TE buf-fer were heat ~ to 65C for 10 minutes, and the DNA was extracted subseguently with an equal volu~e of phenol. After one additional phenol/chloroform and 3 ether extractions, the DN~ was precipitated t~o times with ethan~l.
PR81 DN~ (10 ~ug) was digested with 15 units of HindIII for one hour at 37C. Follcwiny purification by phenol extraction the DN~
was incubated with several units d Rlenow fragment and 0.5 mM of all four deoxyribonucleotide triphosphates for one hour at 6-7C.
After repurification, the DN~ was digested with 10 units of NcoI for 90 minutes at 37C and was repurified by phenol extraction-Aliquots (0.25 Jug) of the p520 1 kbp NcoI/PvuII fragments andaliquots (0.05~0.1~ug) of the PR~l fragments were incubated overnight at 14C with an excess of T4 DM~ ligase. This mixture was used to transEorm E. coli C 600. A~out 600 ampicillin-resistant transformants were isolated and examined by Grunstein colony hybridization using a nick translated p51 Sau3AJXh I prcmoter fragment as a probe (ap-proximately 108 cpm/~g) (Fig. llb).
Ten positives were grcwn up and D~A was prepared frcm these clo~
nes. Aliquots of their DMAs were di~ested with XhoI and were analyz-ed on 5~ polyacrylamide gels~
The presence and sizes o~ XhoI fragments provide~ gcod evidence for the presence in the recGmbinant DNAs of both the hsp 7a kdal gene promoter and the haemagglutinin gene. Fbur out of the 10 clones test-ed showed the expected restriction pattern (two fragments: one lar-ge, about 5 kbp, and one of 400 bp) (Figs llc and d).
A~ditional digestions with NcoI/XhoI and XbaI/NcoI confirmed our initial analysis. PR~4 is one of the four plasmids which have the correct structure.

D. Expression e~pe_ ments usin~l~lasmid PR84 COS-I cells ~Gluzman (1981) Cell 23: 175-182) were maintained in Dulbecco's MEM with 10~ new-born calf serum and were subcultured one day prior to trans~ection to give cultures containing about 106 cells per 6 cm diameter petri dish after a fur~her one day of cul-ture. Transfection of cells was performRd as follows;
m e medium was r~moved Erom cell cultures at room tenperature and the cultures were washed twice wi~h 5 ml of phosphate-buffered saline (PBS)o Subsequently one ml of the follcwing ~reparations per dish were added~
1. 1 ml of Dulbecco's MEM with no serum, but containing DE~E
Dextran ~500Jug) and chloroquine t200Jug)~
2. As 1, but containing in addition, lOJug of plasmid 520 DN~.
3. As 1, but containing in addition, lO~ug of plasmid PRB4 DNA.
Transfections were perfonmed on 5 dishes oE cells for each of the three conditions stated above.
Cell cultures were maintained at room temperature for 30 minu-tes with cccasional gentle rocking to distribute the transfection solution over the entire cell culture. After removal of the trans-fection solution, 5 ml per dish of Dulbecco's MEM with 10% serum was added. Cell cultures were further incubated at 37C for 36 hours.
Half of the two c~ltures were incubated at 42C for the last 4 hours - 3~ -of this period. After this period, the medium was removed and cul-tures were washed once with a ~ulbecco's MEM containing 3 mg/l of L,methionine and 1% new-born calf serum. Cell cultures were label-led by addition of 1 ml of this same low methionine m~dium per dish but in the presence of 35S-~ethionine (50 JuCi/ml, 968 Ci/mmol). Cul tures were incubated for one hour at either 37C or 42C in the case of the heat-shock sanplesO After this period of labelling, the 35S-Methionine medium was remcved, the cultures washed three times with ice-cold PBS, and cells were scraped off in 1 ml of NET buffer (0.1 M NaC1, 0.01 M Tris-HCl pH 7.8, 0.001 M EDTA, 0.5% NP 40). After vigorous pipetting to break dcwn the ou~er cell membranes, nuclei and cellular debris were removed by centrifugation at 5000 x g for 5 minutes at 4C. m e cytoplasmic supernatants were removed and aliquots were taken for immune precipitation as follows:
Ali~uots (500 ~ ) of cytoplasmic extracts were kept on ice, 5 ~
of specific an~ibodies were added and the solutions were mixed gent-ly at 4C for two hours. The antibodies used were the following: Nor-mal rabbit antiserum, antiserum raised against purified E. coli ~-galactosidase, and antiserum raised against PR$ influenza virus obtained from Dr. J. Shekel, MRC Laboratories, Mill Hill, London, U.K. Anti HA antisera was preadsorbed for 30 minutes in NET buffer using cytoplasmic extracts of unlabelled COS cells prior to use for immune precipitation of labelled samples.
Antigen-anti~ody ccmplexes were isolated by the addition of 50~ 1 of a 50% suspension of protein A-Sepharose ~ (Pharmacia), continuing the shaking at 4C for one hour, and recovering the bound conplexes by centrifugation at lS,000 x g for one minute. m e complexes on pro tein ~-Sepharose ~ were washed and centrifuged four times using 1 ml of NET and three times using 1 ml of ~IPA buffer (0.15 M NaCl, 0O05 M Tris-~Cl pH 7.5, 0.02 M EDTA, 1 M Urea, 1% Triton x 100, 1% sodium deoxycholate). The final Sepharose pellets were boiled for 3 minu-tes in 100 ul of sample buffer (Laemmli (1970) Nature 227: 680-685).
After centrifugation at 15,000 x 9 for 1 minute, the samples were subjeted to electrophoresis on 10~ polyacrylamide gels as indicat-ed by Laemmli (1970) In the case of cell cultures transfected with plasmid PR84 D~
and in which the labelled cytoplasmic extracts were immune precipi-~ 35 ~ 1 ~ 1 2027 tated with antiserum raised against PR~ influenza virus, the gel analy-sis by fluorcgraphy (Laskey and Mills (1975~ Eur. J. Biochem. 56:
335 - 341) clearly indicated the presence of a protein which is in-distinguishable in size (~ = 75~000) from authentic glycosylated haemagglutinin (Elder et al. ~197~) Virology 95: 343-350). Some pos-sible prccessing to give polypeptides of Mr approximately 42,000 and 36,000 was observed in scme of these experLments. Synthesis of haemag-glutinin occurred both when labelling was performed at 37C and to a lesser extent at 42C. No such specific polypeptides were observ-ed using any other transfecting DNA than PR84, nor in immune preci-pitations other than with anti PR8 antisera.
In a second series of similar trans~ection experiments, COS cells transfected with PR84 were subjected to a period of heat-shock o~
five hours at 42C and were subsequently labelled with 35S-Methio-nine as described above but at 37C for 14 hrs. Subsequent immuno-precipitations with anti H~ antisera indicated a larger quantity of HA product than in non heat-treated s~.~les labelled under similar conditions.
These results suggest that the Drosa~h la heat-shock promotor fiunctions better at the heat shock te~perature in monkey cells than at the 37C control temperature. Sone activity of this prom~tor at 37C is perhaps to be e~pected as we have previously noted the use of 37C as the heat-shcck temperature for Droso~hila cells~ m e im-proved synthesis of HA when cells are returned to 37C after a pe riod of heat shock suggests that the translational control of the Drosoehila heat-shock promotor fragment, in particular the osophi-la rikosome hinding site is non-functional or poorly functional in monkey cells at least at the higher te~perature. m us it is possi-ble that although promotion of transcription occurs at 42C using the Drosophila DN~ sequences, ribosome binding and initiation of trans-lation may occur using the HA ribosome hinding site, optimLlly at 37C.
According to the present invention, expression vectors are pro-vided which yield efficient transformation of eukaryotic hosts and high levels of expression within the host. In particular, the trans-criptional/translational control region of a 70 kdal heat-shock pro-tein has been isolated and joined to a replication system derived ~ 36 - 1 3~2~27 from a sLmian virus. A human influenza virus haemagglutinin gene was inserted into the resulting vector under the control of the said con-trol region, and the resulting plasmid used to transfect a mammalian cell culture and apparently glycosylated haemagglutinin was produced.

Construction of plasmid F629 A further Drosophila control element/HA gene construction was m~de to extend the utility of plasmid constructions 5uch as PR~4.
The HA gene being a viral gene is rather specialized as a model for any eucaryotic gene with its 3' end signals, and its RN~ leader sequenr ce, hence the PR~4 construction contains both an HA RNA leader and a Drosophila RN~ leader.
As a more general exa~ple of the construction o a gene expres-sion unit using heat-shock control elements plasmid p629 was cons-tructed. m is pla$mid contained in addition to the Dros~2_ila heat-shock promotor the entire heat-shock RNA leader region, but was at-tached to a ~A gene lacking the first 40 amino acid codons. For conr venience, a plasmid p622b (a gift and personal communica~ion of R.
Voell~y~ was e~ployed; this plasmid contains a 450 bp XhoI~ II frag-ment carrying the Dros~phila hsp 70 prom~tor and ~A leader sequen-ce (repreænted by the XhoI/Sau3A fragment of plasmid pRW15 de~crib-ed earlier). p622b is a derivative of the C05 cell vector pSVod. The sequence around the lII site is given below:

...... hsp 70 ~ A~CG~IC ~ ................ HA gene.. i.~
1~L_--450 bp~ II oI Ba~BI
XhoI

Experimen~al construction ofJ~_29 15 ug of p622b were digested with 12 units (per 15 min) of ~
and 15 ug of PR 3/ms/34 with 15 units of XhoI. Digests were for 2 hrs at 37C. m e DMAs were then extracted three times with phenol, three times with ether and precipitated two times with EtOH. Both DNAs were resuspended in 20 ul of TE buffer and ends were filled in a reaction involving several units of DNA polymerase (Klenow frag-13~2027 ment) and 0.5 m~ of each of the four dNTPS (total volume: 100 ul;reaction at 6C for 2 hrs). ~oth D~s were again phenol-extracted and purified as above and again dissolved in 20 ul of TE ~ufferO They were then digested individually with 15 uni~s of Ba~HI ~to~al volu-me 100 ul; 37C for 2 hrs). Follcwing phenol extraction the DNAs were collected by EtOH precipitation and dissolved in 15 ul of TE buffer Different sets of ligations containing 1 ul of 622b digest and 1-5 ul of PR 8/ms/34 digest in a total volume of 25 ul were carried out (enzyme: 1-2 ul, T4 DNA ligase, ~ew England Biolabs) incubated over-night at 14. The ligase was inactivated at 65C for 10 min and the samples were then digested with several units of ClaI and Sall (suc-cessively, using the conditions suggested by New Ehgland Biolabs) for ~ hrs at 37C in volumes of 50 ul. The DNA mixtures were then used to transform E. coli MK1061.
About 300 individuzl transformants were characterized further using the Grunstein colony hybridization assay. Two separate sets of hybridizations were carried out: prokes used were: a) the 450 bp Xhc- lII fragment from F622b and b) the 1775 bp HindIII/BamHI gene fragment from P~8O4.76. Fragments had been separated in low melting agarose gels and were suhsequently eluted from these gelsO They were then labelled by nick translation (107-108 cEm/ug; 106 cFm~filter)~
8 colonies that hybridized to koth probes were selected, (see Figs 12c and d) and DN~ was prepared from them using the mini-DNA prepa-ration procedure described previously. The DMAs were digested with XhoI or with ~ EcoRI and the digests were analyzed on 5% poly-acrylamide gelsO Xhol was e~pected to yield two fragment~ with sizes of 450bp and of about 5kb. The BglII/RI digest should give fragments of 1130 bps, 1100 bps and of about 2.6 kb. Three of the eight clo-nes tested had the correct restriction patterns (see Figs 12a and b).
-Expression experLments using plasmid p629 Transfection of C06 I cells was performed in a manner analogousto tha~ described for the expression experiments using plasmid PR~4.
I = ne precipitations performed as described previously, precipitat-ed one major and a number of minor polypeptides of M.Wt around 75,000 daltons. When 35S-Methionine labelling was performed in the presence of 1 u~/ml tunicamycin, one major band (Mr =
75,000) alone was observed, suggesting that the minor bands were the result of glycosylation. Synthesis of HA protein was greater duriny a labelling period of 14 hrs at 37C
following a period of heat-shock than in parallel labeling periods in the absence o~ heat-shock.

The cultures of the micro organisms are on file at "The American Typeculture collection in Rockville, U.S.A.
and -three cultures exist of E.Coli containing plasmids PRV-15,PR-84 and PR-81."

Claims (37)

1. A host cell transformed with a gene expression unit, comprising: (a) an expression control region of about 650 bp from a eucaryotic HSP70-gene carrying promoters, operators, activators, cap signals, ribosomal binding signals and leader sequences from an eucaryotic heat-shock protein gene; and (b) a structural gene of interest under the transcriptional and translational control of said expression control region (a);
and wherein the host cell comprises cells which are either from bacterial origin or from at least one species similar to those from which the said expression control region was isolated and in which control region (a) linked to the said structural gene (b) of interest is present in order to produce the product of said structural gene.

2. A host cell as in claim 1, wherein the structural gene of interest is inserted into the gene expression control region.

3. A host cell as in claim 1, wherein the heat-shock gene control region is derived from the genomic DNA of Drosophila melanogaster.

4. A host cell as in claim 3, wherein the heat-shock gene control region is derived from any one of the genes encoding the 70 kilodalton heat-shock proteins of Drosophila melanogaster.

5. A host cell as in claim 1, wherein at least part of the functional heat-shock gene control region is of synthetic origin.

6. A host cell as in claim 1, in which said expression control region further comprises at least one selectable marker linked to the expression control region which allows for selection of transformed hosts.

7. A host cell as in claim 1, in which said expression control region further comprises a procaryotic replication system which allows propagation of the expression control region, and at least one antibiotic resistance gene.

8. A host cell as in claim 1, wherein said expression control region is linked to a fragment containing an eucaryotic extrachromosomal replication system.

9. A host cell as in claim 1, wherein the expression control region is linked to a cellular or viral transcription enhancer element to allow for constitutive expression of the gene of interest.

10. A host cell as in claim 1, wherein a heat-shock promoter element is linked to the complete RNA coding region of the structural gene of interest.

11. A host cell as in claim 1, wherein an eucaryotic heat-shock gene promoter RNA leader segment is linked to the protein-coding region of the structural gene of interest.

12. A host cell as in claim 1, wherein an eucaryotic heat-shock gene control element is linked to at least part of the DNA-coding region of the structural gene of interest.

13. Plasmid pRV15.

14. Plasmid PR84.

15. A method for preparing the host cell of claim 1, comprising introducing the expression unit into the host cells by cotransformation with a selectable marker.

16. A method as in claim 15, wherein the expression unit is introduced into the host cells with an amplifiable gene.

17. A method as in claim 15, wherein the expression unit is introduced into the host cells by transformation or transfection.

18. An expression vector comprising: an eucaryotic extrachromosomal replication system; an expression control region of about 650 bp from an eucaryotic HSP70-gene carrying promoters, operators, activators, cap signals, ribosomal binding signals and leader sequences from an eucaryotic heat-shock protein gene; and at least one inserted protein encoding sequence gene under the transcriptional control of the expression control region.

19. An expression vector as in claim 18, said vector being free of a translational start codon between the expression control region and the insertion site.

20. An expression vector as in claim 18, having a translational start codon upstream from the insertion site.

21. An expression vector as in claim 18, wherein the heat-shock gene control region is derived from the genomic DNA of Drosophila melanogaster.

22. An expression vector as in claim 21, wherein the heat-shock gene control region is derived from the heat-shock gene which encodes the 70 kilodalton heat-shock protein of Drosophila melanogaster.

23. An expression vector as in claim 18, further comprising a prokaryotic replication system which allows stable maintenance in prokaryotic hosts.

24. An expression vector as in claim 18, further comprising at least one selectable marker which allows for selection of transformed hosts.

25. An expression vector comprising:

an eukaryotic extrachromosomal replication system;

a heat-shock gene control region substantially homologous to a 650 base pair sequence from an eucaryotic HSP70-gene carrying promoters, operators, activators, cap signals, ribosomal binding signals and leader sequences which controls transcription of the 70 kilodalton heat-shock protein in Drosophila melanogaster; and at least one inserted protein encoding sequence gene under the transcriptional control of the heat-shock gene control region.

26. A DNA construct of less than 15 kbp which includes an expression control region of about 650 bp from an eucaryotic HSP70-gene carrying promoters, operators, activators, cap signals, ribosomal binding signals and leader sequences from an eukaryotic heat-shock gene and at least one inserted protein encoding sequence gene under the transcriptional control of the heat-shock gene control region.

27. A DNA construct as in claim 26, wherein the heat-shock gene encodes the 70 kilodalton heat-shock protein of Drosophila melanogaster.

28. A method for producing competent gene products, said method comprising:

joining a structural gene to an expression control region of about 650 bp from an eucaryotic HSP70-gene carrying promoters, operators, activators, cap signals, ribosomal binding signals and leader sequences from an eukaryotic heat-shock gene;

introducing said structural gene and control region into a eukaryotic host cell suitable for expression of the structural gene; and growing said host, whereby said product is produced.

29. A method as in claim 28, wherein the structural gene and control region are joined together in an expression vector having an eukaryotic extrachromosomal replication system.

30. A method as in claim 28, wherein the structural gene and control region are introduced to the host by co-transformation with a selectable marker.

31. A method as in claim 28, wherein the expression control region is derived from a heat-shock gene which encodes a 70 kilodalton heat-shock protein in Drosophila melanogaster.

32. A method as in claim 28, wherein the heat-shock gene is derived from a different eukaryotic species from the host.

33. A method as in claim 28, wherein the structural gene is a mammalian gene.

34. A method as in claim 33, wherein the host is a mammalian cell culture.

35. A method as in claim 28, wherein the structural gene is a human gene.

36. A method as in claim 35, wherein the host is a mammalian cell culture.

37. A method as in claim 36, wherein the host is a COS I
cell culture.