Recombinant protein synthesis and plasmid instability in continuous cultures of Escherichia coli JM103 harboring a high copy number plasmid

Escherichia coli JM103[pUC8] was employed as a model to investigate the behavior of a recombinant microbial system harboring a plasmid at high copy numbers. Experiments with batch and continuous cultures of recombinant and plasmid-free cells were conducted in a well-controlled bio-reactor. In batch experiments, plasmid copy number varied typically from an average of 500 during the exponential growth phase to as high as 1250 during the stationary phase. While the segregational plasmid instability was negligible in batch experiments, severe segregational instability occurred in continuous experiments conducted over a range of dilution rates, resulting in complete loss of plasmid-bearing cells from the continuous cultures within few residence times after transition to continuous operation. The profound differences in the specific growth rates and mass yields of the plasmid-free and plasmid-bearing cells resulting from the extra metabolic burden on the plasmid-bearing cells mainly due to excessive plasmid DNA content was the major cause for the plasmid instability. Plasmid multirnerization was detected in batch and continuous cultures and was found to have significant influence on the effective copy number and was partially responsible for the severe segregational instability in continuous cultures. A quasi-steady state representative of plasmid-bearing cells was established in the initial portion of each continuous culture experiment. Due to the profound growth rate differential between the two types of cells, transients of considerable duration were observed in each continuous culture experiment (initiated with a pure culture of plasmid bearing cells) following the slow accumulation of plasmid-free cells near the end of the quasi-steady state. Significant variations in various culture parameters (including a rapid decline in the plasmid-bearing fraction of the total cell population) occurred during this period, leading ultimately to a steady state for a culture dominated entirely by plasmid-free cells. In continuous cultures, plasmid copy number during the quasi-steady states increased with decreasing dilution rate from 50 (at 0.409 h(-1)) to 941 (at 0.233 h(-1)). Production of the plasmid-encoded protein (beta-lactamase) in these experiments was maximized at an intermediate dilution rate, corresponding to an optimum copy number of about 450. A similar optimum copy number was observed in batch cultures. Significant excretion of beta-lactamase was observed at both low and high dilution rates.     

Growth kinetics of Eschericia coli containing temperature-sensitive plasmid pOU140

The growth dynamics of Escherichia coli with the temperature-sensitive plasmid, pOU140, were examined. Recombinant cells exhibited nearly identical kinetic behavior to host cells at low culture temperatures and low copy numbers. However, at higher temperatures, in which the copy number was significantly increased, the recombinant cells showed decreased stability along with lower growth rates and substrate yields as compared to the host. Furthermore, the production of a constitutive cloned-gene protein was shown to increase with temperature in an Arrhenius fashion when culture temperature was varied between 38 and 42 degrees C. These results suggest that temperature can be used to quantitatively control the production of a desirable plasmid-coded gene product.     

Impact of targeted vector design on Co/E1 plasmid replication

The demands for recombinant proteins, in addition to plasmid DNA, for therapeutic use are steadily increasing. Bacterial fermentation processes have long been and still are the major tool for production of these molecules. The key objective of process optimization is to attain a high yield of the required quality, which is determined, to a large extent, by plasmid replication rates, metabolic capacity and the properties of the specific gene construct. When high copy number plasmids are used, the metabolic capacity of the host cell is often overstrained and efficient protein production is impaired. The plasmid copy number is the key parameter in the exploitation of the host cell, and can be maximized by optimal control of the flux ratios between biosynthesis of host cell proteins and recombinant proteins.     

Effects of recombinant plasmid content on growth properties and cloned gene product formation in Escherichia coli

Plasmid-host cell interactions have been investigated experimentally using Escherichia coli HB101, plasmid RSF1050 which contains the origin of replication of pMB1, and four other closely related copy number mutant plasmids. Growth characteristics of these recombinant strains and beta-lactamase activity expressed from a plasmid gene were investigated in Luria broth (LB) and in minimal medium (M9) containing in some cases casamino acids or different concentrations of alpha-methylglucoside, a competitive inhibitor of glucose transport. Maximum specific growth rates in LB and minimal media were reduced for increasing plasmid content per cell. Plasmid copy number increased when specific growth rate was reduced by changing medium composition. Growth rates of high copy number strains were less sensitive to alpha-methylglucoside than lower copy number strains and the plasmidfree host. The overall efficiency of plasmid gene expression, measured as the ratio of beta-lactamase specific activity to plasmid content, decreased significantly with increasing plasmid content in LB medium.     

Stability of plasmid vector plJ303 in Streptomyces lividans TK24 during laboratory-scale fermentations

Plasmid plJ303 stability in Streptomyces lividans cultures has been studied by measuring plasmid copy number under various growth conditions. An increase in mean plasmid copy number was normally seen during early rapid growth in both shaken culture and stirred vessel fermentations at 28 degrees C. Maximum copy numbers were consistently attained in early stationary phase followed by a decline (of variable amount) upon further incubation. The imposition of environmental stress (high growth temperature, i.e., 37 degrees C, and low dissolved oxygen tension, i.e., <5% air saturation) led to a plasmid copy number of zero and a 50% reduction, respectively. Interestingly, the relative proportions of plasmid topoisomers changed with time since progressively more supercoiled forms were observed throughout the stationary phase. Plasmid dimers were also observed in some cultures, and no evidence of structural plasmid instability was found. In general, this host-vector system seemed remarkably stable under normal growth conditions. However, copious organic acid production by the host was observed and was thought to be undesirable for good heterologous gene expression of a secreted protein. (c) 1993 John Wiley & Sons, Inc.     

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.     

A novel structured kinetic modeling approach for the analysis of plasmid instability in recombinant bacterial cultures

The instantaneous specific growth rate of a recombinant bacterial culture is directly calculated using a simple structured kinetic modeling approach. Foreign plasmid replication and foreign protein expression represent metabolic burdens to the host cell. The individual effects of these plasmid-mediated activities on the growth rate of plasmid-bearing cells are estimated separately. The dynamic and steady state simulations of the model equations show remarkable agreement with widely observed experimental trends in plasmid copy number and foreign protein content. The model provides an important tool for understanding and controlling plasmid instability in recombinant bacterial fermentations. The modeling framework employed here is suitable for studying the metabolism and growth of a variety of microbial cultures.     

Sociobiological Control of Plasmid Copy Number in Bacteria

All genes critical for plasmid replication regulation are located on the plasmid rather than on the host chromosome. It is possible therefore that there can be copy-up “cheater” mutants. In spite of this possibility, low copy number plasmids appear to exist stably in host populations. We examined this paradox using a multilevel selection model. Simulations showed that, a slightly higher copy number mutant could out-compete the wild type. Consequently, another mutant with still higher copy number could invade the first invader. However, the realized benefit of increasing intra-host fitness was saturating whereas that of inter-host fitness was exponential. As a result, above a threshold, intra-host selection was overcompensated by inter-host selection and the low copy number wild type plasmid could back invade a very high copy number plasmid. This led to a rock-paper-scissor (RPS) like situation that allowed the coexistence of plasmids with varied copy numbers. Furthermore, another type of cheater that had lost the genes required for conjugation but could hitchhike on a conjugal plasmid, could further reduce the advantage of copy-up mutants. These sociobiological interactions may compliment molecular mechanisms of replication regulation in stabilizing the copy numbers.     

Cloning and maintenance of the housefly sodium channel gene using low copy number vector and two sequential host strains

In spite of the long history of recombinant DNA technology, some genes have not been successfully cloned in Escherichia coli. This is probably due to the toxic effects of the expressed foreign gene product on E. coli. In initial attempts to clone the full-length Vssc1 voltage-sensitive sodium channel α-subunit gene from houseflies, we used one of the most popular vectors and hosts but were unable to retrieve any intact clone. By using two vectors with different copy numbers and two alternate E. coli host strains, we found that the combined use of a low copy number vector (pALTER-1) and an E. coli host strain that suppresses plasmid replication (ABLE-K) is essential to obtain intact full-length Vssc1 clone. However, since the ABLE-K strain was not a suitable host for the long-term maintenance of Vssc1 gene due to its recombination-positive genotype, it was necessary to transfer the Vssc1 plasmid from the primary host to a secondary host with a recombination-minus genotype (Stbl2) to minimize the chances of deletion or rearrangement. We believe that this cloning strategy, with a low copy number vector and the sequential use of two E. coli strains, will be also applicable for the cloning of other toxic genes.     

Plasmid presence changes the relative levels of many host cell proteins and ribosome components in recombinant Escherichia coli

Relative levels of many individual proteins in Escherichia coli HB101 strains with 0, 37, 56, and 240 plasmids per chromosome were determined by computer image analysis of two-dimensional gel electrophoresis patterns. The plasmids investigated had very similar sequences except for small domains encoding the represser of plasmid replication. At the intermediate plasmid copy number of 56, levels of several of the TCA cycle enzymes (oxoglutarate dehydrogenase complex, succinate thiokinase, and succinate dehydrogenase) as well as in aspartate transcarbamoylase increased. At a plasmid copy number of 240, higher amounts of PEP carboxylase as well as several of the heat shock proteins were observed. Furthermore, at high plasmid levels, significant decreases occurred in growth rate, pyruvate kinase I, pyruvate dehydrogenase complex, unadenylated glutamine synthetase, aspartate transcarbamoylase as well as in several of the proteins involved in translation. Decreases in ribosome content as well as in the free 30S and 50S ribosomal subunit pool fractions were also observed in separate analyses. These results indicate that recombinant DNA manipulations can cause major alterations in numerous host cell properties which could significantly influence cloned protein production or metabolic engineering endeavors.     

Development of small high-copy-number plasmid vectors for gene expression in Caulobacter crescentus

Caulobacter crescentus is a bacterium with a distinctive life cycle and so it is studied as a cell development model. In addition, we have adapted this bacterium for recombinant protein production and display based on the crystalline surface protein (S)-layer and its C-terminal secretion signal. We report here the development of small, high-copy-number plasmid vectors and methods for producing an obligate expression host. The vectors are based on a narrow-host-range colE1-replicon-based plasmid commonly used in Escherichia coli, to which was added the replication origin of the IncQ plasmid RSF1010. C. crescentus strains were modified to enable plasmid replication by introduction of the RSF1010 repBAC genes at the recA locus. The small (4.0-4.5 kb) plasmids were in high copy numbers in both C. crescentus and E. coli and amenable to rapid methods for plasmid isolation and DNA sequencing. The method for introducing repBAC is suitable for other C. crescentus strains or any bacterium with an adequately homologous recA gene. Application of the vector for protein expression, based on the type I secretion system of the S-layer protein, when compared to constructs in broad-host-range plasmids, resulted in reduced time and steps required from clone construction to recombinant protein recovery and increased protein yield.     

Evaluation of strains derived from Escherichia coli W as hosts for the expression of penicillin G-acylase-encoding gene cloned on the recombinant plasmid pKA18

The potential for production of penicillin G-acylase (PGA), encoded by the chromosomal genepga ⁱ, of four strains belonging to a genealogical line derived from the strainEscherichia coli W, was evaluated in a medium with and without the inducer phenylacetic acid (PA). These strains were used as hosts of the recombinant plasmid pKA18, in which the structural genepga ^c isolated from the strain RE3, the best host strain of a line giving the highest production, was cloned. The presence of the inducer reduced the copy number of the plasmid in all recombinant strains. Only in recombinant strain RE3 (pKA18) the reduction of the gene dosage ofpga ^c resulted also in the reduction of the amount of PGA synthesized by the cells. The reduced activity of the cells did not result from a segregation of plasmid-free clones. Also the growth rate was decreased by 20 and 40% in the host and recombinant strains, respectively. The host strain RE3 showing the highest production of PGA was also the best host of the recombinant plasmid in terms of the segregational stability and copy number (198 copies per chromosome). The recombinant strain RE3 (pKA18) also provided the highest production of PGA.     

Host-plasmid interactions in Saccharomyces cerevisiae: effect of host ploidy on plasmid stability and copy number

The segregational stability of two chimaeric plasmids has been examined in an isogenic series of haploid, diploid and tetraploid strains of Saccharomyces cerevisiae, constructed by transformation-associated spheroplast fusion. For the highly unstable, ARS-based plasmid YRp7M, a significant increase in its segregational stability was observed with increasing ploidy, while the relatively stable, 2 microns-based plasmid pMA3a showed only a small increase in stability in strains of higher ploidy. The copy number of both pMA3a and the endogenous 2 microns plasmid increased in proportion with the host cell ploidy, while the copy number of TRp7M was increased in the higher ploidy strains but did not correlate with ploidy. These results suggest that the copy numbers of both the 2 microns plasmid and a plasmid derived from it are controlled by a nuclear gene and that, in addition, there are 2 microns sequences, other than those required for the FLP-mediated recombination system, that play a role in maintaining copy number.     

Evolutionary competitiveness of two natural variants of the IncQ-like plasmids, pRAS3.1 and pRAS3.2

Plasmids pRAS3.1 and pRAS3.2 are natural variants of the IncQ-2 plasmid family, that except for two differences, have identical plasmid backbones. Plasmid pRAS3.1 has four 22-bp iterons in its oriV region, while pRAS3.2 has only three 6-bp repeats and pRAS3.1 has five 6-bp repeats in the promoter region of the mobB-mobA/repB genes and pRAS3.2 has only four. In previous work, we showed that the overall effect of these differences was that when the plasmid was in an Escherichia coli host, the copy numbers of pRAS3.1 and pRAS3.2 were approximately 41 and 30, respectively. As pRAS3.1 and pRAS3.2 are likely to have arisen from the same ancestor, we addressed the question of whether one of the variants had an evolutionary advantage over the other. By constructing a set of identical plasmids with the number of 22-bp iterons varying from three to seven, it was found that plasmids with four or five iterons displaced plasmids with three iterons even though they had lower copy numbers. Furthermore, the metabolic load that the plasmids placed on E. coli host cells compared with plasmid-free cells increased with copy number from 10.9% at a copy number of 59 to 2.6% at a copy number of 15. Plasmid pRAS3.1 with four 22-bp iterons was able to displace pRAS3.2 with three iterons when both were coresident in the same host. However, the lower-copy-number pRAS3.2 placed 2.8% less of a metabolic burden on an E. coli host population, and therefore, pRAS3.2 has a competitive advantage over pRAS3.1 at the population level, as pRAS3.2-containing cells would be expected to outgrow pRAS3.1-containing cells.     

Reciprocal intrapool variation in plasmid copy numbers: a characteristic of segregational incompatibility

An experimental analysis of the concept that incompatible plasmids occupy a common intracellular pool from which copies are drawn at random for replication and assortment is presented. Intrapool variations in an incompatible heteroplasmid strain are inevitable and it is shown that these variations can be exploited by differential selection to amplify one plasmid at the expense of the other. Constant overall copy number is demonstrated for isogenic wild-type replicons and also for isogenic copy mutants whose copy numbers are so great that segregational incompatibility cannot be measured. In the test system used, that of the Staphylococcus aureus plasmid pT181, the rate of replication is probably determined by the availability of a trans-active initiator protein, RepC. In heteroplasmid strains containing wild-type and dominant copy mutant plasmids, although intrapool variation occurs, the total copy number is not constant but varies as a consequence of selection for or against the mutant plasmid. This is because all of the RepC is synthesized from the mutant plasmid (the wild-type is hyper-repressed) and therefore the selection affects the supply of RepC at the same time that it affects the copy number of the plasmid. None of these effects are seen with single plasmids or with compatible pairs.     

Investigation of the instability of plasmids directing the expression of Met-prochymosin in Escherichia coli

The causes of the instability of a multicopy plasmid, pCT70, which directs the expression of calf prochymosin in Escherichia coli, were investigated. Plasmid pAT153 and its derivative, pCT54, were stable for more than 90 generations in continuous culture with glucose limitation. The multicopy plasmid pCT66, which expressed very low levels of prochymosin due to poor translational efficiency, and low copy number plasmids which efficiently expressed the prochymosin gene, were also stable. These results indicated that high level translation of the recombinant gene was the cause of the instability of pCT70. The maximum specific growth rate of E. coli(pCT70) was reduced by 30% compared with E. coli(pCT66). To fulfil the requirements of a production system, a dual origin plasmid with controllable copy number was developed. Both this plasmid (pMG165) and a derivative which contained the prochymosin gene (pMG168) were stable when maintained at low copy number. When the copy number of plasmid pMG168 was increased by putting replication under the control of the lambda PR promoter and the cI857 temperature sensitive repressor, expression of prochymosin was achieved. This strategy enables large-scale production of prochymosin without the need for antibiotic selection or other methods of preventing plasmid loss.     

Modulating heterologous protein production in yeast: the applicability of truncated auxotrophic markers

The use of auxotrophic Saccharomyces cerevisiae strains for improved production of a heterologous protein was examined. Two different marker genes were investigated, encoding key enzymes in the metabolic pathways for amino acid (LEU2) and pyrimidine (URA3) biosynthesis, respectively. Expression plasmids, carrying the partly defective selection markers LEU2d and URA3d, were constructed. Two CEN.PK-derived strains were chosen and insulin analogue precursor was selected as a model protein. Different truncations of the LEU2 and URA3 promoters were used as the mean to titrate the plasmid copy number and thus the recombinant gene dosage in order to improve insulin productivity. Experiments were initially carried out in batch mode to examine the stability of yeast transformants and to select high yielding mutants. Next, chemostat cultivations were run at high cell density to address industrial applicability and long-term expression stability of the transformants. We found that the choice of auxotrophic marker is crucial for developing a yeast expression system with stable heterologous protein production. The incremental truncation of the URA3 promoter led to higher plasmid copy numbers and IAP yields, whereas the truncation of the LEU2 promoter caused low plasmid stability. We show that the modification of the level of the recombinant gene dosage by varying the degree of promoter truncation can be a strong tool for optimization of productivity. The application of the URA3d-based expression systems showed a high potential for industrial protein production and for further academic studies.     

Escherichia coli plasmid production in fermenter

Escherichia coli harboring a recombinant plasmid was grown in a fermenter to study the effects of selected process parameters on the growth of the microbe and on plasmid amplification with chloramphenicol treatment. Eighteen fermentations were carried out according to a statistical experimental design in which the fermentation temperature, pH, and turbidity of culture at the onset of plasmid amplification were the selected independent process variables. Static regression models describing the process were derived from the experimental results. It turned out that recombinant plasmid copy numbers could be influenced by controlling fermentation temperature and pH. The maximal copy number during bacterial growth phase and the optimal plasmid production were found to require fermentation conditions different from those needed for optimal bacterial growth and cell division. The conditions also differed significantly from those routinely used in research laboratories for plasmid preparation. The chloramphenicol treatment increased the plasmid copy number compared with chromosome numbers up to fivefold. Some of the data suggest that under certain conditions the number of chromosome molecules in E. coli cells may rise during the plasmid amplification stage. Statistical experimental design, a nucleic acid sandwich hybridization technique for plasmid quantification, and regression models proved to be useful in this study.     

A comparison of mathematical model predictions to experimental measurements for growth and recombinant protein production in induced cultures of Escherichia coli

Recombinant cell growth and protein synthesis by a recombinant Escherichia coli under various inducing conditions are compared to the predictions of a mathematical model. The mathematical model used was a combination of two literature models: (1) an empirical kinetic model for recombinant growth and product formation and (2) a genetically structured model of the lac promoter-operator on a multicopy plasmid. The experimental system utilized was recombinant E. coli CSH22 bearing the temperature-sensitive plasmid pVH106/172, which codes for the synthesis of beta-galactosidase and the other lac operon genes under the control of a lac promoter. Mathematical model predictions for recombinant beta-galactosidase yield and specific growth rate were compared with fermentation measurements of these same quantities for conditions of chemical induction with cyclic AMP and IPTG, copy number amplification (by shifting culture temperature), and combined chemical induction and copy number amplification. The model successfully predicted experimental product yields for most cases of chemical induction even though the product yields varied from 0.34 x 10(3) to 1500 x 10(3) units/g cell mass. The kinetic model also correctly predicted a decline in the specific growth rate with increasing levels of plasmid and recombinant protein. The model was less successful at predicting product amplification at high copy numbers. A comparison of model predictions and experimental results was also used to investigate some of the assumptions used in constructing the mathematical models.