Design of microaerobically inducible miniR1 plasmids

Impact statement Plasmid DNA manufacture is an essential step to produce gene therapy agents and next‐generation vaccines. However, little attention has been paid toward developing alternative replicons that can be coupled with large‐scale production conditions. Our results demonstrate that the miniR1 replicon can be efficiently induced by oxygen limitation when a copy of the regulatory protein RepA under control of a microaerobic promoter is used. The results are potentially attractive for industrial applications.


Impact statement
Plasmid DNA manufacture is an essential step to produce gene therapy agents and next-generation vaccines.However, little attention has been paid toward developing alternative replicons that can be coupled with large-scale production conditions.Our results demonstrate that the miniR1 replicon can be efficiently induced by oxygen limitation when a copy of the regulatory protein RepA under control of a microaerobic promoter is used.The results are potentially attractive for industrial applications.
Plasmid DNA (pDNA) is the active pharmaceutical ingredient in the so-called DNA vaccines 1 and in gene therapy products.The first pDNA vaccine for use in humans, which has shown high protection against SARS-CoV-2, was recently approved in India 2 .Furthermore, pDNA is often used as a template for in vitro transcription to produce mRNA vaccines 3 .Therefore, current and future demands of pDNA will require efficient production processes that can be implemented on a large scale, particularly considering the physical constraints in large-scale bioreactors, like imperfect mixing and mass transfer limitations.The oxygen required by Escherichia coli cells (the preferred host for pDNA production) in cultures can be difficult to meet, leading to local or global oxygen limitation.Although oxygen limitation causes undesired metabolic deviations, it can also result in increased pDNA yield (Y pDNA/X ) 4,5 .Therefore, environmental conditions in industrial bioreactors could be used to improve pDNA production.The majority of the plasmids used for DNA vaccines contain pMB1-derived replicons, particularly pUC 1 .High pDNA yields can be obtained with pUC plasmids if the culture temperature is increased to 40-45°C, which triggers runaway replication 6 .However, this temperature increase also triggers overflow metabolism and markedly reduces the viability of cells 5,7 .Moreover, the pUC origin of replication contains a cruciform sequence that is sensitive to endonuclease activity 8 , which may reduce the stability of pUC vectors.Only a few options have been proposed in addition to pUC replicons for high-yield pDNA production.Namely, the pCOR plasmids contain the R6K replicon, and were modified to yield relatively high plasmid copy numbers (PCN) per chromosome 9 .More recently, the R1 replicon was used to construct vectors that yielded PCN of several hundreds upon thermal induction 10 .In a different report, a plasmid named pminiR1 was assembled.pminiR1 contains a synthetic R1 replicon, in which some of the natural R1 sequences were eliminated 4 .It was shown that the Y pDNA/X obtained with pminiR1 was similar to that of the high PCN plasmid pUC57kan.In the present study, pminiR1 was produced in aerobic, microaerobic, and biphasic cultures with a regime change from aerobic to microaerobic conditions.pminiR1 was modified to create a microaerobically inducible version by overexpressing the positive replication control element (the protein RepA) upon oxygen limitation.
Plasmid pminiR1 contains a minimal set of sequences to allow its replication and selection (Figure 1A).For instance, the sequence of copB, which is a negative regulator, is also not included.Moreover, the parB locus originally present in plasmid R1 is not included in pminiR1.The parB locus contains the genes hok (coding for a very stable killing protein) and sok (coding for unstable antisense mRNA that regulates hok expression).In addition to antibiotic resistance, the parB locus provides another mechanism to stabilize plasmid R1, known as the "postsegregational killing" of plasmid free cells 11 .Therefore, the stability of pminiR1 is expected to be controlled only by antibiotic resistance.
Production of plasmid pminiR1 was first characterized under aerobic or microaerobic batch cultures of E. coli W12, which constitutively expresses the Vitreoscilla hemoglobin (VHb).Expression of VHb improves the growth and metabolic performance of E. coli under both aerobic and oxygen-limited conditions 7 .The growth profiles of aerobic and microaerobic cultures for pminiR1 production are shown in Figure S1.
Under aerobic conditions, exponential growth was observed only during the first 4-6 h.Later on, growth was approximately linear (Figure S1A).Linear growth was also reported in previous studies in the production of pDNA containing the R1 replicon 4,10 .Interestingly, the pDNA yield from biomass (Y pDNA/X ) increased markedly when the cell growth was linear (Figure S1A).The cell growth under microaerobic conditions was very slow and both pDNA concentration and Y pDNA/X increased steadily during the first 10 h of culture (Figure S1B).The Y pDNA/X at the end of the aerobic cultures was 6.4 ± 0.7 mg/g, while under the microaerobic regime, it reached 8.2 ± 0.4 mg/g.A previous study also reported an increase of Y pDNA/X due to oxygen limitation; however, the values shown in Figure S1 are higher than those of the aforementioned report 4 .Possible causes for this are different media and temperature, as well as pH control in the present study.The final Y pDNA/X in microaerobic cultures was also higher than that reported by Bower and Prather 10 , in which a plasmid containing the R1 replicon was produced in aerobic cultures at 30°C and shifted to 42°C.
To further increase Y pDNA/X upon oxygen depletion, microaerobically inducible pUC replicons that substantially improve pDNA production under oxygen-limited regimes have been designed 12 .In the present study, a similar design principle was applied: a second copy of the positive replication control gene (repA), placed under the transcriptional control of a microaerobic promoter (P St ) 13 , was inserted in pminiR1 (Figure 1B).The microaerobically inducible plasmid was named pminiR1-MAInd.The initial assumption was made that increased abundance of RepA would increase plasmid replication, provided that the amount of CopA does not increase in the same proportion, in agreement with previous simulations using mathematical models 14 .The copA promoter strength is 12.5 transcripts/min 15 , while for repA, it is around 1.4 transcripts/min 16 , which is nearly nine times lower.Consequently, it has been established that the synthesis of the RepA protein is a rate-limiting factor for replication initiation 17 .Therefore, it is expected that increased repA expression may lead to higher PCN.The production of pminiR1 and pminiR1-MAInd was evaluated under biphasic conditions, as described in Materials and methods in Supporting Information.The result showed the growth profile of the biphasic culture of pminiR1 and pminiR1-MAInd (Figure S2).The Y pDNA/X increased for both plasmids after change in the regime to microaerobic conditions, reaching values higher than those achieved under constant regimes.This is advantageous for large-scale cultures, where a gradual transition to microaerobic conditions would occur.The maximum Y pDNA/X was obtained at 13 h of culture for both plasmids, reaching 9 ± 1 mg/g for pminiR1 and 12 ± 1 mg/g for pminiR1-MAInd (Figure S2).This means increases of 33% and 84% over the original plasmid in biphasic and in fully aerobic cultures, respectively.
To confirm the relation between the increased Y pDNA/X and repA expression, the PCN and the ratio of repA copies to copA copies per cell were measured and are reported in Figure 1C.In cultures at constant regimes, the repA/copA expression ratio increased 31% under the microaerobic condition, compared to the aerobic condition, which led to a 3.9-fold increase of PCN.The absolute PCNs are lower than those reported by Bower and Prather 10 .However, those authors performed cultures at 30°C, which improves the stability of low-copy-number plasmids in E. coli.Microaerobic conditions were also used for the production of the pUC plasmid (pVAX1) in batch mode.The authors reported a 61% increase of Y pDNA/X , compared to aerobic conditions 7 , which is similar to the results shown in Figure 1C.Microaerobic cultures at a very low oxygen transfer rate (OTR, 10 mmol/l h) resulted in five times higher Y pDNA/X values for pVAX1, compared to aerobic cultures at OTR max of 110 mmol/l h 18 .Therefore, it may be possible to improve the results shown in Figure 1C by decreasing the OTR max of the culture.
The biphasic regime resulted in increased repA/copA expression ratios, compared to constant regime cultures.Notwithstanding the increase of Y pDNA/X of pminiR1 observed in biphasic cultures compared to constant regimes, the PCN increase was almost unchanged (Figure 1C), meaning that there was no direct equivalence between PCN and Y pDNA/X .This was also observed by Bower and Prather 10 and contrasts with pUC plasmids 12 .The repA/copA expression ratios in cells bearing pminiR1-MAInd were more than 4-fold higher than that for pminiR1, in agreement with substantial increases of PCN and Y pDNA/X (Figure 1C).The PCN values obtained by the end of biphasic cultures of cells bearing pminiR1-MAInd under microaerobic condition were 3-and 12-fold higher than those attained in a constant microaerobic and aerobic regime, respectively, using pminiR1.This change is substantially higher than that reported for a microaerobically inducible pUC plasmid.In this report, the PCN was 2.3-fold higher during the microaerobic phase, compared to the aerobic phase of a fedbatch culture 12 .This could be related to the fact that the positive control molecule for miniR1 is a protein, contrary to the case of pUC plasmids, in which the PCN is controlled by RNA.The half-life of the majority of RNA molecules in E. coli is between 3 and 8 min 19 , while the half-life for proteins (excluding abnormal or unstable proteins) is several hours 20 .Therefore, the additional expression of RepA under microaerobic conditions could lead to more stable effects than the expression of RNA.
Plasmid topology is a relevant factor when pDNA is used as an active pharmaceutical ingredient 1 .It is generally recommended that at least 80% of the pDNA is present in the supercoiled isoform (SCF) 1 .We analyzed the supercoiled content of both plasmids obtained from the different regimes.The result showed that under both aerobic and microaerobic regimes, the sc-pDNA fraction of pminiR1 was 80%.In biphasic cultures, the sc-pDNA fraction during the aerobic phase of the sc-pDNA was 80%, while in the microaerobic phase, it increased to 90%.Similar results were obtained for pminiR1-MAInd (Figure S3).
Here, we present an alternative to deal with microaerobic conditions that can easily occur during the production of pDNA at any culture scale, from shake flasks to large bioreactors.The R1 replicon can be an alternative to traditional pUC replicons.Therefore, a minimal R1 plasmid was modified to induce replication upon transition to microaerobic conditions.In the experiments presented here, it is considered that microaerobic conditions are present when dissolved oxygen tension (DOT) is below 10%.The modified miniR1 contains an extra copy of the gene repA under control of the microaerobic promoter P St .Although the DOT for optimal induction of P St has not been reported, it has been demonstrated that the promoter from the Vitreoscilla hemoglobin is maximally induced at DOT below 5% 21 .Since both promoters control the expression of microbial globins, we chose DOT of 2% for induction of repA as an initial approximation.While precise control at such a low DOT in industrial bioreactors is difficult, our experiments show the differences between two contrasting conditions: fully aerobic and microaerobic conditions.
The presented results demonstrate that the transition to microaerobic conditions is better than constant microaerobicity to increase the production of plasmids containing the minimal R1 replicon.Furthermore, our biodesign proved that overexpression of the gene repA is an efficient way to increase pDNA yields and PCN upon transition to oxygen limitation.Calculated over the entire whole culture time of cultures shown in Supporting Information, the global productivity in biphasic cultures was 1.61 ± 0.12 mg/l h for pminiR1 and 1.71 ± 0.15 for pminiR1-MAInd, which is an increase of only 6.2%.However, the SCF increased from ~80% to ~90%.Therefore, the increased pDNA yields and SCF of pminiR1-MAInd make it advantageous for downstream operations.Overall, we show that the minimized R1 replicon can be an interesting option to the traditional pUC replicon for achieving high yields.Microaerobic conditions can increase the PCN of pminiR1.The designed inducible plasmid is therefore an alternative for oxygen transfer limitations in bioreactors, and therefore, can help to efficiently scale up pDNA production processes.

Figure 1 .
Figure1.Design of the miniR1 plasmids and the main results of plasmid copy numbers and repA expression ratios in aerobic and microaerobic cultures.(A) Scheme of the plasmid pminiR1, containing the minimized R1 replicon.(B) Scheme of the plasmid pminiR1-MAInd, containing an extra copy of repA under control of the microaerobic promoter P St .copA, gene coding for antisense RNA that lowers the transcription rate; ntpII, neomycin phosphotransferase gene; oriR1, origin of replication or R1 replicon; P 2 , repA promoter; P AmpR , promoter of the ampicillin resistance gene; P St , promoter of a globin from Salmonella typhi; repA, replication initiation protein gene; ssiA, single-strand initiator (primosome assembly site); tap, translational activator peptide required for RepA synthesis.(C) Main results of repA and copA expression levels, Y pDNA/X , and plasmid copy number (PCN) per chromosome under the different conditions studied.Samples correspond to the following hours of culture: a 10, b 12, c 3, and d 11.Time profiles of the cultures are shown in Supporting Information.