1. Introduction
The liver is an essential organ involved in a variety of functions including detoxification, vitamin and glucose storage, iron metabolism, regulation of hormones, production of most plasma proteins, and metabolism of carbohydrates, fat and proteins [
1]. Hepatocytes constitute the bulk of cells in the liver parenchyma and are affected by the majority of monogenic liver inherited disorders. Consequently, the liver is a prime target for gene therapy, and many liver-targeted therapies are under investigation [
2,
3,
4,
5,
6,
7,
8,
9]. A common gene therapy strategy consists of gene supplementation, where a functional form of an affected gene is delivered to the target cells, which allows the phenotypic correction of a disorder. For decades, adeno-associated viruses (AAVs) and lentiviruses (LVs) have been used as vectors for gene supplementation [
10,
11]. These vectors are able to release ectopic DNA in the nucleus of transduced hepatocytes that is responsible for transgene expression.
AAVs have a single-stranded DNA (ssDNA) genome of 4.7 kb length that when they reach the nucleus are converted into double-stranded DNA (dsDNA) by host proteins prior to transcription [
12]. For gene therapy applications, recombinant AAVs (rAAV) are generated, where the ssDNA viral genome is replaced by a therapeutic expression cassette flanked by the AAV inverted terminal repeats (ITR). Instead, self-complementary AAV vectors (scAAV) were designed to deliver their genome in a dsDNA form due to a truncation in one of the ITRs [
12,
13]. Transduction with scAAV vectors produces higher levels of the transgene and in a faster way compared to single-stranded AAV (ssAAV) vectors [
14]. However, its cloning capacity is considerably reduced [
12]. rAAV episomal DNA does not replicate and it can be silenced or lost, especially during cell division [
11]. Nevertheless, in non-growing tissues, rAAV gene expression can last for years [
15]. Integration of rAAV genomes is rare, random and occurs preferentially at chromosomal breakage sites [
16]. Integrated rAAV genome expression might be silenced through histone modification and chromatin condensation [
17,
18,
19]. Thus, rAAV transgene expression is primarily derived from episomal DNA forms.
LVs, as well as recombinant LV vectors (rLVs), used in gene therapy applications, have a single-stranded positive polarity RNA genome that is retrotranscribed into dsDNA by the viral retrotranscriptase [
10], generating the DNA template for integration. Integration into the host cell is driven by the viral integrase and it is required for viral gene expression [
20,
21,
22]. The cloning capacity of rLV vectors exceeds that of rAAV vectors, reaching up to 10 kb in size [
23]. Integration overcomes the loss of episomal DNA during cell division; however, concerns about insertional mutagenesis and oncogenic potential exist. To avoid these problems, integration deficient lentiviral vectors (IDLV) were developed. IDLVs carry point mutations in the viral integrase that avoid viral genome integration [
24]. As a consequence, non-integrated viral genomes accumulate in the form of 1-long terminal repeat (LTR) and 2-LTR circular episomal DNA [
25,
26], which maintain transgene expression capacity, albeit to a reduced level compared to their integration-competent LV (ICLV) counterparts [
27]. As it occurs with AAV expression, IDLV episomal DNA expression can be lost upon cell division [
28,
29].
In the last few decades, rAAV and rLV vectors have been extensively studied in preclinical settings as viral gene therapy vehicles. In order to maintain efficient therapeutic protein expression, high vector doses are sometimes required, which can induce activation of innate and adaptive responses against the viral vectors decreasing its transgene expression [
11]. Moreover, it can also induce adverse effects, including hepatotoxicity, nephrotoxicity, neurotoxicity and viral genome integration into the host genome, causing undesired mutations and driving proto-oncogene expression [
30,
31]. For these reasons, development of more efficient and safer viral vectors, as well as methods to improve viral vector transduction, are still needed. In this regard, several authors have explored the possibility of using chemical compounds as adjuvants of viral vector transduction. Indeed, DNA-damaging agents, including topoisomerase I (TOP1) and II (TOP2) inhibitors (e.g., camptothecin (CPT) and Etoposide (Eto), respectively), produced an increase in the rAAV transduction efficiency in cell culture in vitro [
32,
33]. Topoisomerase inhibitors are known to induce double-strand break formation in cellular DNA [
34]. In this context, DNA damage breakage sites induced by Eto and γ-irradiation increased rAAV genome integration [
16], suggesting that the increased rAAV transduction observed by other authors might be a consequence of AAV DNA integration induced by the treatments. However, epigenetic mechanisms have been shown to trigger silencing of rAAV genome expression from integrated viral DNA [
17,
18,
19]. Thus, it is still unclear the mechanism by which topoisomerase inhibitors increase AAV-derived transgene expression.
Integration is a key process in ICLV infection that is driven by viral integrase and allows robust lentivirus gene expression [
20,
21,
22]. It has been reported that integration-defective human immunodeficiency virus (HIV) strains are still capable of integrating their genomes, albeit at very low levels compared to their ICLV counterparts [
35,
36]. Compounds that increase ICLV transduction have also been identified [
37]. Groschel et al. described that CPT and Eto induce cell cycle arrest and promote ICLV integration into the host chromosomal DNA [
38]. Induction of DNA damage by treatment with H
2O
2 or by exposure to γ-irradiation, prior to transduction, increased IDLV integration [
39]. Moreover, the presence of DNA damage agents, such as bleomycin and Eto, during viral transduction was able to increase IDLV but not ICLV viral genome copy numbers in macrophages [
35]. In contrast to those observations, a different study, performed in monocyte-derived macrophages, showed that induction of DNA damage by CPT and Eto blocks HIV-1 infection after completion of viral DNA synthesis, at the step involving 2-LTR circle and provirus formation [
40].
Nevertheless, how topoisomerase inhibitors modulate viral vector transgene expression remains unclear and further research is needed to shed light on the specific mechanisms responsible for the observed effects. In this study, we use a systematic approach to evaluate the phenotypic effect of CPT and Eto treatments on rAAV and rLV vector transduction in non-dividing hepatic cells in vitro. On the one hand, we analyzed the effect of the compounds on the induction of DNA damage and cell viability and correlated the results with a quantitative analysis of viral vector gene expression. On the other hand, we used different rAAV (ssAAV vs. scAAV) and rLV (ICLV vs. IDLV) vectors, all expressing the same mCherry reporter gene, to show that CPT and Eto increase their expression by increasing the integration of viral episomal genomes in hepatic cells.
2. Materials and Methods
2.1. Cells, Antibodies, Plasmids and Reagents
The origins of HepG2-NTCP [
41] and HEK-293T [
42] cells have been described previously. All cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM) (cat. nº: D6429-500ML, Sigma-Aldrich, St. Louis, MI, USA) supplemented with 10% fetal bovine serum (FBS; cat. nº: 35-079-CF, Corning, NY, USA), 10 mM HEPES (cat. nº: H0887-100ML, Sigma-Aldrich), 100 units/mL penicillin and 100 mg/mL streptomycin, (cat. nº: P4333-100ML, Sigma-Aldrich) and 2 mM L-glutamine (cat. nº: G7513-100ML, Sigma-Aldrich) in 5% CO
2 at 37 °C. To avoid cell overgrowth during the experiments, culture medium was supplemented with 2% dimethyl sulfoxide (DMSO; cat. nº: D2650, Sigma-Aldrich) starting one day before any treatment was applied or before any viral transduction and it was kept during the entire length of the experiments. Camptothecin (CPT; cat. nº: S1288, Selleckchem, Houston, TX, USA), etoposide (Eto; cat. nº: S1225, Selleckchem) and actinomycin D (ActD; cat. nº: A9415, Sigma-Aldrich) were obtained from commercial sources. Aliquots of stock solutions were prepared in DMSO and stored frozen at −20 °C until they were used in the experiments at the concentrations indicated in each experiment. The pLenti-C-Myc-DDK plasmid (cat. nº: PS100064, Origene, Rockville, MD, USA) was used for the cloning of the mCherry reporter gene. The lentivirus packaging vectors (pRSV-Rev, pMDLg-pRRE and pMD2.G) were obtained from the Addgene repository (Addgene plasmid #12251, #12253 and #12259). The AAV packaging plasmid pDP3 was obtained from PlasmidFactory (#PF0433, Germany).
2.2. Molecular Cloning
2.2.1. Generation of pLenti-mCherry-NLS Plasmid for Lentiviral Production
The DNA sequence corresponding to the mCherry open reading frame was amplified by PCR using the following primers: AsiSI-mCherry-Fwd (5′-AGCTGCGATCGCATGGTGAGCAAGGGCGAGG-3′) and MluI-mCherry-Rev (5′-AGCTACGCGTCTTGTACAGCTCGTCCATGCC-3′), and the CHC-mCherry plasmid as a template with the Supreme NZYProof DNA polymerase (cat. nº: MB28302, NzyTech, Lisbon, Portugal), according to the manufacturer´s instructions. The polymerase chain reaction (PCR) conditions were as follows: 1 cycle at 96 °C for 4 min; 25 cycles of 30 s at 96 °C, 30 s at 60 °C and 30 s at 72 °C; and 1 cycle at 72 °C for 5 min. The column-purified PCR product and the destination plasmid vector pLenti-C-Myc-DDK were subjected to AsiSI (cat. nº: R0630L, New England Biolabs (NEB), Ipswich, MA, USA) and MluI-HF (cat. nº: R3198L, NEB) digestion and subsequently ligated using T4 DNA ligase (cat. nº: M0202S, NEB). The sequence of the resulting pLenti-mCherry plasmid was verified by Sanger sequencing.
The pLenti-mCherry-NLS plasmid was generated by the insertion of the nuclear localization signal (NLS) of the cellular c-myc gene followed by three consecutive stop codons at the end of the mCherry open reading frame in the parental pLenti-mCherry plasmid. To do so, the following oligonucleotides 5′-CGCGTCCTGCTGCTAAGAGAGTGAAACTGGATTGATAATAGC-3′ and 5′-TCGAGCTATTATCAATCCAGTTTCACTCTCTTAGCAGCAGGA-3′ were subjected to annealing reactions to produce a short double-stranded DNA product with 5′ and 3′ overhang ends ready for directed ligation into a MluI-HF and XhoI (cat. nº: R0146L, NEB) digested pLenti-mCherry plasmid. The sequence of the resulting pLenti-mCherry-NLS plasmid was verified by Sanger sequencing and was used for the generation of lentiviruses as described below.
2.2.2. Generation of pAAV-EALBAAT-mCherry-NLS and dsAAV-EALBAAT-mCherry-NLS Plasmids for AAV Production
The DNA sequence corresponding to the mCherry open reading frame followed by the NLS sequence of c-myc was amplified by PCR from the pLenti-mCherry-NLS plasmid using the following primers: 5′-ACTGCCATGGTGAGCAAGGGCGAGGAGG-3′ and 5′-ACTGGAATTCTAGAGTCGCGGCCGCTATTATCAATCCAGTTTCACTCTC-3′, and the Speedy NZYTaq 2× Green Master Mix (cat. nº: MB36201, NzyTech) following the manufacturer’s instructions. The column-purified PCR product and the destination plasmid vector pAAV-EALBAAT-EGFP-PA were subjected to NcoI-HF (cat. nº: R3193L, NEB) and EcoRI-HF (cat. nº: R3101L, NEB) digestion and subsequently ligated using T4 DNA ligase. The expression cassette was then cloned in a backbone plasmid with a mutated ITR to generate the self-complementary rAAV (dsAAV). The sequence of the resulting pAAV and dsAAV-EALBAAT-mCherry-NLS plasmids was verified by Sanger sequencing.
2.2.3. Generation of pMDLg_pRRE-D64A Plasmid for Integration-Deficient Lentivirus (IDLV) Production
The D64A mutation [
43] was introduced by site-directed mutagenesis in the parental pMDLg_pRRE lentivirus packaging plasmid. To do so, a mutagenic PCR was performed using 5′-ATGGCAGCTAGCTTGTACACATTTAG-3′ and 5′-ATTCCTGGGCTACAGTCTAC-3′ primers with Supreme NZYProof DNA polymerase (cat. nº: MB28301, NzyTech) and the PCR product was subsequently circularized with KLD reaction mix (cat. nº. M0554S, NEB) following the manufacturer’s instructions. The sequence of the resulting pMDLg_pRRE-D64A plasmid was verified by Sanger sequencing and was used for the generation of IDLVs as described below.
2.3. Viral Vector Production and Titration
2.3.1. ICLV and IDLV Production and Titration
ICLV and IDLV expressing the mCherry-NLS reporter gene in their genome were produced in HEK-293T cells by co-transfection of all necessary packaging plasmids (pRSV-Rev, pMDLg-pRRE and pMD2.G) as previously described [
44]. Cell supernatants were collected at 40–48 h post-transfection, filtered through 0.45 µm filters, aliquoted and kept at −80 °C until needed. Titration was carried out by inoculation of HepG2-NTCP cells with two-fold serial dilutions of lentiviral particles. Quantitation of the number of and the mean fluorescence intensity signal in mCherry-positive cells was used to determine the amount of each lentiviral particle to be used in the experiments. As expected, ICLV transduction was much more efficient than IDLV transduction.
2.3.2. rAAV Production and Titration
Production of the rAAVs was performed by double transfection in HEK-293T cells using the pAAV plasmid with the expression cassette and a pDP3 plasmid that includes the
rep and
cap AAV genes and the adenovirus genes essential for replication. HEK-293T cells (2% FBS DMEM; 65% cell confluency) were co-transfected using linear polyethyleneimine at 25 kDa (cat. nº: 23966-100, Polysciences, Warrington, PA, USA). Vector particles were obtained from cells and the supernatant. After 72 h, the supernatant was treated with polyethylene glycol solution (PEG8000, 8% final concentration; cat. nº: 89510, Sigma-Aldrich) at 4 °C for 48 h, then centrifuged for 15 min at 3000 rpm, and the viral particles present in the pellet were resuspended in lysis buffer (50 mM Tris-Cl, 150 mM NaCl, 2 mM MgCl
2, 0.1% Triton X-100) and kept at −80 °C. Cells were treated with lysis buffer and kept at −80 °C. Then, both the supernatant and cells were subjected to 3 freeze/thaw cycles, centrifuged and finally, they were treated with DNase and RNase solutions. This lysate was purified in an iodixanol gradient (15, 25, 40 and 54% iodixanol) by ultracentrifugation (69,000 rpm, 16 °C, 2.5 h; in Beckman type 70 Ti rotor in Beckman) according to previously published methods [
45] and concentrated by Ultra-15 mL Amicon columns (cat. nº: C7715, Amicon
®; Millipore, Bedford, MA, USA). AAV vector genomes were extracted using the High Pure Viral Nucleic Acid Kit (cat. nº: 11858874001, Roche, Switzerland) according to the manufacturer’s specifications. The quality of the vector production was checked by two techniques: viral genome quantification by real-time-quantitative PCR (qPCR) assay and capsid protein detection by sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS-PAGE) combined with SYPRO Ruby staining. Real-time-qPCR was performed to quantify the number of AAV genomes (viral genomes (vg)/mL)), using the GoTaq
® qPCR Master Mix (cat. nº: A6001, Promega, Madison, WI, USA) and primers specific for ITRs [
46], while SDS-PAGE was used to analyze the capsid protein ratio. Vectors were stored at −80 °C until use.
2.4. Compound Treatment
HepG2-NTCP cells were seeded at high density in 12-well plate (4 × 105 cells per well) or 96-well plate (4 × 104 cells per well) formats, as indicated in each experiment. The following day, the medium was discarded and freshly prepared 2% DMSO-containing medium was added to the cells and they were further incubated for another twenty-four hours. The next day, cells were treated for one hour with the indicated concentration of compounds diluted in 2% DMSO-containing medium. After the treatment, the medium was discarded, and cells were processed immediately for double-strand break detection or cytotoxicity analysis, or they were subjected to viral transduction or metabolic labeling as indicated below.
2.5. Detection of Double-Strand Breaks by γH2AX Immunofluorescence
HepG2-NTCP cells plated in a 96-well plate format and treated as indicated in each figure legend were subjected to immunofluorescence analysis with specific antibodies to detect the phosphorylated form of H2AX (γH2AX), a marker of double-strand break formation. To do so, the compound-containing medium was discarded right after one hour of treatment. Cells were washed with phosphate-buffered saline (PBS) and incubated in 4% formaldehyde (FA)-PBS solution for twenty minutes at room temperature. Then, they were extensively washed with PBS to remove the fixative and they were processed for immunofluorescence analysis as previously described [
47]. In brief, cells were incubated in blocking buffer (1xPBS, 10% FBS, 3% bovine serum albumin (BSA), 0.3% Triton X-100) for one hour at room temperature, washed three times with PBS and then incubated for one hour with a 1:200 dilution of a rabbit monoclonal antibody against γH2AX (cat. nº: 9718, Cell Signaling Technology, Danvers, MA, USA) prepared in binding buffer (1xPBS, 3% BSA, 0.3% Triton X-100). After extensive washes with PBS, cells were incubated for one hour in a mixture of 2 μg/mL Alexa Fluor
TM 488-conjugated goat anti-rabbit IgG cross-adsorbed secondary antibodies (cat. nº: A-11008, Invitrogen, Waltham, MA, USA) and 0.5 μg/mL Hoechst 33342 dye (cat. nº: H3570, Invitrogen) for nuclei staining. The fluorescence signal was then imaged and pictures were taken in a SparkCyto plate reader (Tecan, Austria) after extensive washes with PBS. Images were analyzed as indicated below.
2.6. Cytotoxicity Analysis
The toxicity of compounds was evaluated by quantitation of metabolic activity using an MTT-formazan assay. HepG2-NTCP cells plated in a 96-well plate format were subjected to compound treatment at concentrations indicated in the corresponding figure legend. One hour later, the culture medium was discarded, cells were washed with complete medium and subjected to MTT assays (cat nº: M2128, Sigma-Aldrich) using previously described procedures [
48].
2.7. Viral Transduction
Appropriate amounts of lentiviral (ICLV or IDLV) and adeno-associated viral (ssAAV or scAAV) stocks diluted in medium were used to inoculate HepG2-NTCP cells. Sixteen hours later, viral inoculum was replaced by freshly prepared 2% DMSO-containing medium and the cells were further incubated at 37 °C for different lengths of time, as indicated in each figure legend. Viral transduction efficiency was determined by mCherry fluorescence analysis and nucleic acid quantitation as described below.
2.8. Visualization of Cellular RNA Synthesis by Metabolic Labeling with Ethyl Uridine
HepG2-NTCP cells were plated at a density of 4 × 104 cells per well in a 96-well plate format. The day after, the medium was replaced by a 2% DMSO-containing complete medium. Twenty-four hours later, cells were treated with 2.5 μM of CPT, 50 μM of Eto, 1.25 μM of actinomycin D (as a control of RNA synthesis inhibition) or vehicle (DMSO; as control) for one hour, after which they were metabolically labeled with 0.5 mM ethyl uridine (EU). After 30 min of labeling, the medium was discarded, cells were washed and the incorporation of EU into the nascent RNA was revealed in situ after cell fixation and through a click-chemistry reaction using the Click-iT RNA HCS Assay kit (cat nº: C10327, Invitrogen), following manufacturer’s instructions. The fluorescence signal was then imaged and pictures were taken in a SparkCyto plate reader as indicated below.
2.9. Fluorescence Image Acquisition and Analysis
HepG2-NTCP cells plated in a 96-well plate format were subjected to compound treatment followed by vector viral transduction as indicated above. At the indicated times after transduction, the culture medium was discarded, cells were washed with PBS and fixed by incubation in 4% FA-PBS solution for ten minutes at room temperature. Cells were then extensively washed with a solution of 0.3% Triton X-100 in PBS for one hour to remove the fixative and they were incubated for 30 min with a 1xPBS solution containing a 0.5 μg/mL Hoechst 33342 dye. Viral vector-derived mCherry, γH2AX (in immunofluorescence experiments), EU labeling and nuclei fluorescence staining were imaged and pictures were taken in a SparkCyto plate reader after extensive washes with PBS. Images of 2456 × 2052 pixels at a 16-bit gray scale were acquired with a 10× objective. All images were taken with exactly the same exposure settings. The Fiji/ImageJ v1.0 software analysis package [
49] was used to quantitate the number of nuclei and mCherry-positive cells, and the mean intensity signal of each image.
2.10. Nucleic Acid Analysis
HepG2-NTCP cells plated in a 12-well plate format were subjected to compound treatment followed by vector viral transduction as indicated above. Forty-eight hours later, the medium was discarded, cells were washed once with PBS and nucleic acids were extracted as explained below.
Total RNA was extracted following the guanidinium isothiocyanate extraction protocol as previously described [
50]. In total, 2 μg of total RNA purified from each sample was subjected to DNase treatment (dsDNase; cat. nº: 15205063, Fisher Scientific, Waltham, MA, USA) in a final volume of 10 μL, following the manufacturer’s instructions. Half of the material was used to quantify mCherry and GAPDH mRNA (for normalization) in each sample in a two-step RT-qPCR assay using the MultiScribe Reverse Transcriptase (cat. nº: 4319983, Applied Biosystems, Waltham, MA, USA) and the PowerTrack
TM SYBR Green qPCR Master Mix (cat. nº: A46109, ThermoFisher Scientific, Waltham, MA, USA). Next, 10-fold serial dilutions of plasmids containing each target sequence were prepared as standard curves to be used with each corresponding pair of primers: mCherry (5′-TTCATGTACGGCTCCAAGGC-3′ and 5′-TGTAGATGAACTCGCCGTCC-3′) and GAPDH (5′-TGGAAGATGGTGATGGGATTT-3′ and 5′-AGGTGAAGGTCGGAGTCAACG-3′). The results were normalized using GAPDH mRNA levels in each sample and were displayed in the figures as the number of copies per 100 ng of total RNA.
Total DNA was extracted using the NZY Tissue gDNA Isolation kit (cat. nº: MB13503, NzyTech) following the manufacturer’s instructions. The RNase treatment step was included right after sample lysis to make sure RNA-free DNA was obtained. Next, 5 ng of purified RNA-free DNA from each sample was used for direct quantitation of total mCherry and GAPDH DNA content (for normalization) by qPCR using the PowerTrackTM SYBR Green qPCR Master Mix. Then, 10-fold serial dilutions of plasmids containing each target sequence were prepared as standard curves to be used with each corresponding pair of primers: mCherry (same primers as the ones used in mCherry RT-qPCR, above) and GAPDH (5′-CATGGTGCCAAGCCGGGAGA-3′ and 5′-GGGTCGGGTCAACGCTAGGC-3′).
DNA integration was analyzed using an in-house optimized two-step PCR protocol. First, 5 ng of purified RNA-free DNA was subjected to a PCR protocol using the Speedy NZYTaq 2× Green Master Mix and an mCherry specific primer (Cherry1Alu: 5′-GTGAGCAAGGGCGAGGAGGATAACATGGC-3′) together with an AluII specific primer (5′-GCCTCCCAAAGTGCTGGGATTACAG-3′). The PCR conditions were: 1 cycle at 95 °C for 1 min; 20 cycles of 2 s at 94 °C, 5 s at 60 °C and 25 s at 72 °C; and 1 cycle at 72 °C for 2 min. This first PCR allows the amplification of mCherry integrated events. Next, 2.5 μL of 1:10 diluted PCR products was used as templates in a nested-qPCR assay using the same qPCR conditions and mCherry primers described above for total mCherry DNA and mCherry mRNA analysis.
2.11. Statistical Analysis
GraphPad Prism v.10.0.0 software was used to analyze the data, prepare the graphs and perform the statistical analysis. All experimental results (immuno-/fluorescence intensity analysis, MTT assays, DNA and RNA analysis) are displayed in the graphs as the mean ± standard deviation (S.D.). Each experiment was performed at least two times in triplicate wells as stated in each figure legend. Normal distribution of the data was confirmed using the Shapiro–Wilk normality test. For data distributed normally, the differences among the means between multiple groups (more than two) were analyzed by one-way ANOVA followed by Dunnett’s multiple comparison test, as indicated in each figure legend. Data not distributed normally were analyzed by a nonparametric Kruskal–Wallis test followed by Dunn’s multiple comparison test, as indicated in each figure legend. The statistical significance was set as: ns, not significant; * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001.
4. Discussion
Gene therapy is a promising strategy to treat and cure inherited metabolic liver disorders [
11]. Expression of the wild-type form of a mutated allele in the affected tissues can correct a disorder and, depending on the disease, achieve life-long benefits. To do so, systemic administration of high vector doses is sometimes required to achieve therapeutic levels in solid organs, such as the liver. This may induce the appearance of undesired adverse events that need to be overcome. This has motivated the search for more effective viral vectors and also the use of small molecules that could act as adjuvants [
32,
33]. However, the use of small molecules may bring with it the appearance of new risks, such as those derived from an increased integration of viral genome used for therapy. This justifies the need to evaluate the risks derived from the use of compounds during viral vector transduction.
In this study, we demonstrated that a very short, one hour treatment with topoisomerase inhibitors prior to viral vector transduction is enough to increase reporter expression of rAAVs as well as rLVs in hepatic cells. These viral vectors enter the cells through different cell entry receptors and express the transgene under different promoters. Since the increased reporter gene expression occurs in both viral vector transductions, we conclude that the effect of topoisomerase inhibitor treatment is independent of the entry pathway, the promoter and the viral vector used. These results suggest that topoisomerase inhibitors used in this study target a common step or feature that is shared in both viral vector transductions. Given that the compounds were added prior to viral vector transduction, it is unlikely that the observed effect is a direct consequence of the compounds on viral particles themselves or on the delivery of their genomes. Instead, we favor the hypothesis that the compounds act directly on the host cell by predisposing it in a way that vector genome expression is enhanced. In this regard, we demonstrated that cell viability and host RNA biogenesis are unaffected by the treatments, ruling out the possibility that the increased viral vector expression is a consequence of toxic effects or increased general cellular gene expression. Instead, both compounds increased the phosphorylation of histone H2AX, a hallmark of DNA damage, indicating that DDR was induced upon these treatments.
Expression from rAAV vectors depends on the formation of genome concatemers and their circularization, which produces stable transcriptionally active structures. Since DDR activation favors these processes [
54], induction of DDR by treatment with topoisomerase inhibitors could explain the increased mCherry intensity signal observed upon rAAV transduction. The magnitude of the effect with Eto treatment was always higher than that of CPT. These differences might be explained by their intrinsic capacity to induce double-strand breaks or they might point to the existence of mechanistic differences between TOP1 and TOP2 in the regulation of viral vector expression. Indeed, TOP1 inhibition by administration of CPT has been proposed to increase ssAAV DNA repair [
55], suggesting that TOP1 negatively regulates the establishment of functional AAV episomal DNA. However, in a different study, treatment with TOP2 inhibitors increased both ssAAV and scAAV transgene expression [
33], suggesting that modulation of AAV expression by TOP2 occurs downstream of ssAAV DNA repair. These results pointed out that inhibition of topoisomerases may have differential effects on specific steps of AAV transduction depending on the inhibitor and/or the experimental conditions used. However, our results, obtained in non-dividing hepatic cell culture conditions, showed that ssAAV and scAAV expression increased in a similar way upon topoisomerase inhibitor treatment, ruling out the possibility that topoisomerase inhibitors regulated rAAV second strand DNA synthesis, a step that is already completed in the scAAV vector. Thus, topoisomerases do not seem to play a role in ssAAV genome repair in hepatic cells, in contrast to what has been previously published in heart-derived cells [
55]. This discrepancy strongly suggests the existence of cell-type and/or tissue-specific differences that affect the efficiency of rAAV transduction, and that strategies aiming to increase rAAV transduction efficiency need to consider these singularities.
In the case of rLV transduction, topoisomerase inhibitors increased both ICLV- and IDLV-derived expression, but with a much stronger increase in the IDLV transduction system. As expected, when we compared both types of lentiviruses, ICLV produced a higher reporter accumulation compared to IDLV transduction. Both LVs produced equivalent amounts of genomic DNA copy numbers; however, mRNA production was enhanced in ICLV compared to IDLV transduction, further supporting the idea that integration is necessary for optimal rLV expression [
20,
21,
22]. For all these reasons, we hypothesized that the IDLV expression increase observed upon topoisomerase inhibitor treatment could be explained by an increase in non-specific and integrase activity independent integration of the IDLV genome, produced as a consequence of the DNA damage induced by the compound treatment. In agreement with that, treatment with CPT and Eto increased viral genome integration in both rLVs with a particularly higher impact on IDLV, supporting the idea that topoisomerase inhibitors increased episomal DNA integration, especially when viral integrase activity is abolished [
35,
38,
39]. The increased viral genome integration correlated with a higher viral mRNA production upon treatment. Interestingly, treatment with topoisomerase inhibitors rescued IDLV-derived reporter mRNA expression to levels comparable to those observed in ICLV-transduced cells. Topoisomerase inhibition did not change the intracellular viral genome DNA copy number, discarding an effect on rLV entry, genome retro-transcription, or DNA homeostasis. This is in contrast to previously reported results derived from HIV-infected 293T cells, where CPT and Eto treatments produced an increase in the accumulation of total reverse transcription products, which are the precursors of 2-LTR episomes and the dsDNA integration template [
38]. This discrepancy suggests that topoisomerases, and therefore, the inhibition of their activity, produce differential cell type-dependent effects that in turn might regulate differential rLV transduction. The smaller increase in viral genome integration observed in ICLV transduction could be explained by the fact that ICLV can also produce non-integrated episomal DNAs as those produced upon IDLV transduction. Thus, the increased integration observed in ICLV transduction upon CPT and etoposide treatment could probably be related to the effect of the compounds on non-integrated viral genomes produced during ICLV transduction.
When comparing both compounds, Eto treatment produced a much higher effect on viral genome integration and viral mRNA production than CPT treatment. These results agree with the fact that Eto induced more DNA damage than CPT in hepatic cells. Thus, our data confirm the notion that the extent of DNA damage induced by these compounds determines the rate of episomal DNA integration and therefore its expression [
39]. This is especially important for IDLV integration, since IDLV does not contain integrase activity. In contrast, ICLV integration, given the functional role of its viral integrase, does not depend on damaged DNA for its integration and expression. This agrees with the fact that ICLV integration barely increased compared to IDLV integration upon compound treatment. Taken together, these results strongly suggest that the increased IDLV viral DNA integration in the host chromosomal genome is the main mechanism by which DNA damage-inducing agents, such as CPT and Eto, produce an increase in IDLV reporter gene expression in hepatic cells.
In summary, our study highlights the need to monitor DNA damage and undesired integration of viral episomal DNAs into the host genome when studying chemical compounds that increase viral transduction. This study has set up the necessary tools for the discovery of new compounds with the potential to increase viral vector transduction efficiency without triggering their integration into the host genome.