Characterization of plasmid yield

We constructed a new DNA vaccine vector, pDMB02-GFP, containing the runaway R1 replicon as described in the Methods section (Figure 1). The vector also carries the kanamycin resistance gene as well as the sequences necessary for expression of therapeutic genes in a eukaryotic host. For this work, we have included GFP as a placeholder for the therapeutic gene sequence. The specific yield of the new vector was compared to that of both the parent vector (pCP40) and a pUC-based DNA vaccine vector (pVAX1-GFP) after a mid-exponential phase temperature shift from 30°C to 42°C in LB medium (Figure 2). The pUC-based vector behaved as expected – the specific yield remained low (less than 1 mg/g DCW) at 30°C, and increased to about 4 mg/g DCW after temperature induction. Both of the R1-based plasmids produced higher specific yields than the pUC-based plasmid – a difference only partially accounted for by the larger size of the R1 plasmids. Interestingly, while temperature-induced amplification was observed for the R1 vectors at early time points (2 and 4 hours after the temperature shift), at later time points the yield at 30°C was higher than or the same as that at 42°C. It should be noted that the parent vector, pCP40, contains the phage lambda major leftward promoter (p_L) upstream of the origin. Despite using a host (DH5α) that does not contain the p_L repressor protein, we did not observe the plasmid instability alluded to by Remaut et al. [16]. The completely de-repressed phage promoter may not affect pCP40 stability because there is no recombinant protein sequence immediately downstream of p_L.

To further investigate the production capabilities of pDMB02-GFP, the specific yield after a temperature shift later in exponential phase (OD₆₀₀ = 1) was compared to the yields obtained after a mid-exponential phase shift (Figure 3). Typically, the temperature-shifted cultures reached a lower final optical density than the cultures that remained at 30°C, likely due to heat stress (Figure 3, A and B). A period of increased growth rate was observed between 0 and 2 hours post shift – possibly due to the culture transiently being at optimal growth temperature for E. coli (37°C) – followed by growth arrest. The timing of the temperature shift did not significantly impact the yield of the 42°C cultures, but resulted in higher yields at 30°C, likely due to the increased elapsed culture time. At 30°C, the cultures were typically in stationary phase at the later sampling times. It is possible that the high plasmid yields measured at these time points resulted from the phenomena of stationary-phase plasmid amplification [17] and/or amplification in response to nutrient starvation [18]. The maximum yield achieved in these experiments (Figure 3) is on par with the maximum yield produced by pCP40 (Figure 2), confirming that pDMB02-GFP did not lose any production capacity during the construction process.

There are several possible explanations for the experiment-to-experiment variability observed in the maximum yield of pDMB02-GFP. The use of a rich medium (LB) may be partially responsible, as the exact medium composition can vary from run-to-run. However, the mechanism of R1 replication is not designed for tight control of plasmid copy number. Instead, the goal is to prevent the copy number from dropping below one per cell – a concern that is more relevant for the low-copy, wild-type R1 plasmid [10]. This may result in increased clone-to-clone variability when the plasmid copy number is significantly higher than one, as observed in the runaway mutants.

Impact of seed growth phase on yield

In all of the experiments described above, shake flask cultures were inoculated directly from working seed banks. However, it has been reported previously that the growth phase of the seed can significantly impact the productivity of resulting cultures when using runaway replication plasmids for recombinant protein production [14]. To test the sensitivity of DH5α[pDMB02-GFP] to seed age, we inoculated flasks using seeds in late exponential, early stationary, and late stationary phase and measured the specific plasmid yield after 24 hours at 30°C. The data show that the plasmid production capacity of DH5α[pDMB02-GFP] cultures inoculated using late-stationary seeds is greatly reduced compared to cultures inoculated using seeds earlier in their respective growth phases (Figure 4). In addition, the reduction in culture productivity corresponds to an increase in the plasmid content of the seed culture, suggesting that high plasmid content in the seed negatively impacts the productivity of resulting cultures. These results are consistent with the findings of a previous report [14]. Remaut et al. [16] also alluded to the fact that runaway R1 plasmids can be unstable after prolonged growth, even at low temperatures. Direct inoculation from the working seed bank gave yields comparable to those obtained from late exponential/early stationary seeds, so we continued to use this as our standard protocol for maximum run-to-run consistency.

Quantification of repA gene expression

We were surprised to observe that pCP40 and pDMB02-GFP did not show the dramatic temperature-induced amplification reported in the early literature characterizing the runaway R1 replicon [9, 16]. In particular, we expected greater copy number control before the temperature shift; however, we are not the first group to observe that runaway replication is not completely suppressed at 30°C [14, 19]. To further investigate the temperature-dependent behavior of pDMB02-GFP we pursued additional studies at the RNA and protein levels.

In the R1 replicon, the repA gene codes for a protein required for initiation of replication at the origin and is transcribed from both the P _repA and P _copB promoters (Figure 5). copA is an antisense RNA that binds to and inhibits translation of repA mRNA, and CopB is a tetrameric repressor of P _repA . The two point mutations that lead to the runaway replication phenotype reduce the efficiency of P _copA and cause a temperature-dependent increase in transcription from P _copB [11]. The replication initiation protein, RepA, is likely the limiting factor in plasmid replication, owing to the fact that multiple copies of the protein are required for initiation of replication [10]. In the results that follow, we chose to focus our analysis exclusively on repA. Our experiments were all in the high-copy-number regime (compared to wild-type plasmid R1), and under these conditions, CopB is likely present in sufficiently high amounts to completely repress P _repA such that all repA mRNA is transcribed from P _copB [20]. Also, the antisense RNA control element, copA, is small and unstable [21] and as such we were unable to obtain reliable measurements of its expression.

Role of RepA protein expression levels

The repA mRNA expression data pointed to the possibility of a post-transcriptional limitation on plasmid yield at 42°C. One option for relieving this limitation is to increase expression of RepA. To this end, we changed the RepA start codon in pDMB02-GFP from GTG to ATG, resulting in the plasmid pDMB-ATG. There is evidence in the literature that genes with a GTG start codon are typically translated several-fold less efficiently than genes with an ATG start codon [22]. We chose a start codon mutation instead of a promoter replacement to minimize disruption of the other elements of the replicon. Also, since RepA is primarily cis acting [23], supplying RepA exogenously from either the chromosome or an additional plasmid was not a viable strategy for increasing RepA availability.

To qualitatively evaluate the expression of RepA protein in cultures of DH5α[pDMB02-GFP] and DH5α[pDMB-ATG] we used a Western blot to detect RepA using polyclonal antiserum obtained from rabbits [24]. For pDMB02-GFP, a band corresponding to RepA (33 kDa) was visible on the blot after 2 hr at 42°C and after 8 hr at 30°C, with the intensity of the RepA band being higher in the 30°C lysate (Figure 6: Lanes 6 and 7). This suggests that temperature-induced expression of RepA from pDMB02-GFP was occurring two hours post-shift, but by 8 hours, protein was either no longer being translated or was degraded, possibly by heat-shock-induced proteases [25]. A higher molecular weight band is visible in all lanes, but since this band is also present in the negative control lanes (Figure 6: Lanes 1 and 3), it is most likely due to cross-reaction with either the primary or secondary antibody rather than expression of RepA. As with specific plasmid yield, the trends in RepA expression were also different for pDMB-ATG. A RepA band is visible in the lanes corresponding to the 42°C samples at both 2 and 8 hours post-shift, and in the lane containing the 30°C sample at 8 hours post-shift (Figure 6: Lanes 11 – 13). The band for the 8 hr, 42°C sample is the most intense. It is clear that the kinetics of RepA expression are different for pDMB-ATG, and that RepA expression persists longer after the temperature shift when compared to expression from pDMB02-GFP.

Taken with the RNA expression data, the protein expression data for pDMB02-GFP suggest RepA protein levels may be limiting replication at 42°C. While we cannot deconvolute whether more RepA expression is leading to higher copy number or vice versa, it is clear that RepA protein is depleted after 8 hours at 42°C for the wild-type plasmid, and that the start codon mutation increases RepA protein levels at this time point. Despite this limitation, we have shown that high yields of pDMB02-GFP can be produced during constant-temperature growth at 30°C without a temperature shift. This is particularly advantageous for production-scale runs, as it reduces the complexity of the process and allows the culture to achieve a higher final cell density and higher volumetric yields. In addition, the mutant plasmid pDMB-ATG seems to perform better at 42°C, making it the more attractive vector if a temperature-shift process is desired.

Plasmid quality

For biopharmaceutical applications, it is important to assess the quality of the plasmid DNA produced in addition to the yield. While there is some debate over the clinically-relevant differences between the various plasmid isoforms (supercoiled, nicked, and linear), the FDA currently recommends that plasmid DNA biopharmaceuticals contain predominantly the supercoiled isoform [26]. Gel electrophoresis analysis of samples of pDMB02-GFP and pDMB-ATG collected from cultures with and without a temperature shift show a high percentage of supercoiled DNA (Figure 7). Also, the trends in plasmid quantity are consistent with the specific yield measured for these samples (Table 1 and Table 2).

These data show that neither the elevated temperature nor the start codon mutation had a negative impact on plasmid quality. Thus, pDMB02-GFP and pDMB-ATG are capable of producing not only high yields of plasmid DNA, but high quality product as well.

E. coli strains and plasmids

E. coli strain DH5α [F^- φ80lacZ ΔM15 Δ( lacZYA argF)U169 deoR recA1 endA1 hsdR17(r_k^-, m_k⁺) phoA supE44 thi-1 gyrA96 relA1 λ^- and the plasmid pVAX1 (2999 bp) were purchased from Invitrogen (Carlsbad, CA). pVAX1-GFP (3642 bp) was constructed by cloning the superfolding green fluorescent protein (sGFP) gene [27] into the multi-cloning site of pVAX1. The sGFP gene was obtained from pTrcsGFP, a gift from the Gregory Stephanopoulos laboratory (Department of Chemical Engineering, Massachusetts Institute of Technology, Cambridge, MA, USA). pCP40 (5029 bp) was constructed by Remaut et al. [16] and was obtained from the Belgian Coordinated Collections of Microorganisms BCCM/LMBP plasmid collection (accession number LMBP 951).

To construct pDMB02-GFP, a 1938-bp fragment of pVAX1 containing the human cytomegalovirus (CMV) immediate-early promoter/enhancer, bovine growth hormone (BGH) polyadenylation signal, and kanamycin resistance gene was PCR-amplified using primers containing AvrII and SbfI restriction sites. A 3073-bp fragment of pCP40 containing the R1 origin of replication along with the repA copA, and copB gene sequences was also PCR-amplified with the same restriction sites. The pVAX1 and pCP40 fragments were ligated to construct the plasmid pDMB02. The sGFP gene was cloned into the NheI and XhoI restriction sites downstream of the CMV promoter/enhancer. A Kozak sequence was also inserted at the start of the sGFP gene by adding the sequence ACC before the start codon and a valine codon (GTG) following the start codon. The Kozak sequence should help facilitate sGFP gene expression in mammalian cells [28]. The resulting vector, pDMB02-GFP, was 5668 bp in size (Figure 1). Correct construction was confirmed by restriction digests and sequencing.

The start codon of repA in pDMB02-GFP was mutated from GTG to ATG using site-directed mutagenesis resulting in the plasmid pDMB-ATG. Primers containing the mutation were used to amplify pDMB02-GFP using 20 cycles of PCR with Phusion high-fidelity DNA polymerase (New England Biolabs; Ipswich, MA). The primers were purified using a reverse phase cartridge by the vendor (Sigma-Aldrich; St. Louis, MO) and their sequences are shown below:

5’-GTGAAGATCAGTCATACCATCCTGCACTTACAATGCG-3’

5’-GCAGGATGGTATGACTGATCTTCACCAAACGTATTACCG-3’

After PCR, the template plasmid was digested using DpnI, and after clean-up the reaction was used to transform ElectroMAX DH10B cells (Invitrogen). Positive transformants were selected on LB/agar plates containing 50 μg/mL kanamycin, and the presence of the start codon mutation was verified by sequencing.

Preparation of working seed banks

Frozen working seed banks of DH5α[pDMB02-GFP] and DH5α[pDMB-ATG] were prepared by transforming subcloning-efficiency, chemically-competent DH5α (Invitrogen) with purified plasmid. Positive transformants were selected on LB/agar plates containing 25 μg/mL kanamycin. After overnight incubation at 30°C, a single colony was used to inoculate 3 mL of LB medium containing 25 μg/mL kanamycin. The culture was incubated overnight at 30°C. The next day, 500 μL of overnight culture was used to inoculate 50 mL of LB medium containing 25 μg/mL kanamycin in a 250-mL shake flask. The culture was incubated at 30°C until mid-exponential phase (OD₆₀₀ approximately equal to 0.5), at which time 900 μL of culture was added to 900 μL of cold 30% (v/v) glycerol in a cryogenic vial and immediately stored at −80°C. Working seed bank vials were discarded after two freeze-thaw cycles.

Culture conditions

Difco LB Broth, Miller (BD; Franklin Lakes, NJ) was used for shake flask cultures and contained 10 g/L tryptone, 5 g/L yeast extract, and 10 g/L NaCl. Cell density was monitored using OD₆₀₀ measurements on a DU800 spectrophotometer (Beckman Coulter; Indianapolis, IN). All cultures were mixed and aerated by agitation at 250 rpm unless otherwise specified.

Measurement of plasmid copy number

Plasmid copy number was determined from 100 μL of culture using the quantitative PCR (qPCR) method described by Bower et al. [29] In short, SYBR Green was used to detect amplification of plasmid-based and chromosomal gene targets from total DNA samples diluted 5-fold (if necessary) to be within the linear range of the assay. The ΔΔC_T method [30] was used to calculate plasmid copy number.

Measurement of plasmid DNA specific yield

Plasmid DNA was quantified from crude lysates prepared from OD₆₀₀ = 10 cell pellets using the method described previously [29]. Briefly, crude lysates were run on a Gen-Pak FAX anion-exchange column (Waters Corporation; Milford, MA) and pDNA was eluted using a NaCl gradient. Plasmid detected by absorbance at 260 nm was quantified using a standard curve of known quantities of purified plasmid DNA. Specific yield could then be calculated using the correlation that 1 OD₆₀₀ unit = 0.4 g DCW/L culture.

Measurement of repA mRNA expression

repA mRNA expression was measured using a quantitative real-time PCR assay. RNA was purified from OD₆₀₀ = 1 pellets using the Illustra RNAspin Mini RNA Isolation Kit (GE Healthcare; Piscataway, NJ). 1 g/L lysozyme was used for cell lysis in the first step of the protocol. Due to the high plasmid DNA content of the strains used in this work, an additional DNase digestion was required after the RNA purification step to remove contaminating DNA. 43 μL purified RNA was digested with 2 μL DNase I (New England Biolabs) in 5 μL of the supplied reaction buffer for 10 min. at 37°C. RNA was purified from the other reaction components using the RNA Cleanup protocol from the RNeasy Mini Kit (Qiagen; Valencia, CA). The RNA content of each sample was measured using a NanoPhotometer (Implen; Westlake Village, CA), and 800 ng of RNA was converted to cDNA using the QuantiTect Reverse Transcription Kit (Qiagen). Control reactions containing water instead of the reverse transcriptase enzyme were included for each sample.

Quantitative PCR was performed on a 7300 Real-Time PCR System (Applied Biosystems; Carlsbad, CA). The desired reference sample was diluted 2- to 1000-fold to prepare a standard curve. repA mRNA was detected using gene-specific primers (forward primer: 5’-CAGAGCTTAAGTCCCGTGGAAT-3’, reverse primer: 5’-TGACGTTCTCTGTTCGCATCA-3’) designed by Primer Express 3.0 software (Applied Biosystems). Each 25-μL reaction contained 1X Brilliant II SYBR Green QPCR High ROX Master Mix (Agilent Technologies; Santa Clara, CA), 200 nM each of the forward and reverse primers, and the experimental sample diluted 100-fold. The thermal cycling conditions were a 95°C hold for 10 min., followed by 40 cycles of 95°C for 30 sec. and 60°C for 1 min. Dissociation-curve analysis was also performed to check for the presence of primer dimers or non-specific products. Results were analyzed using the Applied Biosystems Sequence Detection Software (v. 1.3.1).

Immuno-detection of RepA protein

Cloning and expression of repA

To verify that the band being detected on the Western blots was indeed RepA, a positive control vector was constructed by cloning the repA gene into the BamHI/HindIII sites of pETDuet-1 (EMD Millipore; Billerica, MA), in-frame with an N-terminal 6X His tag. The resulting plasmid, pETDuet-repA, was used to express RepA-His in E. coli BL21 Star (DE3) (Invitrogen). 50-mL LB cultures of BL21 Star (DE3) containing either pETDuet-repA or pETDuet-1 and 100 μg/mL ampicillin were grown at 30°C with 250 rpm shaking and induced with 0.5 mM IPTG at an OD₆₀₀ of approximately 0.5. Six hours after induction, 20-mL aliquots of culture were harvested by centrifugation. Lysates were prepared from the pellets using disruption with 0.1 mm glass beads (Scientific Industries) in buffer containing 7 M urea, 0.1 M NaH₂PO₄, and 0.01 M Tris–HCl at pH 8.0.

Plasmid quality assessment

Plasmid DNA purified from OD₆₀₀ = 2 pellets using the Zyppy Plasmid Miniprep Kit (Zymo Research Corporation; Irvine, CA) was run on a 0.7% agarose gel at 90 V for 60 min. to separate the supercoiled, nicked (open-circle), and linear isoforms. The separated DNA was visualized by staining the gel with 0.5 μg/mL ethidium bromide.