An ideal gene therapy vector should enable persistent transgene expression without limitations of safety and reproducibility. Here we report the development of a non-viral episomal plasmid DNA (pDNA) vector that appears to fulfil these criteria. This pDNA vector combines a scaffold/matrix attachment region (S/MAR) with a human liver-specific promoter (α1-antitrypsin (AAT)) in such a way that long-term expression is enabled in murine liver following hydrodynamic injection. Long-term expression is demonstrated by monitoring the longitudinal luciferase expression profile for up to 6 months by means of in situ bioluminescent imaging. All relevant control pDNA constructs expressing luciferase are unable to sustain significant transgene expression beyond 1 week post-administration. We establish that this shutdown of expression is due to promoter methylation. In contrast, the S/MAR element appears to inhibit methylation of the AAT promoter thereby preventing transgene silencing. Although this vector appears to be maintained as an episome throughout, we have no evidence for its establishment as a replicating entity. We conclude that the combination of a mammalian, tissue-specific promoter with the S/MAR element is sufficient to drive long-term episomal pDNA expression of genes in vivo.
Gene therapy vectors derived from modified viruses are the most effective gene delivery systems in use today. However, their efficacy at gene transfer is tempered by potential toxicity.1, 2, 3, 4 An ideal synthetic non-viral vector for human gene therapy must provide sustainable levels of therapeutic gene expression without compromising the viability of the host at either the cellular or somatic level. This would be best achieved through the delivery of a plasmid DNA (pDNA) expression system that is maintained as a functional, extra-chromosomal (episomal) entity following delivery. The development of such expression systems remains a main strategic task for non-viral gene therapy research and is a prerequisite for any future long-term clinical application.
Recent studies involving pDNA vectors in vivo have focused on monitoring longitudinal transgene expression in the liver as a function of time, owing to the ease with which functional pDNA may be delivered by hydrodynamic injection to the murine liver.5, 6, 7, 8 According to the results of one of these studies, viral promoters were found to promote the highest initial levels of transgene expression but subsequently expression declined sharply. Expression levels of the initially most effective of these vectors containing a cytomegalovirus (CMV)/chicken-β-actin fusion promoter element dropped subsequently 31 000-fold within 2 months. This decline was attributed to transcriptional silencing of the promoter.7 It has been hypothesized that the shutdown of viral promoters in the liver could be a result of DNA methylation processes9 and/or because of the lack of transcription factors in the liver that can support their continuous expression.10 Moreover, some cytokines such as α- or γ-interferon have been shown to actively inhibit the CMV promoter.11 Previous studies utilizing DNA vectors have shown that a longer duration of transgene expression is possible in vivo when driven by a mammalian promoter such as α1-antitrypsin (AAT) in combination with the entire genomic loci12 or viral elements.13 These studies, however, did not investigate the mechanisms behind the sustenance of expression.
Several alternative strategies have been devised to reduce episomal pDNA silencing in the liver in vivo. One of the most common approaches has been to generate pDNA mini-circles, essentially devoid of extraneous bacterial DNA.14, 15, 16 Although in some cases these minimally sized systems have led to improvements in the duration of liver expression in vivo, the inefficient and expensive purification procedures to obtain sufficient mini-circles for in vivo application have so far outweighed any potential benefit that these vectors might have.
Recent developments in pDNA episome research suggested that the creation of an episomal system for long-term in vivo expression might be possible by a combination of appropriate DNA sequences.
Two novel classes of pDNA episomes that are small enough to be readily manipulated for such approach have recently been investigated. One of them comprises self-replicating circular expression systems that contain the Epstein–Barr virus17 elements oriP and EBNA1.18 However, Epstein–Barr virus is a causative agent of infectious mononucleosis as well as of various malignancies in humans, and hence the therapeutic potential of episomal systems containing Epstein–Barr virus elements may be limited.
By contrast, the other system containing scaffold/matrix attachment regions (S/MARs), which are ubiquitously present in all eukaryotic genomes, is less likely to present such limitations. S/MARs can function as cis-acting elements for episomal maintenance.17 Mammalian S/MAR elements such as the human β-interferon gene cluster S/MAR sequence have been shown in vitro to render S/MAR-containing pDNA vectors resistant to chromosomal integration19, 20, 21 to avoid epigenetic silencing22 and to promote extra-chromosomal stability.23 Importantly, these properties are conferred without the need for virus-encoded proteins or selective pressure.21, 24 Mitotic stability appears to be provided by direct interaction of the S/MAR element with components of the nuclear matrix,23, 25 including the scaffold attachment factor A protein, which is the principal component of cellular chromatin and chromosomes.26 Furthermore, this interaction probably brings the S/MAR plasmids into contact with the host replication machinery, which is located on the nuclear matrix, thereby facilitating replication once per cell cycle and hence extra-chromosomal stability. Impressively, an S/MAR-containing non-viral vector has been recently used to develop genetically modified pigs.27 An S/MAR element has also been incorporated, among other elements, into lentiviral vectors28 and into a plasmid vector already optimized for long-term expression.13 The mechanism by which S/MAR elements provide the unique ability to sustain long-term transgene expression in vivo has, however, so far not been specifically investigated.
We report here the first major study specifically examining transgene expression from simple S/MAR-containing plasmid episomes in the murine liver after hydrodynamic delivery, which shows that prolonged transgene expression in vivo depends on both the S/MAR element and the choice of promoter.
According to previous research in vitro, S/MARs endow pDNA expression systems with three fundamental properties: episomal maintenance, resistance to epigenetic silencing and extra-chromosomal, mitotic stability. To explore if these properties can be maintained in vivo, we prepared a set of different pDNA vectors based on the structure of a known S/MAR element-containing plasmid. Each of them was designed to express the luciferase transgene but was otherwise modified with respect to promoter and the presence or absence of the S/MAR element (for detailed plasmid maps, see Supplementary Figure 1). The pDNA systems prepared were pEPI-Luc (CMV promoter; S/MAR element), pEPI-Luc control (CMV promoter; no S/MAR element), pLucA1 (human AAT promoter; S/MAR element), and pLucA1 control (human AAT promoter; no S/MAR element).
AAT-S/MAR plasmids, but not CMV-S/MAR plasmids, mediate persistently high levels of transgene expression in vivo
Each of the pDNA expression systems was separately administered to groups of six mice by hydrodynamic injection to obtain high initial levels of transgene expression in liver hepatocytes following injection. Longitudinal expression was then observed over a period of 6 months using a Xenogen in vivo bioluminometer. Results from this set of time-course experiments are illustrated for each of the pDNA vectors used (Figure 1). All pDNA constructs were able to mediate high levels of transgene expression and subsequent protein activity in hepatocytes up to 24 h post-administration. However, in the case of three of the plasmids pEPI-Luc, pEPI-Luc control and pLucA1 control, the levels of transgenic luciferase were found to decline abruptly within a week after the injection (F=8.3, P<0.01). In contrast, pLucA1 was able to mediate high levels of transgene expression and subsequent protein activity in hepatocytes up to 6 months after administration (F=5.05, P<0.05). By way of illustration, transgene expression levels after 1 month mediated by pEPI-Luc, pEPI-Luc control and pLucA1 control systems were all only about 0.03% of those determined at 24 h post-administration. In contrast, transgene expression levels after 1 month mediated by pLucA1 were still at least 10% (P=0.02) of the 24 h post-administration level. Expression from this vector is up to 20 times higher at 3 months than that of all the others at 1 week. Importantly, in the absence of the S/MAR moiety, pLucA1 behaves no differently to pEPI-Luc. Within 5 days, luciferase expression levels in this group are less than 3% of the 24 h post-administration value. This vast difference in the behaviour of pLucA1 compared with the other pDNA expression systems under investigation suggests a powerful synergy between the human AAT promoter and the S/MAR element of pLucA1 providing a continuous high level of transgene expression whereas such synergy obviously does not operate between the S/MAR element and CMV promoter in pEPI-Luc. Moreover, in a follow-up study that illustrates the specific synergy required for sustained expression, we replaced the AAT promoter with the ubiquitous mammalian promoter elongation factor 1α. Transgene expression mediated by this vector, even with the inclusion of an S/MAR element, is not sustained beyond 14 days following hydrodynamic delivery (Supplementary Figure 3).
It has been shown previously that the function of the S/MAR vectors depends not only on an active transcriptional unit linked to the S/MAR but also that transcription should run into this element.29 In this study, it was shown that on the 2-kb-long β-interferon S/MAR, a cryptic termination site is present approximately 1500–1700 bp into the S/MAR element. Consistent with this, we performed northern blot analysis of transcripts isolated from the liver 24 h after hydrodynamic delivery, which revealed an mRNA transcript of pLucA1 plasmid of approximately 3.6 kb long and an mRNA transcript of the pLucA1-control plasmid of 2 kb (data not shown). This confirms that as in the in vitro study, after hydrodynamic injection the transcription of the transgene in the pLucA1 plasmid goes through the luciferase gene (2 kb long) and into the S/MAR element but terminates prematurely at position 1600 bp, due to the S/MAR cryptic termination site.
The tissue distribution of transgene expression after vector administration was analysed by immunohistochemistry for the four different plasmids used. Initially, 24 h after administration, high levels of luciferase expression were observed, particularly in hepatocytes adjacent to the central veins of the liver lobules, gradually declining outwards towards the periphery. Interestingly, a high degree of transgene expression is observed after 3 weeks in hepatocytes near the liver capsule probably due to a ‘congestion’ effect at this final barrier. The decrease in expression with distance from central veins most likely correlates with a decrease in the effects of hydrodynamic injection pressure with distance from the vein (Figures 2a–d). Luciferase expression was not detected in other non-parenchymal cells. No inherent luciferase activity was seen in the liver tissue taken from control mice not subjected to administration of pDNA (data not shown). The persistence of luciferase expression in hepatocytes at 3 weeks and 6 months after administration of pLucA1 (Figures 2e and f) agrees very well with corresponding data determined by in situ bioluminescent analysis (Figure 1).
S/MAR plasmids do not become established as actively replicating episomes in hepatocytes
Given the clear difference in behaviour between pLucA1-mediated expression and levels of expression mediated by the other pDNA expression systems under investigation, we strove to find mechanistic explanations for the superiority of pLucA1. Initially, we set out to determine if pLucA1 pDNA had become actively established as a replicating episome in the hepatocytes or was only being passively maintained and thereby lost during hepatocyte turnover. We performed quantitative PCR to determine the relative copy number of plasmid molecules throughout the duration of the experiment. Our analysis demonstrated that proportionally the number of pDNA constructs in the liver reduced over time; there were still, however, a significant number of plasmid molecules in the liver 6 months after administration. For the pLucA1 vector, it was determined that there were >20 000 copies of vector per cell 24 h after administration, which dropped to ∼200 copies per cell at 7 days. After 6 months, ∼50 copies of plasmid per cell remained. Each of the administered plasmid vectors exhibited a similar copy number profile. In related experiments, plasmid vectors were rescued by bacterial transformation. For pLucA1 plasmid, the average number of bacterial colonies obtained from DNA isolated from the liver 6 months after administration (1–3 colonies per plate) was less than the colonies obtained at 3 weeks post-administration (7–10 colonies per plate), verifying the decrease in pDNA copy numbers with time. The sharp decline in vector number in the immediate weeks following administration is likely due to the administration procedure; transduced cells damaged in the infusion process will be eliminated and unstable pDNA in the nucleus degraded.30 It has been suggested that, eventually, only properly established vector molecules, influenced by higher nuclear architecture in the hepatocyte nucleus, will be able to express the luciferase transgene for extended periods of time.31 Without episomal establishment, these will of course be lost during hepatocyte turnover.
We also performed a more direct procedure to determine whether the vectors replicated after administration. In this experiment, rapid liver regeneration was induced by performing 70% hepatectomies on mice 3 weeks after administration of the pDNA constructs. Luciferase expression levels were then examined daily for up to 32 days after resection by in situ bioluminescent analysis (Figure 3), making sure that mice were imaged for luciferase from all sides, as liver mass topology can change slightly during post-hepatectomy hepatocyte proliferation. Liver regeneration after hepatectomy normally takes place by a process in which the remaining hepatocytes undergo one or two cell division cycles until the initial liver mass is reconstituted (after approximately 14 days). During liver regeneration, any actively established pDNA vector would replicate and spread throughout the reconstituted tissue leading to persistence of transgene expression, whereas passively maintained pDNA would not. In our case, resection and full regeneration of murine liver led to a reduction in luciferase expression to almost background levels (Figure 3), although residual luciferase expression mediated by pLucA1 appeared to remain a little higher than the expression levels mediated by the three other pDNA vectors (F=12.26, P=0.01 on day 32). Similar results were obtained when longitudinal luciferase transgene expression after partial hepatectomy was examined in all mice 6 months post-injection (results not shown).
We also investigated the replicative state of the pDNA vectors by performing replication-dependent restriction assay using enzymes that specifically cut DNA that have replicated in mammalian cells. For this, 20 μg of extracted liver DNA was isolated at several time points after pDNA administration and digested with StuI and then with DpnI, MboI or Sau3AI enzyme. The latter three enzymes recognize the same cutting site (GATC), but the activity of DpnI and MboI depends on the methylation pattern of the DNA. DpnI cuts only those DNA molecules that retain the bacterial methylation pattern, whereas MboI cuts only those DNA molecules that have lost the bacterial methylation pattern by replicating twice in mammalian cells. Sau3AI cuts regardless of the methylation status and acts as a control.21 Southern blot analysis demonstrating the restriction of pDNA isolated from the liver at different time points throughout the experiment is provided in Figure 4a. Comparison of the restriction pattern of the pLucA1 plasmid isolated from the liver of injected mice at 1 week post-injection (Figure 4a, lanes 4–6) and 6 months post-injection (Figure 4a, lanes 7–9) with the restriction pattern of pLucA1 isolated from Escherichia coli cells (Figure 4a, lanes 1–3) demonstrates that pLucA1 did not replicate during the period of 6 months in the mouse liver. More specifically, double digestion of pLucA1 with DpnI–StuI in all cases shows the expected fragments of an adenine methylated GATC motif (Figure 4a, lanes 2, 5 and 8), whereas digestion of pLucA1 with MboI enzyme was blocked, verifying the presence of methylated adenine in the same GATC motif (Figure 4a, lanes 1, 4 and 7). In this case, the StuI digestion shows the linearized pLucA1 plasmid at the expected size (7.6 kb). Finally, digestion with Sau3AI–StuI enzymes is not blocked by any kind of methylation in the GATC motif, and serves as a control for efficient digestion of pLucA1 (Figure 4a, lanes 3, 6 and 9).
In conclusion, the pLucA1 plasmid did not show any detectable replication in the liver of adult MF1 mice for a period of 6 months, verifying the results from partial hepatectomy experiments. These results suggest that the investigated pDNA vectors are maintained after administration in the hepatocytes in a passive episomal state.
Hydrodynamically injected pDNA vectors are maintained by hepatocytes in an episomal state
To provide physical proof for the passive episomal state of our constructs in the hepatocytes, we isolated DNA from the liver tissue obtained by partial hepatectomy, at 3 weeks post-administration, or following the killing of the animal 32 days after the partial hepatectomy when the liver had regained its original mass. DNA was digested with StuI enzyme, which has a single restriction site in all four pDNAs (pEPI-Luc, pLucA1 and their non-S/MAR controls) and was then subjected to Southern blot analysis. A representative blot can be seen in Figure 4b showing an individual band of expected size in all pDNA lanes. Hence all pDNAs appear to remain episomal at 3 weeks post-administration even though luciferase expression had declined to near-background levels in all cases, with the notable exception of pLucA1-mediated luciferase expression (Figure 1). Further evidence for episomal maintenance was also provided by plasmid rescue of pEPI-Luc and pLucA1 isolated from kanamycin-resistant E. coli after transformation with total DNA from the livers of hydrodynamically treated mice. In this case, only intact free pDNA from the isolated liver sample would produce bacterial colonies on plates containing kanamycin, the resistance marker present on all four injected plasmids. The restriction patterns of these plasmids were consistent with unmodified non-integrated plasmid constructs (data not shown). On the basis of these data, we surmised that pDNA silencing and not pDNA shedding is the reason for the substantial decline in longitudinal pDNA-mediated expression over the first 3 weeks after administration. Importantly, without partial hepatectomy, the episomal status of pLucA1 was maintained even up to 6 months post-administration (Figure 4b, lane 2). The passive episomal nature of pLucA1 is, however, emphasized by its disappearance from DNA isolated from de novo hepatocytes from the newly regenerated liver 32 days after partial hepatectomy (Figure 4b, lane 3). Although we cannot exclude some low-level integration of the pDNAs, we conclude that this does not contribute significantly to the observed expression patterns.
The promoters of silenced pDNA constructs are methylated
As all pDNAs were retained in hepatocytes as episomes following hydrodynamic administration, we hypothesized that the difference in expression between pLucA1 and the other pDNA vectors could be due to the extent of DNA methylation. DNA methylation in animals is mediated by a class of enzymes called DNA methyltransferases, which catalyse methylation of the cytosine residue in the 5′-dCpG-3′ (CpG) dideoxynucleotide motif. The presence of a 5-methyl-cytosine base in the promoter of specific genes inhibits the binding of transcription factors and other proteins. This form of methylation also attracts methyl-DNA binding proteins and histone deacetylases that modify the chromatin structure around a transcription start site. Accordingly, we elected to determine the methylation status of CpG-rich areas present in the promoters of pDNA isolated from the liver tissue after hydrodynamic administration at the time of partial hepatectomy. The CMV promoter of both pEPI-Luc plasmids contains three CpG-rich areas (forming a CpG island of 444 bp long) whereas the AAT promoter contains only one such CpG-rich area (100 bp long). Indeed, the CMV promoter is well known to become heavily methylated in vivo hence prohibiting further active transcriptional activities.9 Furthermore, the CpG islands of the CMV promoter are also supplemented by several other dCpG dideoxynucleotides throughout the promoter length that themselves also have equal probabilities of becoming methylated. Therefore, to assess the CMV and AAT promoter methylation status in pDNAs isolated 3 weeks post-administration, we employed a restriction enzyme/PCR methodology, which utilizes several enzymes that are inhibited by cytosine methylation within their specific restriction sites. The methylation status of both types of promoter is summarized in Figure 5. Detailed information regarding enzymes utilized is provided in Supplementary Table 1. The CMV promoter located in both pEPI-Luc pDNAs was methylated to various degrees ranging from 20 to 80%, regardless of the presence or absence of the S/MAR element. In the case of pLucA1, the AAT promoter was not methylated whereas in the case of pLucA1 control (without S/MAR), up to 85% promoter methylation was observed. The AAT methylation status in pLucA1 isolated 6 months post-administration was also assessed and was found to remain essentially unchanged from that before (Supplementary Table 1). Therefore, in mechanistic terms, the success of pLucA1 at mediating long-term transgene expression appears to be due to the lack of significant AAT promoter methylation, suggesting that the accompanying S/MAR element in pLucA1 has an unexpected capacity to modulate promoter methylation (and hence prevent transcriptional silencing) thereby enabling long-term transgene expression in the liver from a pDNA episome.
Repeated administration of pDNA restores initial levels and similar profiles of transgene expression in hepatocytes
To establish whether methylation was fundamental to the silencing of pDNA expression or whether the liver had become refractory to transgene expression, we repeated the administration of the pDNA constructs 2 weeks after the initial hydrodynamic injection and again after a further 3 weeks following the second administration. After each subsequent administration, the initial levels of transgene expression and subsequent decline were indistinguishable from the original (Figure 6). This suggests that the hepatocytes were capable of generating transgene expression from the pDNA constructs but that these vectors could be silenced in a similar manner as after a single administration.
The clearly established oncogenic risks involved in the clinical application of integrating viral vectors highlight an urgent need for the design of novel, safe and efficient approaches to gene therapy utilizing either non-integrating viral vectors or synthetic non-viral constructs with the ability to provide long-term therapeutic transgene expression. The non-viral approach requires the design of episomal pDNA expression systems that should preferably comprise no bacterial or viral components and be devoid of non-human elements for sustained episomal maintenance and replication in human cells. Moreover, specifically designed and selected promoters should drive the expression of the transgene at endogenously equivalent and clinically relevant levels and the application of such expression systems should, ideally, be adapted specifically to target the relevant cells for the treatment of the disease in question. Accordingly, the maintenance and expression of such pDNA expression systems should be without adverse or toxic consequences for the host at either the cellular or somatic level. Here we show unparalleled long-term transgene expression from a simple non-viral pDNA expression system, pLucA1, following hydrodynamic delivery to the murine liver. This is the first demonstration in vivo of the specific ability of S/MAR sequences in combination with a tissue-specific mammalian promoter to confer sustained episomal gene expression from a standard plasmid vector. Furthermore, we have demonstrated that the transgene expression is likely to be sustained as a result of an unexpected protective effect of the S/MAR/promoter domain on methylation-sensitive sites in the promoter element of the pLucA1 expression system.
The detailed molecular mechanism underlining pDNA silencing and control of silencing is presently not known. What is known is that the chromatin fate of heterologous DNA such as pDNA can be altered in many ways after entry into the nucleus. Histone acetylation, DNA methylation and phosphorylation of histones all work to influence the status of the heterologous DNA and the expression of transgenes.32 Trimethylation of histone 3 lysine 9 (H3K9), methylation of histone 3 lysine 27 (H3K27) and histone 4 trimethylated lysine 20 together with hypoacetylation of histone tails are all well-known hallmarks of heterochromatin DNA and are linked to the silencing of transgene expression.33 On the other hand, open or active chromatin is associated with acetylated histones (especially lysines 9 and 14 of histone 3 tail) and methylation of lysine 4 histone 3 (H3K4). In a very recent study, antibody-specific chromatin immunoprecipitation analysis was employed to investigate the chromatin changes of plasmids following hydrodynamic injection to the murine liver and it was observed that expression of a protein (hAAT) driven by either the RSV or the elongation factor 1α promoter was low after 35 days, despite initial high expression. The observed transgene silencing was attributed to hypoacetylation of histones and methylation of H3K9 and H4K20 among others.34 When the CpG-rich bacterial backbone of these plasmids was removed (creating mini-circles), expression of hAAT was found to last longer, and methylation markers were lower, indicating the negative effect of CpG dinucleotides on gene expression. Even a simple reduction of CpG sequences on pDNA enhances transgene expression, implying that CpG methylation suppresses expression.35 In similar reports,36, 37 it was verified that CpG dinucleotides could negatively affect transgene expression from pDNA in vivo, even with linearization of the transgene expression plasmid before hydrodynamic injection. Finally, other groups have reported the negative effect of CpG methylation in vitro,38 and recently the extensive methylation of CpG islands of a CMV promoter-driven adenoviral vector following delivery into the muscles of mice was shown, with methylation levels similar to our results.9 Here we show that the CpG-rich area of the AAT promoter remains unmethylated in the presence of an S/MAR element, whereas the presence of the same S/MAR element was unable to protect a corresponding CMV promoter from such CpG methylation and transcriptional silencing. Additionally, the S/MAR element cannot prevent the shutdown of expression from a nonspecific promoter elongation factor 1α. It is worth noting however that luciferase expression from some of the control plasmids was slightly higher than background levels, which could be explained by the various degrees of promoter methylation observed. When another known DNA motif (GATC) was checked for methylation in vitro, it was shown that pDNA vectors with S/MAR were protected from cytosine methylation in GATC motifs present in their CMV promoter, whereas vectors without S/MAR were rapidly silenced due to such methylation.22 In our studies, we could not detect any methylation of C in GATC in pLucA1, but we did observe a 33% (2/6 mice) methylation of the C in GATC of the pLucA1 control at 3 weeks post-administration (data not shown). Such an outcome suggests that the unexpected benefits of combining an S/MAR element with a promoter to suppress methylation and hence enable episomal long-term transgene expression are linked to a requirement for the promoter element to have a tissue-specific origin and function. Indeed, consistent with this suggestion, a recent study39 showed that inclusion of a tissue-specific promoter (β-globin) within a lentiviral vector enabled sustainable secretion of hFIX ex vivo, whereas similar studies utilizing tissue-specific promoters clearly demonstrate their advantages in gene expression over the immunogenic, transiently expressed CMV promoter.40, 41 Additionally, no prolonged luciferase expression or DNA persistence was observed after intravenous lung administration of an S/MAR-containing pEPI-1 vector with a CMV-driven expression cassette.42 Together with our data, these results suggest that S/MAR-containing vectors are prone to promoter attenuation in vivo if not combined with constitutive or tissue-specific promoters. The exact mechanism of how an S/MAR element may protect a particular promoter in vivo from methylation now requires further characterization but it can be assumed that cell- or tissue-specific factors play a decisive role in this process.
Recently, it has been suggested that gene silencing is not a single event, but rather a series of events that may be initiated from methylation of CpG sites and spread towards the promoter of the transgene leading to gene silencing.43 Indeed in transgenic mice, DNA methylation has been shown to spread from an integrated retrovirus to flanking sequences in the mouse genome.44 It was speculated that a ‘dynamic equilibrium’ may exist between ‘forces’ that support the spread of DNA methylation and ‘forces’ that actively block this spread.43 Particular DNA sequences have been identified as such blocking forces. For instance, it was observed that an A-rich sequence in the human GSTP1 gene could form a boundary between methylated and unmethylated DNA.45 We cannot exclude that an A-rich S/MAR element may behave in a similar way to protect the AAT promoter of pLucA1 from methylation and silencing, but why an A-rich S/MAR element should act as a blocking sequence for the AAT promoter of pLucA1 and not for the CMV promoter of pEPI-Luc remains unclear. It is also known that the protein-binding partners of S/MAR elements will change according to their topological status and environment as demonstrated by cloning an 800 bp S/MAR element from the upstream border of the human β-interferon domain at various positions within a transcribed region of 4.3 kb.46 At a distance of about 4 kb from the transcriptional start site, the S/MAR was found to support transcriptional initiation, whereas at distances below 2.5 kb transcription was essentially shut down. However, again this does not provide an obvious explanation for the different functional behaviour of the S/MAR element in pLucA1 and pEPI-Luc. We suggest that the addition of an S/MAR element into an AAT plasmid allows it to function and exist in an unmethylated state as it would endogenously, whereas the natural methylation and silencing of the CMV promoter in the liver can obviously not even be prevented by incorporation of an S/MAR element into a vector with this promoter. We suggest that the addition of an S/MAR element into an AAT plasmid confers a conformational structure to the AAT promoter, which allows it to interact with cell-specific factors that keep it in an unmethylated state and function as it would have endogenously. However, the S/MAR is not able to act in this way on the CMV promoter, which is naturally methylated and silenced after an initial expression period in the liver.
On the basis of in vitro studies,19, 20, 21, 22, 47 our original hypothesis was that the S/MAR element might also help to establish pDNA expression systems as actively maintained, stable replicating episomes in the liver. The rapid decline in the transgene expression in the first few days after hydrodynamic delivery may be due to a combination of elimination of transduced cells damaged in the infusion process and degradation of any unstable pDNA in the nucleus as previously suggested.30 Thereafter, however, we observed a remarkably stable passive maintenance of the pLucA1 expression plasmid in liver hepatocytes. This plasmid vector is gradually lost during the natural turnover of liver cells with time and disappears rapidly after stimulation of hepatocyte cell division by partial hepatectomy. Interestingly, we observed an approximately fivefold upregulation of luciferase gene expression 1 day after partial hepatectomy, especially with pEPI-Luc and control (Figure 3). Such a transient upregulation has been reported previously48 and was explained by the upregulation of nuclear factor-κB and other transcription factors (AP-1, SP1) during rapid liver regeneration. The CMV promoter contains four nuclear factor-κB binding sites compared to only one in the AAT promoter, which could perhaps account for the higher upregulation effected by pEPI-Luc and control in comparison to pLucA1 and control. Nevertheless, this upregulation is transient and is quickly lost.
To circumvent the slow decline in luciferase expression from pLucA1, we finally investigated the possibility of repeat administration using the hydrodynamic delivery method and found that repeat administration of pLucA1 did indeed completely restore the initial levels of transgene expression. Therefore, infrequent repeat administration should be able to overcome the very slow decline in transgene expression of our AAT promoter and S/MAR combination and thus make prolonged therapeutic applications of this type of vector feasible.
Materials and methods
The plasmids used in this study (Supplementary Information) were pEPI-Luc (kindly provided by Professor Hans Lipps, University of Witten, Germany), in which luciferase expression is driven by the CMV promoter (human CMV, immediate early promoter), pLucA1, in which luciferase expression is driven by the human AAT promoter,10 and which we created by substituting the AseI-digested CMV-Luciferase cassette of pEPI-Luc with an AAT-Luciferase cassette derived from an FspI–XbaI-digested plasmid. The control plasmids were derived by removal of the S/MAR region (obtained from the human β-interferon gene) by AccI digestion. The plasmid backbone is identical on all plasmids and is derived from the commercial plasmid pGFP-C1 (Clontech, Mountain View, CA, USA). The plasmids were amplified in E. coli DH10B cells (Invitrogen Ltd., Paisley, UK) and isolated using an Endotoxin free Maxi prep (Qiagen, Crawley, UK). All restriction enzymes were purchased from NEB Biolabs (New England Biolabs, Hitchin, UK).
Southern and northern blots
For DNA analysis, total liver DNA was extracted using a GenElute mammalian genomic DNA kit (Sigma-Aldrich Company Ltd., Gillingham, UK). The isolated DNA was quantified using a NanoDrop ND1-1000 spectrophotometer (Labtech International Ltd., Ringmer, UK). For Southern analysis, total liver DNA (15 μg) was digested with an appropriate restriction enzyme (StuI)that is a single cutter to all plasmids, separated on 0.8% agarose gels (20 V, 20 mA overnight) and blotted onto nylon membranes (Hybond XL, Amersham plc, Little Chalfont, UK). A 408 bp DNA fragment derived from the restriction digest of a segment of the kanamycin region, which is common to all plasmids, using enzyme AlwNI, was labelled with 32P (Rad-Prime labelling kit, Invitrogen) and used as a probe. The hybridization was performed in Church buffer (0.25 M sodium phosphate buffer (pH 7.2), 1 mM EDTA, 1% BSA, 7% SDS) at 65 °C for 16 h.
For the replication-dependent restriction assay, 20 μg of total liver DNA was digested with StuI and further digested with DpnI, MboI or Sau3AI enzyme overnight. An agarose gel separation and Southern analysis wee then performed as mentioned.
For northern analysis, total liver RNA was isolated using an RNeasy mini kit (Qiagen) and 15 μg total RNA was separated on 1.2% formaldehyde/agarose gels, transferred to a nylon membrane and hybridized with a 32P-labelled (Rad-Prime labelling kit, Invitrogen) luciferase gene probe.
The relative amounts of pDNA in hepatocyte samples were calculated by real-time PCR using an ABI PRISM 7000 sequence detector. PrimerExpress software was used to design oligonucleotide primers (Invitrogen) and TaqMan probe (Eurofine MWG Operon, Ebersberg, Germany) for luciferase to determine the amounts of S/MAR plasmid, and primers and probes specific for the mouse Titin gene49 to enable normalization between the samples by calculating the number of cells used as the input. The primers for luciferase gene were as follows: forward 5′-IndexTermGGCGCGTTATTTATCGGAGTT-3′; reverse 5′-IndexTermCCATACTGTTGAGCAATTCACGTT-3′. The probe sequence was 5′-FAM-IndexTermTGCGCCCGCGAACGACATTTATAAT-TAMRA-3′. Amplification reactions (25 μl) contained 5 μl of template DNA, 12.5 μl of PlatinumQuantitative PCR Supermix-UDG with Rox (Invitrogen), 0.1 mM primers and 0.2 mM probe. Following the initial steps at 50 °C (2 min) and then at 95 °C (10 min), PCR was carried out for 40 cycles of 95 °C (15 s) and then 60 °C (1 min). Serial dilutions of plasmids containing appropriate sequences to produce a standard amplification curve for quantification were carried out and all samples were tested in triplicate.
Luciferase expression from the four constructs was analysed by one-way analysis of variance to assess statistical significance. A post-analysis of variance multiple comparison procedure (Tukey's HSD) was further performed to assess pairwise differences in expression confirmed by analysis of variance with a significance level P=0.05.
Plasmid rescue experiments
DH10B E. coli cells were transformed by electroporation, using 1 μg DNA prepared by total liver DNA isolation. Transformed colonies were selected on agar plates containing 30 mg ml−1 kanamycin. DNA was isolated from individual resistant clones, subjected to restriction analysis (StuI) and analysed by electrophoresis on 0.8% agarose gels.
For methylation analysis, 200 ng of isolated liver DNA was subjected to restriction analysis, using enzymes blocked by methylated cytosines (AciI, FauI, HpaII, HgaI present on AAT promoter or AciI, FauI, HpaII, HgaI SfoI, AfeI present on CMV promoter). The enzyme-treated DNA was subsequently used for PCR analysis with primers (designed with MacVector 7.1.1 software) specific for the AAT and CMV promoters, using standard Taq polymerase (Invitrogen). PCR was conducted in a thermocycler T3 (Biometra) with the following primers (Invitrogen): AAT promoter: forward 5′-IndexTermCGACCAGAATTCCTGCAGGTACCCGCCACCCCCTCCACC-3′; reverse 5′-IndexTermCCTGCAAGATCTGGTACCGGTCCGGAATGCCAAGCTTTTCACTGTC-3′; product size: 327 bp.
Template DNA was initially denatured by heating at 95 °C for 5 min, followed by 30 cycles of denaturation at 95 °C for 45 s, annealing at 59 °C for 30 s and primer extension at 72 °C for 45 s. Incubation for 5 min at 72 °C followed to complete extension.
CMV promoter: forward 5′-IndexTermCCCATATATGGAGTTCCGCG-3′; reverse 5′-IndexTermCCAAAAATAGGATCTCTGGCATG-3′; product size: 1250 bp.
Reactions were first denatured at 95 °C for 5 min, followed by 30 cycles of 95 °C for 45 s, 53 °C for 30 s, 72 °C for 45 s and final extension for 5 min at 72 °C.
PCR products were analysed on a 0.8% agarose gel
CpG islands of the promoter regions were determined using MethPrimer, a web-based program from Urogene that can be freely accessed at http://www.urogene.org/methprimer/index1.html. As a positive control for methylation, a CpG methylase enzyme (M.SssI; NEB Biolabs) was used to methylate all cytosine residues (C5) within the double-stranded dideoxynucleotide recognition sequence 5′dCpG3′ according to the manufacturer's protocol.
Hydrodynamic injection of animals—in vivo bioimaging
MF-1 mice (1–2 months old, 20–25 g) (B&K Universal Ltd, Hull, UK) were rapidly injected, over 5–8 s, via the tail vein with 2.5 ml phosphate-buffered saline containing 50 μg of each plasmid vector DNA using a 27-gauge needle. Animals were given adequate care in compliance with institutional and UK guidelines.
At 24 h and at regular intervals after hydrodynamic injections, mice were injected intraperitoneally with 300 μl of D-luciferin (Gold Biotechnology, Inc., St Louis, MO, USA) (15 mg ml−1 in phosphate-buffered saline), anaesthetized by isoflurane and then imaged for bioluminescence by the IVIS Imaging 50 Series (Xenogen). Bioluminescent imaging (BLI) was performed in a light-tight chamber on a temperature-controlled, adjustable stage while isoflurane was administered by means of a gas manifold at a flow rate of 2%. Images were acquired at a medium binning level (8) and a 20-cm Weld of view. Acquisition times were either 5 or 10 s, depending on the intensity of the luminescence. The Xenogen system reported bioluminescence as photons per second per cm2 per sr in a 2.86-cm-diameter region of interest covering the liver. The autofunction was used to define the minimum for the scale at each time point. This value was 5% of the maximum in each case. Data were analysed by using LivingImage 2.50 software (Xenogen). Background levels of bioluminescence were 1 × 106 photons per second per cm2 per sr.
Histology and immunohistochemistry
For routine histological analysis, formalin-fixed paraffin-embedded liver samples were cut into sections 4 μm in thickness, deparaffinized in histoclear and dehydrated through a series of decreasing concentrations of ethanol. Sections were stained with haematoxylin and eosin to observe liver morphology. For immunohistochemical detection of luciferase, paraffin-embedded sections were pretreated by incubation in 0.01 M citrate buffer and heated in a microwave oven twice for 5 min. Samples were then incubated overnight at 4 °C with a rabbit monoclonal antibody against luciferase (1:100 dilution; Santa Cruz Biotechnology, Santa Cruz, CA, USA). Endogenous peroxidase activity was blocked with 3% H2O2 and methanol. Avidin and biotin treatment was used to prevent endogenous nonspecific staining. The secondary antibody was a biotin-conjugated anti-rabbit goat antibody (1:500 dilution; DAKO, Carpinteria, CA, USA). Colour development was performed with diaminobenzidene (DAKO detection kit; Vector Laboratories, Ltd., Peterborough, UK).
scaffold/matrix attachment region.
Hacein-Bey-Abina S, von Kalle C, Schmidt M, Le Deist F, Wulffraat N, McIntyre E et al. A serious adverse event after successful gene therapy for X-linked severe combined immunodeficiency. N Engl J Med 2003; 348: 255–256.
Hacein-Bey-Abina S, Von Kalle C, Schmidt M, McCcormack MP, Wulffraat N, Leboulch P et al. LMO2-associated clonal T cell proliferation in two patients after gene therapy for SCID-X1. Science 2003; 302: 415–419.
Themis M, Waddington SN, Schmidt M, von Kalle C, Wang YH, Al-Allaf F et al. Oncogenesis following delivery of a nonprimate lentiviral gene therapy vector to fetal and neonatal mice. Mol Ther 2005; 12: 763–771.
Thomas CE, Ehrhardt A, Kay MA . Progress and problems with the use of viral vectors for gene therapy. Nat Rev Genet 2003; 4: 346–358.
Al-Dosari M, Zhang G, Knapp JE, Liu D . Evaluation of viral and mammalian promoters for driving transgene expression in mouse liver. Biochem Biophys Res Commun 2006; 339: 673–678.
Rettig GR, McAnuff M, Liu D, Kim JS, Rice KG . Quantitative bioluminescence imaging of transgene expression in vivo. Anal Biochem 2006; 355: 90–94.
Wilber A, Frandsen JL, Wangensteen KJ, Ekker SC, Wang X, McIvor RS . Dynamic gene expression after systemic delivery of plasmid DNA as determined by in vivo bioluminescence Imaging. Hum Gene Ther 2005; 16: 1325–1332.
Zhang G, Budker V, Wolff JA . High levels of foreign gene expression in hepatocytes after tail vein injections of naked plasmid DNA. Hum Gene Ther 1999; 10: 1735–1737.
Brooks AR, Harkins RN, Wang PY, Qian HS, Liu PX, Rubanyi GM . Transcriptional silencing is associated with extensive methylation of the CMV promoter following adenoviral gene delivery to muscle. J Gene Med 2004; 6: 395–404.
Kramer MG, Barajas M, Razquin N, Berraondo P, Rodrigo M, Wu C et al. In vitro and in vivo comparative study of chimeric liver-specific promoters. Mol Ther 2003; 7: 375–385.
Gribaudo G, Ravaglia S, Caliendo A, Cavallo R, Gariglio M, Martinotti MG et al. Interferons inhibit onset of murine cytomegalovirus immediate-early gene transcription. Virology 1993; 197: 303–311.
Alino SF, Crespo A, Dasi F . Long-term therapeutic levels of human alpha-1 antitrypsin in plasma after hydrodynamic injection of nonviral DNA. Gene Therapy 2003; 10: 1672–1679.
Ehrhardt A, Peng PD, Xu H, Meuse L, Kay MA . Optimization of cis-acting elements for gene expression from nonviral vectors in vivo. Hum Gene Ther 2003; 14: 215–225.
Bigger BW, Tolmachov O, Collombet JM, Fragkos M, Palaszewski I, Coutelle C . An araC-controlled bacterial cre expression system to produce DNA minicircle vectors for nuclear and mitochondrial gene therapy. J Biol Chem 2001; 276: 23018–23027.
Chen ZY, He CY, Kay MA . Improved production and purification of minicircle DNA vector free of plasmid bacterial sequences and capable of persistent transgene expression in vivo. Hum Gene Ther 2005; 16: 126–131.
Darquet AM, Rangara R, Kreiss P, Schwartz B, Naimi S, Delaere P et al. Minicircle: an improved DNA molecule for in vitro and in vivo gene transfer. Gene Therapy 1999; 6: 209–218.
Jackson DA, Juranek S, Lipps HJ . Designing nonviral vectors for efficient gene transfer and long-term gene expression. Mol Ther 2006; 14: 613–626.
Stoll SM, Sclimenti CR, Baba EJ, Meuse L, Kay MA, Calos MP . Epstein–Barr virus/human vector provides high-level, long-term expression of alpha1-antitrypsin in mice. Mol Ther 2001; 4: 122–129.
Jenke AC, Stehle IM, Herrmann F, Eisenberger T, Baiker A, Bode J et al. Nuclear scaffold/matrix attached region modules linked to a transcription unit are sufficient for replication and maintenance of a mammalian episome. Proc Natl Acad Sci USA 2004; 101: 11322–11327.
Lufino M, Manservigi R, Wade-Martins R . An S/MAR-based infectious episomal genomic DNA expression vector provides long-term regulated functional complementation of LDLR deficiency. Nucleic Acid Res 2007; 35: 10.
Piechaczek C, Fetzer C, Baiker A, Bode J, Lipps HJ . A vector based on the SV40 origin of replication and chromosomal S/MARs replicates episomally in CHO cells. Nucleic Acids Res 1999; 27: 426–428.
Jenke ACW, Scinteie MF, Stehle IM, Lipps HJ . Expression of a transgene encoded on a non-viral episomal vector is not subject to epigenetic silencing by cytosine methylation. Mol Biol Rep 2004; 31: 85–90.
Jenke BHC, Fetzer CP, Stehle IM, Jonsson F, Fackelmayer FO, Conradt H et al. An episomally replicating vector binds to the nuclear matrix protein SAF-A in vivo. EMBO Rep 2002; 3: 349–354.
Papapetrou EP, Ziros PG, Micheva ID, Zoumbos NC, Athanassiadou A . Gene transfer into human hematopoietic progenitor cells with an episomal vector carrying an S/MAR element. Gene Therapy 2006; 13: 40–51.
Mearini G, Nielsen PE, Fackelmayer FO . Localization and dynamics of small circular DNA in live mammalian nuclei. Nucleic Acids Res 2004; 32: 2642–2651.
Kipp M, Gohring F, Ostendorp T, van Drunen CM, van Driel R, Przybylski M et al. SAF-Box, a conserved protein domain that specifically recognizes scaffold attachment region DNA. Mol Cell Biol 2000; 20: 7480–7489.
Manzini S, Vargiolu A, Stehle IM, Bacci ML, Cerrito MG, Giovannoni R et al. Genetically modified pigs produced with a nonviral episomal vector. Proc Natl Acad Sci USA 2006; 103: 17672–17677.
Park F, Kay MA . Modified HIV-1 based lentiviral vectors have an effect on viral transduction efficiency and gene expression in vitro and in vivo. Mol Ther 2001; 4: 164–173.
Stehle IM, Scinteie MF, Baiker A, Jenke ACW, Lipps HJ . Exploiting a minimal system to study the epigenetic control of DNA replication: the interplay between transcription and replication. Chromosome Res 2003; 11: 413–421.
Miao CH, Thompson AR, Loeb K, Ye X . Long-term and therapeutic-level hepatic gene expression of human factor IX after naked plasmid transfer in vivo. Mol Ther 2001; 3: 947–957.
Stehle I, Postberg J, Rupprecht S, Cremer T, Jackson D, Lipps H . Establishment and mitotic stability of an extra-chromosomal mammalian replicon. BMC Cell Biol 2007; 8: 33.
Jenuwein T, Allis CD . Translating the histone code. Science 2001; 293: 1074–1080.
Fuks F . DNA methylation and histone modifications: teaming up to silence genes. Curr Opin Genet Dev 2005; 15: 490–495.
Riu E, Chen Z-Y, Xu H, He C-Y, Kay MA . Histone modifications are associated with the persistence or silencing of vector-mediated transgene expression in vivo. Mol Ther 2007; 15: 1348–1355.
Yew NS, Zhao HM, Przybylska M, Wu IH, Tousignant JD, Scheule RK et al. CpG-depleted plasmid DNA vectors with enhanced safety and long-term gene expression in vivo. Mol Ther 2002; 5: 731–738.
Hodges BL, Taylor KM, Joseph MF, Bourgeois SA, Scheule RK . Long-term transgene expression from plasmid DNA gene therapy vectors is negatively affected by CpG dinucleotides. Mol Ther 2004; 10: 269–278.
Siegfried Z, Eden S, Mendelsohn M, Feng X, Tsuberi BZ, Cedar H . DNA methylation represses transcription in vivo. Nat Genet 1999; 22: 203–206.
Hong K, Sherley J, Lauffenburger DA . Methylation of episomal plasmids as a barrier to transient gene expression via a synthetic delivery vector. Biomol Eng 2001; 18: 185–192.
Chang AH, Stephan MT, Sadelain M . Stem cell-derived erythroid cells mediate long-term systemic protein delivery. Nat Biotech 2006; 24: 1017–1021.
Cordier L, Gao GP, Hack AA, McNally EM, Wilson JM, Chirmule N et al. Muscle-specific promoters may be necessary for adeno-associated virus-mediated gene transfer in the treatment of muscular dystrophies. Hum Gene Ther 2001; 12: 205–215.
Schiedner S, Hertel S, Johnston M, Biermann V, Dries V, Kochanek S . Variables affecting in vivo performance of high-capacity adenovirus vectors. J Virol 2002; 76: 1600–1609.
Conese M, Auriche C, Ascenzioni F . Gene therapy progress and prospects: episomally maintained self-replicating systems. Gene Therapy 2004; 11: 1735–1741.
Turker MS . Gene silencing in mammalian cells and the spread of DNA methylation. Oncogene 2002; 21: 5388–5393.
Jahner D, Jaenisch R . Retrovirus-induced denovo methylation of flanking host sequences correlates with gene inactivity. Nature 1985; 315: 594–597.
Millar DS, Paul CL, Molloy PL, Clark SJ . A distinct sequence (ATAAA)n separates methylated and unmethylated domains at the 5′-end of the GSTP1 CpG island. J Biol Chem 2000; 275: 24893–24899.
Schubeler D, Mielke C, Maass K, Bode J . Scaffold/matrix-attached regions act upon transcription in a context-dependent manner. Biochemistry 1996; 35: 11160–11169.
Jenke ACW, Eisenberger T, Baiker A, Stehle IM, Wirth S, Lipps HJ . The nonviral episomal replicating vector pEPI-1 allows long-term inhibition of bcr-abl expression by shRNA. Hum Gene Ther 2005; 16: 533–539.
Loser P, Jennings GS, Strauss M, Sandig V . Reactivation of the previously silenced cytomegalovirus major immediate-early promoter in the mouse liver: involvement of NF kappa B. J Virol 1998; 72: 180–190.
Charrier S, Stockholm D, Seye K, Opolon P, Taveau M, Gross D et al. A lentiviral vector encoding the human Wiskott–Aldrich syndrome protein corrects immune and cytoskeletal defects in WASP knockout mice. Gene Therapy 2005; 7: 597–606.
The work was supported by the Medical Research Council. SNW is a Philip Gray Fellow of the Katharine Dormandy Trust.
About this article
Cite this article
Argyros, O., Wong, S., Niceta, M. et al. Persistent episomal transgene expression in liver following delivery of a scaffold/matrix attachment region containing non-viral vector. Gene Ther 15, 1593–1605 (2008). https://doi.org/10.1038/gt.2008.113
- persistent expression
- episomal maintenance
- S/MAR plasmid
- hydrodynamic delivery
Maintenance of Plasmid Expression in vivo Depends Primarily on the CpG Contents of the Vector and Transgene
Molecular Biology (2020)
Advanced Drug Delivery Reviews (2020)
Novel Non-integrating DNA Nano-S/MAR Vectors Restore Gene Function in Isogenic Patient-Derived Pancreatic Tumor Models
Molecular Therapy - Methods & Clinical Development (2020)
Episomal vectors based on S/MAR and the β-globin Replicator, encoding a synthetic transcriptional activator, mediate efficient γ-globin activation in haematopoietic cells
Scientific Reports (2019)
Episomal minicircles persist in periods of transcriptional inactivity and can be transmitted through somatic cell nuclear transfer into bovine embryos
Molecular Biology Reports (2019)