Copy number amplification of the 2 micron circle plasmid of Saccharomyces cerevisiae

The 2 micron circle is a small double stranded DNA plasmid that occurs at about 60 copies per cell in the nuclei of virtually all strains of Saccharomyces cerevisiae. The plasmid has no apparent phenotypic effect on host cells, and is the basis of many useful vectors for the transformation of yeast. Under certain circumstances, the plasmid is apparently able to replicate more than once per cell cycle; this over-replication allows the maintenance of the plasmid at high copy number. The plasmid has two inverted repeat sequences, and encodes a product that catalyses intra-molecular recombination between these two repeats. Models are proposed whereby recombination leads to copy number amplification. In particular, it is proposed that intra-molecular recombination during replication flips the orientation of one replication fork with respect to the other, so that both forks travel in the same direction around a circular monomer template, generating a large multimer from a monomer and a single initiation of replication.     

Cloning of industrial Saccharomyces 2-microns plasmid variants by in vivo site-specific recombination

Southern analyses defined several industrial Saccharomyces yeast strains with extensive 2-microns DNA polymorphism. Variants included insertions and deletions up to several hundred base pairs. To facilitate the investigation of yeast plasmid evolution we developed a novel method of cloning 2-microns plasmids by taking advantage of 2-microns circle in vivo site-specific recombination and an SMRI gene as a dominant selectable marker. This method can be applied to other organisms for the isolation of plasmid variants and provides a new approach to in vivo plasmid construction.     

The 2-micron plasmid as a nonselectable, stable, high copy number yeast vector

The endogenous 2-microns plasmid of Saccharomyces cerevisiae has been used extensively for the construction of yeast cloning and expression plasmids because it is a native yeast plasmid that is able to be maintained stably in cells at high copy number. Almost invariably, these plasmid constructs, containing some or all 2-microns sequences, exhibit copy number levels lower than 2-microns and are maintained stably only under selective conditions. We were interested in determining if there was a means by which 2-microns could be utilized for vector construction, without forfeiting either copy number or nonselective stability. We identified sites in the 2-microns plasmid that could be used for the insertion of genetic sequences without disrupting 2-microns coding elements and then assessed subsequent plasmid constructs for stability and copy number in vivo. We demonstrate the utility of a previously described 2-microns recombination chimera, pBH-2L, for the manipulation and transformation of 2-microns as a pure yeast plasmid vector. We show that the HpaI site near the STB element in the 2-microns plasmid can be utilized to clone yeast DNA of at least 3.9 kb with no loss of plasmid stability. Additionally, the copy number of these constructs is as high as levels reported for the endogenous 2-microns.     

Use of interplasmid recombination to generate stable selectable markers for yeast transformation: application to studies of actin gene control

A plasmid recombination system has been developed that relies upon interplasmid exchanges for yeast cell viability. Two types of plasmids, one carrying the LEU2 allele inserted within yeast actin gene sequences and the other carrying 2-microns plasmid DNA and an intact actin gene, were constructed. Neither plasmid alone yielded transformants in the haploid Leu- strain AH22, but when cotransformed, a number of colonies were obtained. Southern blot analysis revealed that transformants arose because of recombination events within the homologous actin sequences that transferred the LEU2 gene to the actin gene on the 2-microns plasmid. The recombinant plasmids could be recovered, and sequence analysis of one recombination site revealed that the exchange event was faithful at the nucleotide level. The resulting recombinant plasmids carried a defective actin gene and presumably arose because of a double-crossover event. Deletion mutations that prevented actin gene expression on one donor plasmid enabled the recovery at a high frequency of transformants resulting primarily from single-crossover events between the two plasmids. This was presumably because such events no longer generated an intact actin gene on a multicopy plasmid. Infrequently a transformant from a plasmid with an intact gene was recovered, but in these cases the plasmid was not present in multiple copies. These cells exhibited a slower growth rate, and Northern blot analysis revealed an elevated level of actin mRNA.     

Use of a cis-acting mutation to study the role of FLP-mediated recombination in the maintenance of native yeast 2μm plasmids

The 2μm plasmid encodes a mechanism that ensures the partitioning of the plasmid at cell division. Little is known about the detailed mechanism of this partitioning system; for example, is there equal or unequal distribution of the plasmid molecules at mitosis? The plasmid also encodes a site-specific recombination system that is thought to be involved in plasmid copy-number amplification, although to date there has been no direct evidence that the recombination process itself is important for maintenance. We have identified a natural 2μm variant that has a cis-acting mutation in the FLP-mediated recombination system. We show that this plasmid is unable to amplify in vivo. Our results demonstrate that the average copy number per cell is not affected for the mutant but there is a large clonal variation. This is a direct demonstration that plasmid partitioning results in an unequal distribution of plasmids and that FLP-mediated amplification compensates for this and therefore has an important role in maintenance.     

The FLP protein of the 2-micron plasmid of yeast. Purification of the protein from Escherichia coli cells expressing the cloned FLP gene

Most laboratory strains of the yeast Saccharomyces cerevisiae contain many copies of an autonomously replicating plasmid called 2-micron circle DNA. This plasmid codes for a site-specific recombinase, the FLP protein which promotes recombination across two 599-base pair inverted repeats of the plasmid DNA. We have cloned the FLP gene under the control of a strong Escherichia coli promoter and have hyperproduced the protein in that organism. Cell-free extracts from this source promote highly efficient site-specific recombination in vitro and we have used this activity to purify the FLP protein substantially. The enzyme acts efficiently on circular and linear substrates and requires only monovalent or divalent cations for activity.     

Recombination within the yeast plasmid 2mu circle is site-specific

The multicopy yeast plasmid, 2mu circle, encodes a specialized recombination system. It contains two regions, each 599 bp in length, that are precise inverted repeats of each other and between which recombination occurs readily. In addition, this recombination requires the product of a 2mu circle gene, designated FLP. By examining the products of FLP-mediated recombination of plasmids containing single insertions within one of the repeated regions, we show that this recombination occurs only at a specific site within the repeat. This result was confirmed from analysis of the ability of plasmids containing various deletions within one of the repeated regions to serve as substrates for FLP-mediated recombination. These experiments limit the recombination site to a sequence of less than 65 bp. In addition, by mutational analysis of the recombination potential of a hybrid plasmid containing the entire 2mu circle genome, we have shown that FLP is only the 2mu circle gene necessary for this site-specific recombination. Finally, we describe a sensitive assay for recombination between the repeated sequences of 2mu circle; using it, we demonstrate that even in the absence of FLP gene product, recombination between the repeats occurs at a low but detectable level during meiosis.     

Recombinational instability of a chimeric plasmid in Saccharomyces cerevisiae.

Wild-type strains of Saccharomyces cerevisiae exhibit mitotic recombination between the chimeric plasmid TLC-1 and the endogenous 2mu circle that involves sequence homologies between the two plasmids that are not acted on by the 2mu circle site-specific recombination system. This generalized recombination can be detected because it separates the LEU2 and CAN1 markers of TLC-1 from each other through the formation of a plasmid containing only the S. cerevisiae LEU2 region and the 2mu circle. This derivative plasmid is maintained more stably during vegetative growth than TLC-1, and strains which carry it frequently lose the endogenous 2mu circle. Therefore, TLC-1 can provide a convenient selection for [cir0] cells. Formation of this new plasmid is greatly reduced, but not eliminated, in strains containing the rad52-1 mutation. This indicates that generalized mitotic recombination between plasmid sequences utilizes functions required for chromosomal recombination in S. cerevisiae.     

Mechanism of high-copy-number integration of pMIRY-type vectors into the ribosomal DNA of Saccharomyces cerevisiae

Targeted integration of the yeast plasmid pMIRY2 into the ribosomal DNA (rDNA) of Saccharomyces cerevisiae by homologous recombination results in transformants carrying 100-200 copies of the plasmid per cell which are stably maintained over a large number of generations [Lopes et al., Gene 79 (1989) 199-206]. These properties make pMIRY2 an attractive vector for high-level production of (heterologous) proteins by yeast cells. We have investigated the mechanism underlying high-copy-number (hcn) integration of pMIRY-type plasmids and show that either targeting to a location outside the rDNA locus or use of the wild-type LEU2, instead of the deficient LEU2d gene, as selection marker reduces the copy number to the low value characteristic of standard integrating (YIp-type) yeast plasmids. Further experiments demonstrate that the hcn of pMIRY-type plasmids is achieved by amplification of a small number of copies initially integrated into the rDNA locus. Amplification depends upon the strong selection pressure created by the extremely low expression of the deficient LEU2d gene, but not on the presence of this gene per se. The hcn integration also occurs when either the TRP1 or URA3 gene is used as the selection marker, provided expression of the marker gene is severely curtailed, e.g., by removal of most of its 5'-flanking region.     

Molecular clocks reduce plasmid loss rates: the R1 case

Plasmids control their replication so that the replication frequency per plasmid copy responds to the number of plasmid copies per cell. High sensitivity amplification in replication response to copy number deviations generally reduces variation in copy numbers between different single cells, thereby reducing the plasmid loss rate in a cell population. However, experiments show that plasmid R1 has a gradual, insensitive replication control predicting considerable copy number variation between single cells. The critical step in R1 copy number control is regulation of synthesis of a rate-limiting cis-acting replication protein, RepA. De novo synthesis of a large number of RepA molecules is required for replication, suggesting that copy number control is exercised at multiple steps. In this theoretical kinetic study we analyse R1 multistep copy number control and show that it results in the insensitive replication response found experimentally but that it at the same time effectively prohibits the existence of only one plasmid copy in a dividing cell. In combination with the partition system of R1, this can lead to very high segregational stability. The R1 control mechanism is compared to the different multistep copy number control of plasmid ColE1 that is based on conventional sensitivity amplification. This implies that while copy number control for ColE1 efficiently corrects for fluctuations that have already occurred, R1 copy number control prevents their emergence in cells that by chance start their cycle with only one plasmid copy. We also discuss how regular, clock-like, behaviour of single plasmid copies becomes hidden in experiments probing collective properties of a population of plasmid copies because the individual copies are out of phase. The model is formulated using master equations, taking a stochastic approach to regulation, but the mathematical formalism is kept to a minimum and the model is simplified to its bare essence. This simplicity makes it possible to extend the analysis to other replicons with similar design principles.     

Genetic properties of chromosomally integrated 2 mu plasmid DNA in yeast

We obtained strains of yeast with large segments of 2 mu plasmid DNA integrated at several chromosomal locations by selecting genetically for recombination between a chromosomal sequence carried on a 2 mu-circle-containing hybrid plasmid and a homologous sequence on the chromosome. In all diploids examined, the presence of 2 mu circle sequences causes a marked instability of the chromosome into which the 2 mu DNA is inserted. Although in some cases the loss of genetic markers is due to physical loss of the entire chromosome, in most cases the loss of markers appears to be due to a mitotic homozygotization of markers: the allelic information from the homologous chromosome replaces the information distal to the integrated 2 mu DNA. The instability caused by integrated 2 mu DNA sequences requires the activity of the specialized site-specific recombination system encoded by the 2 mu plasmid. We propose that the presence of integrated 2 mu DNA allows efficient integration of additional copies of the intact 2 mu plasmid by the action of the plasmid-coded special recombination system. Unequal sister-strand exchanges within the inverted repetition would result in the formation of dicentric chromosomes whose breakage during mitosis might begin a cycle analogous to the breakage-fusion-bridge cycle described many years ago in maize.     

In Vivo Cloning of lac Genes in Streptococcus lactis ML3

The isolation and characterization of a Streptococcus lactis ML3 strain which possessed a recombinant lactose plasmid is described. The recombination events generating this plasmid occurred in vivo in a recombination-deficient strain and appeared to be mediated by transposition events. Restriction mapping revealed that the recombinant plasmid, pDA0307, contained a region of the lactose plasmid, pSK08, linked to another resident plasmid, pSK07. Copy number determinations indicated that the lac genes were present at approximately 20 copies per cell in pDA0307, whereas the lac genes are normally present at approximately 10 copies per cell in pSK08. The strain containing pDA0307 displayed a 21 to 54% increase in the expression of the Lac enzyme phospho-beta-d-galactosidase. However, the strain containing pDA0307 both grew and produced lactic acid in milk at rates identical to that of a strain containing pSK08. This result suggests that lac gene dosage of plasmid-linked lac genes was not limiting the rate at which these derivatives of S. lactis ML3 fermented milk.     

Roles of the 2 microns gene products in stable maintenance of the 2 microns plasmid of Saccharomyces cerevisiae.

We have examined the replication and segregation of the Saccharomyces cerevisiae 2 microns circle. The amplification of the plasmid at low copy numbers requires site-specific recombination between the 2 microns inverted repeat sequences catalyzed by the plasmid-encoded FLP gene. No other 2 microns gene products are required. The overexpression of FLP in a strain carrying endogenous 2 microns leads to uncontrolled plasmid replication, longer cell cycles, and cell death. Two different assays show that the level of Flp activity decreases with increasing 2 microns copy number. This regulation requires the products of the REP1 and REP2 genes. These gene products also act together to ensure that 2 microns molecules are randomly segregated between mother and daughter cells at cell division.     

Enhanced error-prone RCA mutagenesis by concatemer resolution

Error-prone rolling circle amplification (RCA) is a promising alternative to error-prone PCR for random mutagenesis. The main disadvantage of error-prone RCA is the low transformation efficiency of the DNA concatemer produced in the amplification reaction. We improved the method by introducing loxP recombination site of bacteriophage P1 Cre recombinase into the target plasmid and reducing the concatemer by Cre recombinase to plasmid-sized units, increasing the number of transformants 50-fold in non-error-prone and 13-fold in error-prone conditions. The efficiency improvement was verified by obtaining 115 ± 57 ceftazidime resistant colonies per recombined RCA reaction from randomly mutated TEM-1 β-lactamase gene library whereas only 9 ± 11 colonies were gained without recombination. Supplementation of the error-prone RCA with Cre/loxP recombination is a simple and useful tool to increase the transformable library size.     

The mcm2 mutation of yeast affects replication, rather than segregation or amplification of the two micron plasmid.

We have studied the maintenance of the endogenous two micron (2 mu) plasmid in a strain of yeast carrying the nuclear mutation mcm2. This mutation, earlier shown to affect the maintenance of yeast minichromosomes in an ARS-dependent manner, also affected the copy number of the 2 mu plasmid. The effect was more pronounced at 35 degrees C leading to the elimination of the plasmid from the cells cultured at this temperature. The mutant cells could be efficiently cured of the circle by transformation with 2 mu ORI-carrying hybrid vectors, an observation consistent with the low copy number of the endogenous plasmid. A chromosomal revertant of this mutant for another ARS(ARS1) was found also to confer stability on the 2 mu ORI-carrying minichromosomes and had elevated levels of the endogenous plasmid. The mutation neither affected the segregation nor the amplification process mediated by site-specific recombination at FRT sites requiring the FLP gene-encoded protein action. ARS131C, an ARS that was unaffected in the mutant at 25 degrees C, could elevate the copy number of a 2 mu hybrid vector in the mutant cells. In view of these results, some aspects of segregation and copy number control of the endogeneous plasmid have been discussed. We propose that the mutation impairs the 2 mu ORI function, leading to its loss.     

Permanent cell lines that show temperature-dependent expression of adenovirus virus-associated RNA.

Temperature-sensitive COS cells, clone E540, have been stably transformed at a restrictive temperature with plasmid pVA1, which contains the adenovirus type 5 virus-associated (VA) genes in addition to the Neor marker. Transformed cell clones, named EVA cells, contained adenovirus DNA in an integrated form while grown at restrictive temperature but accumulated up to 100 to 200 copies of the input plasmid per cell after temperature shift down. Concomitant with this gene amplification, an accumulation of VA RNA was observed, reaching average concentrations of 10(4) to 10(5) copies per cell. The VA RNA synthesized in EVA cells is functional, as judged by inhibition of in vitro eucaryotic initiation factor-2 phosphorylation and enhancement of reporter gene expression. These EVA cell lines may be of use to study the mechanism of VA RNA function in the absence of adenovirus infection.     

Expression of a bacterial gene for guanine synthesis inserted into colicin E1 factor by an in vitro recombination.

Regulation of expression of a bacterial guaA gene inserted into colicin E1 DNA by an in vitro recombination was studied under various growth conditions. In Escherichia coli K-12 cells that carried this hybrid ColEl plasmid the level of guaA enzyme activity was not regulated by the concentration of guanine in the medium, but by the number of plasmid DNA copies. The optimal conditions for amplifying the guaA gene product by chloramphenicol treatment were determined. The level of guaA enzyme activity found under the optimal conditions was about 37 times that in extracts of wild-type E. coli cultured in guanine-free medium. The properties of the promoter for the guaA gene and applicability of this hybrid ColEl plasmid for amplification of various gene products were discussed.     

Determinant of resistance to kanamycin in Streptomycin rimosus: amplification in the chromosome and reversible genetic instability

The study on the kanamycin resistance determinant (Kanr) in an oxytetracycline--producing strain of S. rimosus showed that it was capable of amplifying in the chromosome during selection for increasing the antibiotic resistance level. The amplification of the DNA fragment with a molecular weight of 10.3 MDa containing Kanr amounted to 300 copies per genome, which resulted in a more than 1000-fold increase in kanamycin resistance level. Cloning of the Kanr determinant on plasmid SLP1.2 in S. lividans strain 66 was performed. In Streptomyces lividans strain 66 the Kanr determinant preserved the capacity for amplification in the hybrid plasmid pSU10 integrated into the chromosome. The Kanr determinant in the strains of S. rimosus and S. lividans was characterized by transfers Kanr in equilibrium Kans with a frequency of 1 X 10(-3). It was shown that the mutation in S. lividans strain 66 resulting in phenotype Kans was not connected with the structural Kanr gene on plasmid pSU10 but was localized on the chromosome. Phenotype Kans was promoted by a decrease in the number of the copies of the regulatory genetic element designated RES1. The reverse to phenotype Kanr might be due to one of the following events: amplification to the initial level of RES1 and amplification up to 200 copies per the genome of the hybrid plasmid pSU10 containing the Kanr determinant. Amplification of the Kanr determinant with preserved initial level of RES1 element resulted in a more than 1000 times increase in the resistance level.     

Chromosomal Assignment of Mutations by Specific Chromosome Loss in the Yeast Saccharomyces Cerevisiae

Yeast 2-microns plasmids were integrated near the centromere of a different chromosome in each of 16 cir0 mapping strains of Saccharomyces cerevisiae. The specific chromosomes containing the integrated 2-microns plasmid DNA were lost at a high frequency after crossing the cir0 strains to cir+ strains. A recessive mutation in a cir+ strain can then be easily assigned to its chromosome using this set of mapping strains, since the phenotype of the recessive mutation will be manifested only in diploids having the integrated 2-microns plasmid and the unmapped mutation on homologous chromosomes.     

Induction of DNA amplification in the Bacillus subtilis chromosome.

A system allowing the induction of DNA amplification in Bacillus subtilis was developed, based on a thermosensitive plasmid, pE194, stably integrated in the bacterial chromosome. An amplification unit, comprising an antibiotic resistance marker flanked by directly repeated sequences, was placed next to the integrated plasmid. Activation of pE194 replication led to DNA amplification. Two different amplification processes appeared to take place: one increased the copy number of all sequences in the vicinity of the integrated plasmid and was possibly of the onion skin type, while the other increased the copy number of the amplification unit only and generated long arrays of amplification units. These arrays were purified and shown to consist mainly of directly repeated amplification units but to also contain non-linear regions, such as replication forks and recombination intermediates. They were attached to the chromosome at one end only, and were, in general, not stably inherited, which suggests that they are early amplification intermediates. Longer arrays were detected before the shorter ones during amplification. When the parental amplification unit contained repeats which differed by a restriction site the arrays which derived thereof contained in a majority of cases only a single type of repeat. We propose that the amplified DNA is generated by rolling circle replication, and that such a process might underlie a number of amplification events.