Propagation of an amplifiable recombinant plasmid in Saccharomyces cerevisiae: flow cytometry studies and segregated modeling

Efficient expression of a foreign protein product by the yeastSaccharomyces cerevisiaerequires a stable recombinant vector present at a high number of copies per cell. A conditional centromere yeast plasmid was constructed which can be amplified to high copy number by a process of unequal partitioning at cell division, followed by selection for increased copy number. However, in the absence of selection pressure for plasmid amplification, copy number rapidly drops from 25 plasmids/cell to 6 plasmids/cell in less than 10 generations of growth. Copy number subsequently decreases from 6 plasmids/cell to 2 plasmids/cell over a span of 50 generations. A combination of flow cytometric measurement of copy number distributions and segregated mathematical modeling were applied to test the predictions of a conceptual model of conditional centromereplasmid propagation. Measured distributions of plasmid content displayed a significant subpopulation of cells with a copy number of 4-6, evenin a population whose mean copy number was 13.5. This type of copy number distribution was reproduced by a mathematical model which assumes that amaximum of 4-6 centromere plasmids per cell can be stably partitionedat cell division. The model also reproduces the observed biphasic kinetics of plasmid number instability. The agreement between simulation and experimental results provides support for the proposed model and demonstrates the utility of the flow cytometry/segregated modeling approach for the study of multicopy recombinant vector propagation.     

Plasmidal maintenance of composite DNA derived from polyoma related plasmid, L factor.

Recently, we reported a multicopy mammalian plasmid with a structure related to polyoma. The plasmid, named L factor, was found at a high copy number (5,000 or more per cell) in a subclone derived from mouse L cells. We attempted to utilize L factor as a plasmid vector for mammalian cells. A series of composite DNA consisting of L factor and a foreign (herpes simplex virus tk) were constructed. These DNA could be established as plasmids after transfection to several mouse cell lines, although the copy number of the re-established plasmids was considerably less than that observed for the original subclone. The composite DNA maintained the structure of the original DNA after prolonged culture and the copy number remained constant even with no selective pressure. A composite DNA, with no DNA sequence corresponding to polyoma T antigen, could also be established as a plasmid in a mouse L cell line in which polyoma T antigen is expressed. The potential use of the plasmid is discussed.     

Distribution of minichromosomes in individual Escherichia coli cells: implications for replication control

A novel method was devised to measure the number of plasmids in individual Escherichia coli cells. With this method, involving measurement of plasmid-driven expression of the green fluorescent protein gene by flow cytometry, the copy number distribution of a number of different plasmids was measured. Whereas natural plasmids had fairly narrow distributions, minichromosomes, which are plasmids replicating only from a cloned oriC copy, have a wide distribution, suggesting that there is no copy number control for minichromosomes. When the selection pressure (kanamycin concentration) for minichromosomes was increased, the copy number of minichromosomes was also increased. At up to 30 minichromosomes per host chromosome, replication and growth of the host cell was unaffected. This is evidence that there is no negative element for initiation control in oriC and that there is no incompatibility between oriC located on the chromosome and minichromosome. However, higher copy numbers led to integration of the minichromosomes at the chromosomal oriC and to initiation asynchrony of the host chromosome. At a minichromosome copy number of approximately 30, the cell's capacity for synchronous initiation is exceeded and free minichromosomes will compete out the chromosome to yield inviable cells, unless the minichromosomes are incorporated into the chromosome.     

Toxic effects of excess cloned centromeres.

Plasmids carrying a Saccharomyces cerevisiae centromere have a copy number of one or two, whereas other yeast plasmids have high copy numbers. The number of CEN plasmids per yeast cell was made artificially high by transforming cells simultaneously with several different CEN plasmids carrying different, independently selectable markers. Some host cells carried five different CEN plasmids and an average total of 13 extra copies of CEN3. Several effects were noted. The copy number of each plasmid was unexpectedly high. The plasmids were mutually unstable. Cultures contained many dead cells. The viable host cells grew more slowly than control cells, even in nonselective medium. There was a pause in the cell cycle at or just before mitosis. We conclude that an excess of centromeres is toxic and that the copy number of centromere plasmids is low partly because of selection against cells carrying multiple centromere plasmids. The toxicity may be caused by competition between the centromeres for some factor present in limiting quantities, e.g., centromere-binding proteins, microtubules, or space on the spindle pole body.     

A mathematical model for prediction of plasmid copy number and genetic stability in Escherichia coli

The design of bioreactors for genetically modified bacterial cultures would benefit from predictive models. Of particular importance is the interaction of the external environment, cell physiology, and control of plasmid copy number. We have recently developed a model based on the molecular mechanisms for control of replication of Co1E1 type plasmids. The inclusion of the plasmid model into a single-cell E. coli model allows the explicit prediction of the interaction of cell physiology and plasmid-encoded functions. The model predictions of the copy number of plasmids with the Co1E1 origin of replication carrying a variety of regulatory mutations is very close to that observed experimentally.All of the model parameters for plasmid replication control can be obtained independently and no adjustable parameters are needed for the plasmid model. In this article we discuss the model's use in predicting the effect of operating conditions on production of a protein from a plasmid encoded gene and the stability of the recombinant cells in a continuous culture.     

Kinetic analysis of the effects of plasmid multimerization on segregational instability of CoIE1 type plasmids in Escherichia coli B/r

The effect of plasmid multimerization on segregational instability was investigated using a structured, segregated model of genetically modified Escherichia coli cells. By including the multimerization of plasmids, the model can predict the proportion of each multimer in the total plasmid population. Simulation results suggest that the plasmid copy number is controlled by the total plasmid content (i.e., total number of plasmid origins) in the host cell and that multimerization reduces the total number of independent, monomeric segregation units. However, multimerization is found to have a minor effect on decreasing plasmid segregational stability for multicopy plasmids with average copy number per cell greater than about 25. Also model predictions were used to test whether or not a nonrandom plasmid distribution at cell fission could cause segregational instability. Even in the case of severely biased partitioning, plasmids whose copy number is above 45 per cell do not show significant segregational instability. The results suggest that when the ColE1-type plasmid does not encode and express any large or disruptive foreign proteins, the copy number of 45 per cell may be the threshold at which only growth rate-dependent instability is responsible for overall plasmid instability.     

Linear- and circular-plasmid copy numbers in Borrelia burgdorferi.

Borrelia burgdorferi, the Lyme disease agent, and other members of the spirochetal genus Borrelia have double-stranded linear plasmids in addition to supercoiled circular plasmids. The copy number relative to the chromosome was determined for 49- and 16-kb linear plasmids and a 27-kb circular plasmid of the type strain, B31, of B. burgdorferi. All three plasmids were present in low copy number, about one per chromosome equivalent, as determined by relative hybridizations of replicon-specific DNA probes. The low copy number of Borrelia plasmids suggests that initiation of DNA replication and partitioning are carefully controlled during the cell division cycle. The copy numbers of these three plasmids of strain B31 were unchanged after approximately 7,000 generations in continuous in vitro culture. A clone of B. burgdorferi B31 that did not contain the 16-kb linear plasmid was obtained after exposure of a culture to novobiocin, a DNA gyrase inhibitor. The plasmid-cured strain contains only one linear plasmid, the 49-kb plasmid, and thus has the smallest genome reported to date for B. burgdorferi.     

Cell cycle analysis of plasmid-containing Escherichia coli HB101 populations with flow cytometry

The relationship between cell mass and cell number dynamics for bacteria such as Escherichia coli depends on the cell cycle parameters C and D. Effects of plasmid copy number on these cell cycle parameters have been studied for Escherichia coli HB101 containing pMB1 plasmids propagated at different copy numbers ranging from 12 to 122. Determination of cell cycle and cell size parameters was accomplished using flow cytometry data on single-cell light scattering and DNA content frequency functions in conjunction with a mathematical model of cell population statistics. Two independent methods for estimating C and D intervals based on flow cytometry were developed and applied with essentially identical results. The presence of plasmids decreases the C and D periods, mean cell sizes, and initiation masses for chromosome replication by 14, 24, 38, and 18%, respectively, relative to corresponding values for plasmid-free host cells. Plasmid copy number has a negligible influence on these parameters, suggesting that host-plasmid inter actions which determine these properties are centered on the single plasmid selected for replication according to the random selection model established for ColE1-type plasmids.     

Random diffusion can account for topA-dependent suppression of partition defects in low-copy-number plasmids.

The maintenance of partition-defective (Par-) mini-P1 and mini-F plasmids was studied in topA strains of Escherichia coli, which are defective in topoisomerase I activity. The partition defects were substantially but not completely suppressed in broth-grown cultures. This suppression was not due to a large increase in copy number. However, the absolute number of copies of Par- mini-P1 plasmids per average dividing cell is sufficiently high to account for the modest stability observed if a random distribution of the copies to daughter cells is assumed. The similar number of Par- plasmid copies in wild-type cells are distributed in a considerably worse-than-random fashion. Thus, it is unnecessary to propose, as was suggested previously, that an active, par-independent pathway operates in topA strains to ensure proper segregation of the plasmids to daughter cells. Rather, it seems likely that the lack of topoisomerase I activity aids the random distribution of the partition-defective plasmids, perhaps by facilitating their separation after replication. The results of studies carried out at reduced growth rates were consistent with this view; when topA cells containing Par- mini-P1 plasmids were cultured in minimal medium, in which the copy number of the plasmids per average cell is sharply reduced, very little suppression of the partition defect was observed.     

Clustering versus random segregation of plasmids lacking a partitioning function: a plasmid paradox?

Plasmids lacking a functional partition system are randomly distributed to the daughter cells; plasmid-free daughter cells are formed with a frequency of (1/2)2n per cell and cell generation where 2n is the (average) copy number at cell division. Hence, the unit of segregation is one plasmid copy. However, plasmids form clusters in the cells. A putative solution to this potential paradox is presented: one plasmid copy at a time is recruited from the plasmid clusters to the replication factories that are located in the cell centres. Hence, replication offers the means of declustering that is necessary in a growing host population. The daughter copies diffuse freely and each copy may with equal probability end up in either of the two cell halves. In this way, the random segregation of the plasmids is coupled to replication and occurs continuously during the cell cycle, and is not linked to cell division. The unit of segregation is the plasmid copy and not the plasmid clusters. In contrast, the two daughters of a Par+ plasmid are directed in opposite directions by the plasmid-encoded partition system, thereby assuring that each daughter cell receives the plasmid.     

A model for fluctuations in the fraction of a bacterial population harboring plasmids

In some experimental situations the fraction of bacteria containing plasmids appears to repeatedly decline and increase before the plasmids are eventually lost in a gradual terminal decline. These fluctuations may not be observed unless the composition of the population is determined at various cell densities. This paper gives a mathematical model for this behavior. In the model the decreases in the fraction of cells bearing plasmids record the appearance of plasmidless offspring cells of plasmid-bearing cells. The increases occur because plasmidless cells obtain plasmids from multicopy donor cells when the cell density is high. The periodic dilution of the medium necessary to permit continued growth of the population in test tubes repeatedly alters the balance of these two factors. The gradual terminal decline in the fraction of cells bearing plasmids is the result of a slight growth rate advantage of plasmidless cells. During the entire process the average copy number of the cells decreases, and the time lag before the final decline is the time required for the average copy number to approach one.     

Copy-number control of the Escherichia coli chromosome: a plasmidologist's view

The homeostatic system that sets the copy number, and corrects over-replication and under-replication, seems to be different for chromosomes and plasmids in bacteria. Whereas plasmid replication is random in time, chromosome replication is tightly coordinated with the cell cycle such that all origins are initiated synchronously at the same cell mass per origin once per cell cycle. In this review, we propose that despite their apparent differences, the copy-number control of the Escherichia coli chromosome is similar to that of plasmids. The basic mechanism that is shared by both systems is negative-feedback control of the availability of a protein or RNA positive initiator. Superimposed on this basic mechanism are at least three systems that secure the synchronous initiation of multiple origins; however, these mechanisms are not essential for maintaining the copy number.     

Copy number control by a yeast centromere

Plasmids containing a cloned yeast (Saccharomyces cerevisiae) centromere (CEN3) in combination with a suitable DNA replication system are maintained in yeast at the low copy number typical of a chromosome. In composite plasmids containing CEN3 plus the yeast 2 mu plasmid, the CEN3 copy number control is dominant over the amplification system that normally drives the 2 mu plasmids to high copy number. The CEN3-2 mu composite plasmids are relatively stably maintained in yeast at a copy number of about one per haploid genome, and segregate through meiosis in a typical Mendelian pattern. Some of the CEN3-2 mu composite plasmids isolated from yeast contain deletions of variable size that remove the functional centromere, resulting in loss of the CEN3 control and reversion to high copy number. Formation of the CEN3 deletions requires the specialized recombination system (inverted repeat sequences and FLP gene) of the yeast 2 mu plasmid.     

Copy number and the stability of 2-micron circle-based artificial plasmids of Saccharomyces cerevisiae

The copy number and stability of artificial 2-micron circle-based plasmids have been accurately measured in [Cir+] and [Cir0] strains of Saccharomyces cerevisiae. We conclude that (i) instability and copy number vary greatly from plasmid to plasmid; (ii) instability and copy number are negatively correlated--that is, high copy number is associated with low instability; (iii) it is difficult to reconcile this variability with a strict and direct system of copy number control; (iv) instabilities are much higher than expected from random partition and the observed copy numbers: this may imply partition which is less efficient than random. Even so, (v) the partitioning of 2-micron circle-like plasmids is more efficient than that of ARS-based plasmids, which hints at the existence of a system for the (inefficient) distribution of 2-micron circles.     

The Evolution of Collective Restraint: Policing and Obedience among Non-conjugative Plasmids

The repression of competition by mechanisms of policing is now recognized as a major force in the maintenance of cooperation. General models on the evolution of policing have focused on the interplay between individual competitiveness and mutual policing, demonstrating a positive relationship between within-group diversity and levels of policing. We expand this perspective by investigating what is possibly the simplest example of reproductive policing: copy number control (CNC) among non-conjugative plasmids, a class of extra-chromosomal vertically transmitted molecular symbionts of bacteria. Through the formulation and analysis of a multi-scale dynamical model, we show that the establishment of stable reproductive restraint among plasmids requires the co-evolution of two fundamental plasmid traits: policing, through the production of plasmid-coded trans-acting replication inhibitors, and obedience, expressed as the binding affinity of plasmid-specific targets to those inhibitors. We explain the intrinsic replication instabilities that arise in the absence of policing and we show how these instabilities are resolved by the evolution of copy number control. Increasing levels of policing and obedience lead to improvements in group performance due to tighter control of local population size (plasmid copy number), delivering benefits both to plasmids, by reducing the risk of segregational loss and to the plasmid-host partnership, by increasing the rate of cell reproduction, and therefore plasmid vertical transmission.     

Control of plasmid DNA replication by iterons: no longer paradoxical

Replication origins of a family of bacterial plasmids have multiple sites, called iterons, for binding a plasmid-specific replication initiator protein. The iteron-initiator interactions are essential for plasmid replication as well as for inhibition of plasmid over-replication. The inhibition increases with plasmid copy number and eventually shuts plasmid replication off completely. The mechanism of inhibition appears to be handcuffing, the coupling of origins via iteron-bound initiators that block origin function. The probability of a trans-reaction such as handcuffing is expected to increase with plasmid copy number and diminish with increases in cell volume, explaining how the copy number can be maintained in a growing cell. Control is also exerted at the level of initiator synthesis and activation by chaperones. We propose that increases in active initiators promote initiation by overcoming handcuffing, but handcuffing dominates when the copy number reaches a threshold. Handcuffing should be ultrasensitive to copy number, as the negative control by iterons can be stringent (switch-like).     

Maintenance of pBR322-derived plasmids without functional RNAI

pBR322-derived plasmids that lack the bla gene and 40% of the gene for the replication inhibitor, RNAI, have been constructed. Since the RNAI gene totally overlaps with the gene for the replication primer, RNAII, this primer is similarly defective and also lacks its normal promoter. The primer is presumed to by synthesized either from the counter-tet promoter (plasmid pCL59) or from an inserted lacUV5 promoter (plasmid pCL59-65). Based mainly on the observation that the plasmid Rom protein, which normally assists in the RNAI/RNAII interaction, has no effect on the replication of the RNAI/RNAII-defective plasmids, we suggest that the defective RNAI is not functional while the defective RNAII primer, although less efficient, still allows plasmid replication. The defective plasmids are fully compatible with the intact parent plasmid, indicating that they do not share a common control of replication. In the absence of antibiotics, the bacteria lose the defective plasmid, beginning after 80 generations; under the same conditions, the parent plasmid is retained even after 140 generations. During exponential growth of their host, the number of defective plasmids in a culture increases exponentially with a doubling time either smaller or greater than that of the host cell growth, depending on the growth medium and, in the case of pCL59-65, on the presence or absence of lac inducer IPTG. As a result of these differences in host cell growth and plasmid replication, the plasmids are either gradually diluted out or their copy number continually increases. This shows that, without RNAI, plasmid replication is uncoupled from the host cell growth and not, as usual, adjusted to it. It also implies that the RNAI mechanism is the only means of replication control for ColE1-type plasmids that senses and adjusts the copy number; limiting host factors cannot provide a back-up control to stabilize copy numbers.