FR2665908A1 - Centromeric sequence and DNA comprising such a sequence - Google Patents

Abstract

The present invention relates to a centromeric sequence, characterised in that it is genetically linked to an ARS sequence which is effective in Y. lipolytica, and to a DNA comprising such a sequence.

Description

 The present invention relates to centromeric sequences which are linked to autonomous replication sequences (ARS sequences) effective in the yeast Yarrowia lipolytica. It also relates to the preparation of centromeric vectors and minichromosomes, comprising these sequences, in Y. lipolytica, as well as their applications.

 In eukaryotes, the centromeres, hereafter designated by regions or fragments "CEN" are specific chromosomal regions which attach to the microtubules of the spindle during nuclear divisions and are responsible for the regular segregation of chromosomes at mitosis and at meiosis. There is normally only one centromere per chromosome. If a second centromere is introduced into a chromosome, this will lead to the rupture of the aberrant chromosome during mitosis. Any DNA molecule having a centromeric sequence may after cell division be regularly found in the two daughter cells.

 The "CEN" fragments characterized on the plasmids give them mitotic stability, and although not having large regions of homology, they share certain common features, in particular a region having an A + T base content greater than 90%. .

 These centromeres have so far been characterized at the molecular level only in two yeast species, S. cerevisiae and S. pombe (Clarke L., Trends in Genet. 1990, 6, 150-154).

 They have very different sizes and structures in these two organisms: 100 bp in S. cerevisiae, around 100 kb in S.

whore.

 In the yeast Y.lipolytica, sequences conferring the autonomy of replication (or ARS sequence) have been isolated from plasmids of this yeast.

 The vectors carrying ARS sequences are characterized by a high frequency of transformation and segregated instability.

They can be obtained after constitution of a bank of genomic fragments of Y. lipolytica in an integrative vector complementing an auxotrophy of Y. lipolytica, by transformation of a strain of Y.

lipolytica with this auxotrophy. The transformants are then selected to highlight the extrachromosomal plasmids which carry an ARS sequence; this ARS sequence gives the plasmids a transforming power at least 5 times greater than the transforming power of the original vector. These ARS sequences are functional in Y. lipolytica and can be used for the production of expression vector plasmids which can be maintained in the extrachromosomal state and can be amplified.

 However, the inventors have now found that the functional ARS sequences on plasmids isolated from Y. lipolytica have atypical behavior. They are rare in the genome of Y. lipolytica one for at least 1000 kb of genomic DNA, against one for 10 to 30 kb in other yeasts. The plasmids carrying these sequences have a much higher mitotic stability than that of the ARS plasmids described elsewhere. In addition, the number of copies per transformed cell is low (of the order of 1 to 5), while in other yeasts the transformed cells typically contain several tens of copies of the transforming vector.

 Quite surprisingly, it has been found that some of these autonomous replication sequences are associated with a centromeric sequence.

 This is why the present invention relates to a centromeric sequence characterized in that it is genetically linked to an ARS sequence effective in Y. lipolytica.

 The possibility of isolating the four products of meiosis in the form of ordered tetrads makes it possible to determine the distance between a gene and the centromere of the chromosome on which it is located. The binding to the centromere is measured, after sporulation of a heterogygous diploid for two markers A and B (type AB / ab), by analysis of tetrads.

If A and B are not both linked to a centromere, three types of ascs are possible: parental ditypes (DP) [AB AB ab ab], recombinant ditypes (DR) [Ab Ab aB aB] - and tetra-types (TT ) [AB ab Ab aB1 in proportion lDP: lDR: 2TT. If the two markers are each linked to a centromere, we no longer observe the TT class (Mortimer and Schild (1981) in The Molecular
Biology of the yeast S. cerevisiae, Cold Spring Harbor ed., P. 11-26). There are 4 or 5 chromosomes in the yeast Y. lipolytica and the genetic map currently known does not contain any marker linked to the centromeres (Ogrydziak (1988) J. Basic Microbiol. 28, 185-196).

 To demonstrate a binding to a centromere of an autonomous replication sequence (ARS), the chromosomes are genetically marked in the vicinity of ARS and it is shown that these sequences are indistinguishable from centromeres in tetrad analysis. The labeling is carried out by substitution of a wild chromosomal copy by a deleted gene created in vitro.

 From the plasmids comprising the ARS sequences, it was possible to isolate narrower sequences carrying the ARS sequence. Thus it was possible to characterize in particular 2 different types of ARS sequences carried by the plasmids pINA119 and pINA123. The effective part of these sequences, noted ARS18 and ARS68, is located respectively on DNA fragments of 1.3 kb and 2.3 kb.

 The 1.3 kb fragment of ARS18 was characterized by its nucleotide sequence. The sequences responsible for high frequency transformation and extrachromosomal replication can be physically located on two adjacent regions.

 The present invention therefore relates to a centric sequence characterized in that it is carried by the ARS18 sequence of 1.3 kb, effective in Y. lipolytica.

 The present invention also relates to a centromeric sequence characterized in that it is carried by the ARS68 sequence of 2.3 kb, effective in Y. lipolytica. This fragment, BamHI-BglII, obtained by successive deletions of a 6.6 kb fragment is capable of maintaining plasmids in the extra-chromosomal state.

The present invention also relates to a DNA comprising a centromeric sequence genetically linked to a sequence
ARS effective in Y. lipolytica.

 This DNA can also comprise an adjacent inducible promoter which controls the centromeric sequence. This DNA will preferably be a plasmid, whose amplification thus inducible makes it possible to overcome the constraint of the low number of copies of the plasmids imposed by the centromeric regions. Thus, replicative vectors with a high number of copies are obtained. These plasmids may therefore also contain elements allowing the expression of a protein in Y. lipolytica, according to techniques common to those skilled in the art.

 The insertion of CEN sequences into a plasmid must confer on the latter the properties of a minichromosome, in particular regular segregation into meiosis. If a CEN + strain is crossed with a CEN strain and the diploid is sporulated, 50% of CEN + strains and 50% of CEN strains are found in the offspring. In tetrad analysis, we observe 2+: 2- or 4+: 0- type segregation (and not 1+: 3- as with the ARS vectors).

 The present invention also relates to DNA comprising a centromeric sequence genetically linked to an ARS sequence as defined above, characterized in that it also comprises telomeric sequences at its ends. Indeed, the insertion of centromeric sequences in a chromosome must lead to the formation of dicentric chromosomes known to be unstable, which must break at mitosis and be lost. Some, however, following a break-out cycle, can be stabilized by the addition of telomeres. We then observe a disappearance of the initial chromosome and the appearance of new linear minichromosomes, stabilized by the addition of ciliate telomeres to the free ends. Fully artificial minichromosomes can also be created in vitro, by assembling telomeres, centromers, ARS and selection marker, plus any desirable DNA sequence. On small minichromosomes ((50 kb), the centromeric function is partially extinguished, allowing an increase in the number of plasmid copies in the transformed cells.

The present invention also relates to DNA comprising a centromeric sequence as defined above and characterized in that it comprises a sequence belonging to the genome of
Y. lipolytica, and in that this sequence is modified.

 This is how we can screen for mutant genes in Y.

lipolytica by the use of exchangeable plasmids, based on incom
patibility of CEN vectors between them.

 The present invention also relates to a DNA having one of the structures described above, characterized in that it comprises, in a region adjacent to the centromeric sequence, a sequence allowing its insertion into a site of a chromosome of Y.

lipolytica.

 Plasmids of this type make it possible to target chromosome fragmentation at precise sites, making it possible inter alia to physically locate a gene on a chromosome.

 We can therefore constitute a genetic map of Y. lipolytica thanks to the creation of a strain carrying marked centromeres.

In the examples which follow, reference will be made to the appended figures which represent
FIGURE 1: Substitution of a wild chromosome copy by a
deleted gene created in vitro. Plasmid pINA270 is linear
designed by NotI
FIGURE 2: Definition of the ARS18 and ARS68 regions and their regions
adjacent. The indicated fragments were cloned into a
LEU2 vector and tested for their ability to transform the
integrative (S) or replicative (U) yeast. The
sizes are expressed in kb. The enzymes used are AccI
(A), BstBI (B), BamHI (Ba), BglII (Bg), EcoRV (ex), HindIII
(H), Sau3A (S), TaqI (T) and XbaI (X).

FIGURE 3: pINA242 and pINA450 plasmids. The XhoI sites used for
chromosomal targeting are underlined.

FIGURE 4: Analysis of 13 stable clones after integration of ARS 18.

 Only clone 3 results from a conversion.

FIGURE 5: Separation in pulsed field electrophoresis of
chromosomes of the dreaming strain INAG33122 (MatB, leu
2-35, lys 2-5, scpr2) and two integrants obtained with
pINA176 or pINAl 19. The additional band is indicated
by #, l the position of the chromosomes by #
In 1 and 2: transformants. In 3: recipient strain
In 4: witness S. pombe; In 5: S. cerevisiae
FIGURE 6: ARS18 deletion plasmid. The region flanking ARS18
is shown in dotted lines. P and sp represent the sites
PstI and SphI used for transformation. Enzymes
used are EcoRI (E), ClaI (C), HindIII (H), XhoI (X), BglII
(G), BamHI (B) and SalI (S)
FIGURE 7: Analysis of transformants obtained with ARS68 LEU2 and
ARS18 URA3
FIGURE 8: Creation of dicentric plasmids by in recombination
vivo. After digestion with HindIII, the expected sizes in kb
are, for pINA386: 9.1; for pINA3lî: 4.8; 5.6; for
pINA447: 2.9; 4.8; 12.5.

EXAMPLE 1: OBTAINING ISOGENIC STRAINS CARRYING
STABLE MARKERS FOR USE IN TRANSFOR
MATION.

 The analysis of tretrads presupposes the use of strains giving a good frequency of spore germination and regular segregation at meiosis. This is currently the case only for a few strains of Y. lipolytica resulting from an isogenization program; these strains do not carry markers which can be used in transformation and it is undesirable for obvious reasons to subject them to a cycle of mutagenesis. It would however be desirable to introduce therein such markers, preferably known deletions constructed in vitro. We have constructed such strains by transformation as described below.

 We previously described the creation of the strain E105 by introduction of a marker ura3-302 in the strain AHa resulting from a genetic improvement program using a disruption vector pINA302 (expressing S invertase . cerevisiae, EPA 90401503-9). Here we describe an alternative strategy for generating by transformation strains carrying a defined mutation. As will be seen, this strategy can be generalized to any marker for which the cloned wild gene is available.

 The E105 strain (MatA, his-1, ura3-302) was crossed with an isogenic Ahd strain (MatB, lysi 1-23) and, after sporulation of the diploid, an Exil9 strain (MatB, lyslî-23, ura3-302 ) has been isolated.

This strain was transformed by integration of the plasmid pINA270 (see FIG. 1) targeted at the LEU2 locus by digestion with the enzyme
NotI. pINA270 is derived from the plasmid pINA62 (Gaillardin and Ribet (1987) Curr.

Broom. 11, 369-375) by StuI digestion and religation of the plasmid on itself, thus creating a deletion of 682 bp in the coding part of the LEU2 gene. The 1489 bp HindIII-BamHI fragment of pINA156 (EPA89401556-9) carrying the URA3 gene was then inserted between the sites
HindIII and BamHI of this plasmid to generate pINA270.

 The integration of this vector at the LEU2 locus of AHa generates an imperfect duplication of LEU2 flanking the plasmid. Excision of this plasmid by a homologous recombination event between the duplicated regions may or may not leave in place the deleted copy of the LEU2 gene (see FIG. 1). This excision makes the strain Ura-. In S. cerevisiae, the ura3- strains are known to be resistant to 5-fluoroorotic acid (5FOA), a compound toxic to wild strains because it is converted into 5-fluoroUMP by the product of the URA3 + gene (Boeke et al. (1984) Mol.

Gen. Broom. 197, 345-346). We established that the strain Ura + AHd was effectively incapable of growing on a minimum medium containing glucose (10 g / l), Yeast Nitrogen Base Difco (7.5 g / l), lysine (100 mu / 1) , uracil (15 mu / 1) and 5FOA (1.25 g / l), while the strain Ura-El 19 grew normally on this medium. If up to 106 cells of the strain Eu 19 transformed with pINA270 are spread on this same medium supplemented with leucine (200 mg / 1), this strain Ura + is inhibited by 5-FOA; a hundred resistant clones appeared, however, after 3 days. After purification, these clones were found to be Ura-, 50% of them being more Leu-. By Southern analysis, it can be demonstrated that all of these clones result from excision of the vector as described in FIG. 1. The Leu- clones in particular now carry the leu2-270 deletion, easily detectable by SalI digestion of the Genomic DNA followed by hybridization with a LEU2 probe: the wild band of approximately 5 kb is replaced by a band of approximately 4.3 kb. A MatB strain, lysll-23, ura3-302, leu2-270 was saved as E122.

 The strain E122 was crossed with AHa and, after sporulation of the diploid, isogenic strains MatA, his-i, ura3-302, leu2-270 were isolated. These strains were screened for their ability to be transformed using a vector pINA62 linearized by NotI, and the strain 22301-3 was saved. The presence of the deletions ura3-302 and ieu2-270 in this strain was confirmed by Southern analysis.

 This same process in particular allowed us to create derivatives of these strains or equivalent strains, carrying deletions in the genes of the exocellular protease (XPR2 gene) or in essential genes (RNA genes 7S SCR1 or SCR2) .

EXAMPLE 2 DEFINITION OF REGIONS ADJACENT TO ARS18 AND A
ARS68 USEFUL FOR CHROMOSOME TARGETING.

We have previously demonstrated that the region necessary for the extrachromosomal maintenance of ARS18 corresponds to a fragment
Fully sequenced 1.3 kb BamHI-Sau3A (EP 0329501). An adjacent BamHI-BglII fragment (FIG. 2), on the other hand, only gives rise to chromosomal integrations.

By successive deletions of the 6.6 kb DNA fragment carrying
ARS68, we also defined a BamHI-BglII fragment capable of maintaining plasmids carrying the LEU2 gene in the extrachromosomal state present in pINA386, see FIG. 2), while the adjacent BamHI-BglII fragment (present in pINA242) was incapable of this. and gave stable transformants at low frequency (see Figure 2).

Previous experiments had shown us that it was very difficult to integrate into the genome of plasmids carrying ARS18 or
ARS68 (see also example 4). We wanted to test whether the regions adjacent to the ARS defined above were capable of targeting a plasmid in the vicinity of ARS18 or ARS68. For this, the 2.8 kb BamHI-BglII fragment of pINA119 was introduced into a plasmid carrying the URA3 gene from Y. lipolytica: the resulting plasmid, pINA450 carries a region adjacent to ARS18 (see FIG. 3). The plasmid pINA242 previously mentioned carries a region adjacent to ARS68 (see Figures 2 and 3).

These plasmids do not carry, in the regions adjacent to the
ARS, a unique site that can be used for chromosomal targeting during an integrative transformation. On the other hand, they each carry 3 XhoI sites, 2 in the marker genes (URA3 or LEU2) and 1 in the fragment adjacent to the ARS. After digestion managed by the enzyme XhoI, the linear forms of each of these plasmids were isolated. pINA450 was used to transform
E122 (Ura + selection) and pINA242 was used to transform 22301-3 (selection
Leu +). In these strains which carry a deletion of the URA3 and LEU2 alleles, the plasmids opened by XhoI at the level of the marker genes cannot be integrated by homology and will transform at low frequency; on the contrary, the plasmids opened by XhoI in the regions adjacent to the ARS must transform at high frequency if their integration is stable. More than 1000 Ura + or Leu + transformants have been obtained: by Southern analysis, it can be demonstrated that the majority of they (18 out of 18 tested) contain the vectors integrated into the ARS18 and ARS68 loci.

EXAMPLE 3: SEGREGATION OF LOCI ARS18 AND ARS68.

The transformed strains E122 / pINA450 and 22301-3 / pINA242 were crossed and the diploid D224 was isolated on minimum medium.
MatA ura3-302 leu2-270 ARS18: URA3 ARS68 his-l +
MatB ura3-302 leu2-270 ARS18 Ars68: LEU2 + lys 1-23
After sporulation, 38 tetrads were dissected. The following results were obtained (Table 1).

TABLE 1
Tetrad analysis
Couples of DP DR TT markers
LEU2-URA3 14 24 0 LEU2-LYS11 10 5 23
LEU2-HIS 1 8 7 23
URA3-LYS 11 9 7 22
URA3-HIS 1 6 9 23 LYSîl-HISi 9 8 21
It can be seen that all of the pairs of markers tested exhibit lDP type segregation: 1DR: 2TT except the couple
LEU2-URA3 marking ARS68-ARS 18 for which no tetratype is observed. These results demonstrate that ARS68 and ARS18 are closely linked to different centromeres (DP = DR).

EXAMPLE 4: THE INTEGRATION OF ARS18 IN LOCUS LEU2 LEADS A
BREAKING OF CHROMOSOMES.

 If the BamHI-Sau3A fragment carrying ARS18 carries a complete centromeric function, its integration must generate an unstable dicentric chromosome.

 3 plasmids were used: pINA176, pINA232 and pINAil9 (see Figure 2) which carry ARS18 and LEU2. They were digested as specified below and introduced into a leu2- strain.

 By digesting pINA232 with BamHI which cuts only once in this plasmid, the integration is directed to the ARS18 locus. The transformation frequencies are as high as with the uncut plasmid (Table 2) and half of the transformants are unstable: after 10 generations in non-selective medium, Leu-clones appear. The unstable clones are probably the result of religating the plasmid on itself in vivo.

The stable clones were analyzed by the Southern technique with an ARS18 probe. Two types of events are expected
1- a conversion of the leu2- to LEU2 + allele without
integration of the vector, generating a single band in
Southern corresponding to the resident ARS18 locus, 12 kb
after digestion with the enzyme EcoRI or 2.15 kb after
digestion with BamHI.

2- integration of the plasmid by recombination
homologous to the ARS18 locus which, if the chromosome
resulting is viable, will generate duplicate ARS18
displayed by two EcoRI bands (13 and 6 kb) and two
BamHI bands (11 and 2.5 kb).

TABLE 2
Target Plasmid Frequency of Number of Number of Conclusion
transformation clones analyzed stable clones according to Southern
ARS18 pINA232 / BamHI 19 400 50 20 (40%) 1 LEU2 conversion
19 integrations of
ARS18 plasmid
ARS18 pINA176 / BstBI 70 5 1 (20%) 1 LEU2 conversion
LEU2 pINA176 / Xhol 840 21 1 (5%) 1 integration of
LEU2 plasmid
LEU2 pINA119 / Apal 4,000 20 3 (19%) 2 LEU2 conversions
1 integration of
LEU2 plasmid
In FIG. 4, it can be seen that out of 13 stable clones studied, 12 result from an integration into ARS18, 1 results from a conversion (clone 3). It therefore seems that the presence of 2 CEN sequences repeated in tandem at 9.2 kb distance is tolerated by the chromosome.
S. cerevisiae of chromosomes carrying two centromeres separated by a few kb- are stable for several generations, but they are quickly broken if the distance between them exceeds 40 kb (Haber et al.

(1984) Genetics 106, 207-226). Alternatively, it may be thought that the fragment carrying ARS18 does not carry a complete centrometric function.

We therefore tried to integrate ARS18 at the LEU2 locus which is not a priori linked to a centromere. pINA176 and pINA119 were digested with XhoI or ApaI which linearize these plasmids at the level of ARS18. We observe (Table 2) a drop in the frequency of transformation and a low frequency of stable clones. 4 stable clones were analyzed as above, 2 conversions and 2 integrations are observed, one with pINA176, the other with pINAl 19. In the latter two cases, the structure of the integrated plasmid does not appear to be altered. The chromosomes of these two integrants were separated by pulsed field electrophoresis (technique
FIGE) in parallel with those of the recipient strain. This technique reveals for it 4 chromosomes, of intermediate size between those of S. cerevisiae and those of S. pombe (see Figure 5). In the case of the two transformants studied, we can clearly see an additional chromosomal band appear. This hybrid band to an ARS18 probe.

All of these results demonstrate that the fragment
BamHI-Sau3A of 1.3 kb present in pINA176 and pINAl l9 carries a complete centrometric function in addition to the ARS function. I1 is the same for the 2.3 kb BamHI-BglII fragment carrying ARS68.

EXAMPLE 5: DISRUPTION OF ARS18 IN THE CHROMOSOME.

 Disruption of a centromere in S. cerevisiae leads to the formation of an unstable acentric chromosome which is quickly lost.

This event is lethal in a haploid and is detected only by a sharp drop in the frequency of transformation observed with the disruption vectors (in fact only the conversion events are observed).

In a diploid, if only one of the two homologous centromeres is destroyed, the strain quickly becomes monosomal.

 The ARS18 disruption plasmid, pINA293, was constructed as follows (Figure 6). The central BamHI fragment carrying ARS18 in pINA119 was replaced by a DNA fragment carrying LEU2. The plasmid pINA293 was digested with PstI and SphI and the fragment carrying LEU2 was purified. This fragment, where LEU2 is flanked by 1.8 kb and 0.5 kb of sequences adjacent to ARS18, was introduced into a haploid yeast leu2- or into a homozygous diploid leu2- / leu2-.

The low transformation frequency (300 Leu + clones per llg of transforming DNA) was observed with the haploid strain. 21 clones
Leu + were analyzed by the Southern technique: all of them resulted from a conversion of leu2- to LEU2 + without the chromosomal replacement of ARS18. We deduce that the deletion of ARS18 is probably lethal in the haploid.

 A higher frequency (3000 Leu + / clones, ug of DNA) was observed with the diploid strain; this frequency remains lower than that observed with a control plasmid during the same experiment (18,000 clones are Leu + / μg of DNA). In addition, the transformants appear after 4 to 8 days of incubation, against 3 to 4 days with the control plasmid. This suggests that strains with disruption of ARS18 in the chromosome show slow growth, possibly due to an unstable chromosome.

As the loss of the acentric chromosome results in a Leu- phenotype, it is possible that secondary events of conversion from Leu2- to
LEU2 + occur at low frequency in unstable colonies and regenerate stable Leu + subclones. All the clones analyzed in
Southern resulted from a conversion of leu2.

 These results are compatible with the hypothesis of an essential role of the ARS18 sequence in the maintenance of the chromosome.

EXAMPLE 6: METOTIC SEGREGATION OF ARS18 VECTORS.

 ARS plasmids are very unstable in S. cerevisiae and are most often lost at meiosis giving tetrads 0+: 4- or 1+: 4 preferentially. On the contrary, the ARS-CEN plasmids in this organism preferentially segregate like minichromosomes 2+: 2 or 4+: 0- (if there were 2 or more copies of the vector in the mother cell of the tetrad).

 To test the meiosis behavior of vectors carrying centromeres in Y. lipolytica, a plasmid ARS18-LEU2 was introduced into a diploid pro-1 / +, ura3 / +, leu2-35 / leu2-35. The viability of the spores being poor, only 2 complete tetrads were obtained. The Leu +: Leu- segregations observed were 2+: 2- and 4+: 4-. By mass analysis, the following +: - ratios were observed: 20:12 for LEU2, 19:13 for URA3, 18:14 for PRO-I. This demonstrates that the centromeric plasmids segregate like Mendelian markers and that they therefore have in Y. lipolytica the properties of a minichromosome.

EXAMPLE 7 COMPATIBILITY OF CENTROMERIC VECTORS.

 A series of CEN plasmids has been constructed. These plasmids contain the 1.3 kb BamHI-Sau3A fragment carrying ARS18 and CEN associated with a LEU2 or URA3 marker (pINA176 and pINA311), or the BamHI-BgiII fragment carrying ARS68 and CEN associated with these same markers (pINA386 and pINA443) .

In a first series of experiments, strains ura3, leu2 were transformed by a plasmid ARS18-LEU2, ARS18-URA3 or
ARS68-URA3. These plasmids have equivalent stability. These strains were then further transformed by ARS plasmids carrying the other markers. The transformants were selected for the presence of the new (incoming plasmid) on a medium which also allows the loss of the first plasmid (resident plasmid) (see Table 3). The transformants are then tested for the presence of the resident plasmid. It is found that in almost half of the cases the incoming plasmid has driven out the resident plasmid.

TABLE 3
Plasmid hunting by transformation
Plasmid Plasmid Selection Number of resident Auxotrophs entering clones tested ARS18LEU2 ARS18URA3 YNBleu 235 60% Leu
ARS68URA3 YNBleu 288 71% Leu
ARS18URA3 ARS18LEU2 YNBura 1032 50% Ura
ARS68URA3 ARS18LEU2 YNBura 1064 45% Ura
In a second series of experiments, we analyzed the stability of the transformants which had retained the incoming and resident plasmids. As seen in Table 4, the ARS18-URA3 or
ARS68-URA3 are preferentially chased by the ARS18- plasmids
LEU2 and their mitotic instability (expressed as a percentage loss per generation) becomes very high, of the order of that of ches ARS plasmids
The origin of this marker effect is unknown. These results demonstrate that the ARS-CEN plasmids are not very compatible in Y.

lipolytica.

TABLE 4
Coexistence of CEN plasmids
Coexisting Plasmids Loss of Plasmid by Generation
1 2 1 2 ARS18LEU2 ARS18URA3 3% 14% ARS 18LEU2 ARS68URA3 0.4% 30%
EXAMPLE 8 STABILITY OF DICENTRIC PLASMIDS.

 Most of the Ura + Leu + transformants described in Example 7 segregate clones Leu-Ura +, Leu-Ura- and Leu + ura-. Certain clones, however, were observed which only segregated Ura-Leu-. The structure of the plasmids present in this type of clone was analyzed by the Southern technique or after extraction and re-isolation in E. coli.

As an example, we examine the transformants obtained with
ARS68LEU2 (pINA386) and ARS18URA3 (pINA311) see Figures 7 and 8).

Total DNA is either undigested (wells 1 to 7) or digested by the enzyme
HindIII (wells 8 to 15), and revealed by a pBR322 probe. The molecular weights are estimated using a standard (well 16, lambda DNA digested with HindIII and with HIndIII and EcoRI). The control plasmids pINA386 (wells 6 and 14) and pINA311 (wells 7 and 15) have identical profiles to those found in the total DNA of the yeast transformants by pINA386 (wells 4 and 12) or pINA311 (wells 5 and 13). Three transforming clones or the co-aggregating markers LEU2 AND URA3 are presented in wells 1 to 3 and 8 to 10. We see that these clones contain a new plasmid, pINA447 whose size (approximately 20 kb) corresponds to the sum of the sizes of pINA311 and pINA386 (10.4 and 9.4 kb). The 20 kb plasmid was re-isolated in E. coli and its restriction map was established (Figure 8). This shows that pINA447 is a dimer of pINA311 + pINA386, probably resulting from a homologous recombination in vivo between the two ARS plasmids.

 pINA447 therefore represents an example of a structurally stable dicentric plasmid in yeast. Its mitotic stability has been compared to that of monocentric vectors. After 20 generations in a non-selective medium, pINA311 is lost in 12% of the clones, pINA386 in 10% and pINA447 in 25% of the yeast clones. The stability of the dicentric vectors in mitosis is therefore less good than that of the corresponding monocentrics.

Claims (9)

 1) Centromeric sequence, characterized in that it is genetically linked to an effective ARS sequence in Y. lipolytica.  2) Centromeric sequence according to claim 1, characterized in that it is carried by an ARS18 sequence from Y. lipolytica, 1.3 kb.  3) Centromeric sequence according to claim 1, characterized in that it is carried by an ARS68 sequence of Y. lipolytica, 2.3 kb.  4) DNA comprising a centromeric sequence according to one of claims 1 to 3.  5) DNA comprising a centromeric sequence according to claim 4, characterized in that it further comprises an adjacent inducible promoter and controlling said centromeric sequence.  6) DNA comprising a centromeric sequence according to one of claims 4 or 5, characterized in that it is a plasmid further comprising elements allowing the expression of a protein in Y. lipolytica and in that that it is kept in a strain of Y. lipolytica with a high number of copies.  7) DNA comprising a centromeric sequence according to one of claims 4 or 5, characterized in that it further comprises telomeric sequences at its ends.  8) DNA according to one of claims 4 to 7, characterized in that it further comprises a sequence belonging to the genome of Y. lipolytica, and that this sequence is modified.  9) DNA according to one of claims 4 to 8, characterized in that it comprises, in a region adjacent to the centromeric sequence, a sequence allowing its insertion into a site of a chromosome of Y. lipolytica.