Combination of hepatocyte specific delivery and transformation dependent expression of shRNA inducing transcriptional gene silencing of c-Myc promoter in hepatocellular carcinoma cells

Background A specific targeting modality for hepatocellular carcinoma (HCC) could ideally encompass a liver cell specific delivery system of a transcriptional unit that is active only in neoplastic cells. Sendai virosomes, derived from Sendai viral envelopes, home to hepatocytes based on the liver specific expression of asialoglycoprotein receptors (ASGPRs) which are recognized by the Sendai virosomal fusion (F) proteins. As reported earlier by us and other groups, transcriptional gene silencing (TGS) does not require continuous presence of the effector siRNA/shRNA molecule and is heritable, involving epigenetic modifications, leading to long term transcriptional repression. This could be advantageous over conventional gene therapy approaches, since continuous c-Myc inactivation is required to suppress hepatocarcinoma cells. Methods Exploiting such virosomal delivery, the alpha-fetoprotein (AFP) promoter, in combination with various tumour specific enhancers, was used to drive the expression of shRNA directed against ME1a1 binding site of the proto-oncogene c-Myc P2 promoter, in order to induce TGS in neoplastic liver cells. Results The dual specificity achieved by the Sendai virosomal delivery system and the promoter/enhancer guided expression ensured that the shRNA inducing TGS was active only in liver cells that had undergone malignant transformation. Our results indicate that such a bimodal therapeutic system induced specific activation of apoptosis in hepatocarcinoma cells due to heterochromatization and increased DNA methylation of the CpG islands around the target loci. Conclusions The Sendai virosomal delivery system, combined with AFP promoter/enhancer expression machinery, could serve as a generalized mechanism for the expression of genes deleterious to transformed hepatocarcinoma cells. In this system, the epigenetic suppression of c-Myc could have an added advantage for inducing cell death in the targeted cells. Electronic supplementary material The online version of this article (doi:10.1186/1471-2407-14-582) contains supplementary material, which is available to authorized users.


Background
Hepatocellular carcinoma (HCC) is the sixth most prevalent cancer and the third leading cause of worldwide cancer related deaths [1]. Genes of fetal or embryonic origin are often re-expressed in various tumours and alpha-fetoprotein (AFP) expression has been shown to be re-activated in HCC [2]. In cancer gene therapy, the biggest challenges are cell specific targeting and tumour selective expression of the therapeutic gene. A number of reports have raised the issue of specificity and efficiency of gene transfer [3,4], specifically to neoplastic cells. Specificity can be at two levels. Firstly, it could be at the level of delivery to a particular cell type. Usually, cell type specific antibodies/ligand-receptor units can be used [5][6][7]. These include binding to generalized ligands like the transferrin or folate receptors [8][9][10][11] or antibodies to cell surface antigens [12,13]. Sendai virosomes specifically fuse with the hepatocytes through their fusion (F) protein's terminal galactose moiety, that binds specifically with asialoglycoprotein receptors (ASGPRs) present only on the surface of hepatocytes [14]. The Sendai F-virosomal system, lacking the hemagglutinin neuraminidase (HN) protein, is non-toxic and comparatively non-immunogenic. One of us earlier (D.P.S), has successfully demonstrated the expression of human uridinediphosphoglucuronate glucuronosyltransferase-1A1 (hUGT1A1) gene in the hepatocytes of Gunn rats for the treatment of jaundice [15]. The expression of a transgene might be low due to its lysosomal translocation and failure to integrate into the host genome. Sendai F-virosome mediated delivery overcomes this limitation since the entrapped cargo is directly delivered into the cytoplasm, thus evading the endosomal pathway [16,17]. This could enhance transgene expression and its longevity for therapeutic purposes.
Another level of specificity is at the level of tumour specific promoters [18]. This relies on the fact that several genes, including oncofetal genes are expressed upon cell transformation, implying that the activation of such promoters takes place only in the transformed but not in the normal cells. Such neoplasia activated promoters include carcinogenic embryonic antigen (CEA), prostate specific antigen (PSA), L-plastin, osteocalcin, midkine etc. [19]. For liver neoplasms, it has been shown that an AFP promoter could help achieve a HCC-targeted gene therapy [20][21][22]. Often, the tumour specific promoters are weak which can be augmented by utilizing various tumour specific enhancers [23] without compromising the specificity. The 5′ flanking region of the AFP gene consists of several enhancer like sequences [24] where one of the core enhancer region can augment gene expression in an engineered construct [25]. However, the possibility of other enhancers could also be explored.
c-Myc regulates several cellular processes [26] and is crucial for stem cell maintenance [27]. It is also essential for normal growth and proliferation since its inactivation produces lethal effects [28,29], indicating its level has to be tightly regulated. Down-regulation of c-Myc both in vitro and in vivo has been shown to induce growth inhibition and differentiation of HCC [30][31][32]. ME1a1 binding site between P1 and P2 promoter of c-Myc is required for sustenance of transcriptionally active dual c-Myc promoters [33]. Since the P2 promoter is associated with 75-90% of the c-Myc transcripts [26], it serves as an ideal candidate for targeting therapy. We have previously demonstrated that siRNA against c-Myc could induce TGS in glioma cells, leading to increased cell death [34].
In the current study, we have tried to assess the combination of cell type specific delivery and tumour dependent activation for inducing TGS in hepatocellular carcinoma cells. There are no reports of TGS by shRNA driven through a tumour specific promoter delivered by a target specific vehicle. In order to impart strength and specificity to the induction of TGS, we have first generated novel combinations of the AFP promoter with AFP enhancer as well as with the nuclear factor kappa beta (NFκB) response element to drive the expression of shRNA targeting c-Myc P2 promoter. Usually shRNA has been expressed by constitutive polymerase (pol) III promoters [42] which fail to elicit tumour specificity. However, in this study, we have tried to achieve specificity as well as efficiency in transcription by using pol II based AFP promoter along with various enhancer elements. Since one of the key events in hepatic oncogenesis is the constitutive activation of NFκB transcription factor [43] and AFP [44], we have compared enhancer systems from both in our study.
Our results indicate that the Sendai virosomal delivery, combined with the AFP promoter/enhancer driven shRNA system, has the necessary specificity and efficiency to activate TGS in hepatocarcinoma cells, leading to cell death. The combination of both targeting entities is likely to be of an added advantage for cancer therapeutics. for cell sciences (N.C.C.S), Pune. Cells were maintained in Dulbecco's Modified Eagle's Medium (DMEM; Sigma-Aldrich, Germany) supplemented with 10% calf fetal serum (Biowest, USA). The molecular characterization of Chang Liver cells was done before any experimentation (Additional file 1: Figure S1).
Generation of TGS inducing system: AFP promoter/ enhancer +2 c-Myc shRNA 100 pmoles of both sense and antisense oligonucleotides of c-Myc (with pre-added sticky ends; 5′ BamHI and 3′ HindIII) were suspended in 100 μl of 1X Annealing Buffer (10 mM Tris-HCL pH 8.0, 1 mM EDTA and 100 mM NaCl). Oligonucleotides were heated for few minutes in boiling water. Temperature was maintained at 95-100°C for 5 minutes and was allowed to cool over night to room temperature. Agarose Gel Electrophoresis was performed to analyze and excise annealed oligonucleotides.
Sequence of c-Myc promoter region is shown in Additional file 2: Figure S3. shRNA targeting c-Myc P2 promoter was designed using Invivogen's online siRNA wizard (http://www.sirnawizard.com/construct.php) and chemically synthesized by Integrated DNA Technologies, USA. Sequence of the test and scrambled (Scr; control) shRNA is enlisted in Table 1. AFP Promoter, AFP enhancer -AFP promoter and NFκB response element -AFP promoter regions were amplified up to +2 bp relative to the transcription start site (TSS) from the previously generated luciferase reporter constructs (AFPPr + 25luc; AFPEn-Pr + 25luc; NFκBEn-Pr + 25luc respectively) with primers containing 5′ EcoRI and 3′ BamHI restriction sites. Amplification up to +2 bp would minimize sense strand and ensure efficient processing of shRNA by RNAi machinery [46]. These fusion constructs were cloned along with test c-Myc shRNA (5′ BamHI and 3′ HindIII sticky overhangs) in shRNA expression vector pSilencer 4.1 (Ambion, USA). The generated constructs are shown in Figure 2B and were as follows: AFP promoter +2 -c-myc shRNA (AFPPr + 2myc), AFP enhancer -AFP promoter -c-myc shRNA (AFPEn-Pr + 2myc) and NFκB responsive element -AFP promoter -c-myc shRNA (NFκBEn-Pr + 2myc). Likewise scrambled c-Myc shRNA was cloned downstream to the same promoter/enhancer constructs (AFPPr + 2myc Scr; AFPEn-Pr + 2myc Scr; NFκBEn-Pr + 2myc Scr). c-Myc test and scrambled shRNA were also cloned under cytomegalovirus (CMV) promoter (CMVPrmyc and CMVPrmyc Scr respectively) where CMVPrmyc served as a positive control. Annealing of oligonucleotides and schematic representation of all the clones are shown in Additional file 2: Figure S4-S6. All the clones were confirmed by restriction digestion and further authenticated by DNA sequencing from professional agencies.

Transfection
Cells were plated at 10 5 cells per well in a six-well plate, 3 × 10 5 cells per 25 cm 2 flask or 10 6 cells per 75 cm 2 flask (Corning, USA). Twenty-four hours later, they were transfected with different reporter or shRNA constructs using Lipofectamine™ 2000 (Invitrogen, USA) as per the manufacturer's protocol.

Dual luciferase assay
All the three constructs: AFPPr + 25luc, AFPEn-Pr + 25luc and NFκBEn-Pr + 25luc ( Figure 1A) were transfected in HepG2, Huh7, Chang Liver and CHO cells. After 48 hours, transactivation study was done by Dual Luciferase Assay (Promega, USA) following manufacturer's protocol. The firefly luciferase activity was normalized against Renilla luciferase activity and expressed relative to promoter-less pGl3-Basic control vector. Results are representative of three independent sets of experiments.
Quantitative RT-PCR to evaluate c-Myc down-regulation and shRNA expression  geometric mean of multiple internal control reference genes [47] such as β-Actin, 18S, GAPDH and relative expression was estimated by Relative Expression Software Tool (REST; [48]). Primers utilized are given in Additional file 3: Table S1.
Similarly, for c-Myc shRNA quantitation, 1st-Strand cDNA Synthesis Kit (Agilent) was utilized, as per the manufacturer's protocol, and siRNA expression was estimated by RT-PCR. Custom made c-Myc siRNA specific primer was obtained separately (sequence in the Additional file 3: Table S1). Luciferase shRNA under CMV promoter (CMV -luc shRNA) served as a control.
Cell survival assay 2 × 10 4 cells seeded on to 24 well plates (Corning, USA) were transfected with various AFP promoter/enhancer driven shRNA constructs or their respective scrambled controls. On the 6 th day, cells were subjected to MTT (Sigma-Aldrich, Germany) assay for percent cell survival. Furthermore, cell survival was also evaluated by cell counting, post staining with trypan blue (Sigma-Aldrich, Germany), by following manufacturer's protocol.
Apoptosis study 10 5 cells were seeded in 25 cm 2 cell culture flask (Corning, USA) followed by transfection with various shRNA constructs. On the 6 th day, cells were fixed overnight in 70% ice-cold ethanol. Staining of cells was done using Propidium Iodide (PI; Sigma-Aldrich, Germany) and fluorescence was captured using Flow Cytometer (BD Biosciences, USA). Percentage of apoptotic cells (subG1) and other cell cycle phases were estimated using WinMDI software (http://winmdi.software.informer. com/2.8/).

Western blotting
On the 6 th day, post transfection/virosomal delivery of various c-Myc shRNA constructs, cell lysates were prepared using triple lysis buffer and protein was estimated by Pierce BCA Protein Assay Kit (Thermo-scientific, USA). Proteins were run on 5% to 12% SDS-PAGE gels and electro transferred to nitrocellulose membranes (Bio-Rad, USA). Blocking was done with 4% bovine serum albumin (Sigma-Aldrich, Germany) and Immunoblotting was done with requisite primary antibodies: antiactin (sc-8432), anti-c-Myc (9E10), anti-TERT (sc-377511) and anti-cyclin D3 (sc-6283); from SantaCruz Biotechnology, USA. Detection of specific proteins was done with horseradish peroxidase (HRP) conjugated secondary antibodies using ECL detection system (Applied Biosystems, USA).

Sendai virus culture
Sendai virus (Z strain) was grown in 10 -11 day old embryonated chicken eggs, and extracted by utilizing procedure described in our previous report [16].

Generation of Sendai fusion (F) virosomes and R18 labeling
Sendai F-virosomes were prepared as described earlier [14]. For Octadecyl Rhodamine B Chloride (R18; Invitrogen, USA) labeling, F-virosome (1 mg/ml) suspension was labeled by adding 10 μl ethanolic soution (1 mg/ml) of R18 in falcon tube while vortex mixing. The mixture was incubated in dark at room temperature for 30 minutes. Excess unbound R18 was removed by ultracentrifugation at 1,00,000 g for 1 hour at 4°C. The pellet was resuspended in 10 mM phosphate buffered saline (PBS).

Study of live cell fusion: kinetics of F-virosomes
A measure of Sendai virus fusion with HepG2, Huh7, Chang Liver and CHO cells was done using R18 labeled F-virosomes. Heat inactivation of virosomal F-proteins was performed using procedure described in our earlier reports [16,17]. HepG2, Huh7, Chang Liver and CHO cells (1 × 10 6 cells) were incubated with 2 mg of R18 labeled F-virosomes for 1 hour at 4°C. After incubation, cells were centrifuged at 2000 rpm for 5 minutes to remove unbound virosomes. The pellet was suspended in 100 μl of cold 10 mM PBS. For measuring fusion kinetics, 50 μl of the labeled F-virosome-cell complex was added in a cuvette having 3 ml of PBS with 1.5 mM Ca 2+ (pre-warmed to 37°C). Fusion kinetics was studied by a spectrofluorimeter (FL3-22; Horiba, USA). For data normalization, percent fluorescence dequenching (% FDQ) at a time point was calculated as per the equation: % FDQ = [(F-F 0 )/F t -F 0 )] x 100 where F 0 denotes fluorescence intensity at time point zero, F is the intensity at a given time point and F t is the intensity recorded when 0.1% Triton X-100 was added to the cell-virosome complex and is designated as fluorescence at "infinite" dilution of the probe (100%). Packaging and delivery of AFP promoter/enhancer +2 c-Myc shRNA constructs by Sendai F-virosomes 50 mg of Sendai virus envelope was reduced with 3 mM Dithiothreitol (DTT) at 37°C. Viral genetic material and HN were removed from the virosomal suspension by treatment with non-ionic detergent Triton X-100 for 1 hour followed by ultra-centrifugation. From this detergent extract, supernatant was recovered and mixed with required amount of various AFP promoter/enhancer driven c-Myc shRNA plasmids. This mixture was reconstituted by step-wise removal of detergent by utilizing SM-2 Biobeads (Bio-Rad, USA). Packaging of shRNA plasmids was confirmed by SDS based lysis and running the contents on 0.8% agarose gel. Cells were plated at 10 5 cells per well in a six-well plate, 3 × 10 5 cells per 25 cm 2 flask, or 10 6 cells per 75 cm 2 flask (Corning, USA) followed by transfection with c-Myc shRNA loaded F-virosomes.

CpG methylation study: bisulfite PCR and sequencing
Following virosomal delivery of AFPEn-Pr + 2myc or its scrambled control in HepG2 cells, genomic DNA was isolated on the 6 th day using Gen Elute Mammalian genomic DNA Miniprep Kit (Sigma-Aldrich, Germany). Bisulfite PCR was done using Epi Tech Bisulfite Kit (Qiagen, Germany) as per the manufacturer's protocol. http://bisearch.enzim.hu/ was utilized for designing specific primers. Primers were M13-tagged for sequencing of PCR products.

Assessment of heterochromatization by chromatin immunoprecipitation (ChIP) assay
Chromatin immunoprecipitation (ChIP) assay for H3K9Me2 and H3K27Me3 was done using EZ ChIP kit (Millipore, USA) as per manufacturer's protocol. Input DNA, anti-H3K9Me2 (mAbcam1220), anti-H3K27Me3 (mAbcam6002), anti-histone 3 acetylated (Upstate) and control mouse IgG antibody (Upstate) immunoprecipitated DNA was amplified using primers specific for the target region on the c-Myc P2 promoter listed in Additional file 3: Table S1. Immunoprecipitation percentage was calculated as described by Haring et al. [49]. Centrosome of chromosome 16 served as a positive control, since it has 100% methylated histone tails.
Caspase 3/7 assay for evaluation of apoptosis after virosomal delivery of shRNA Caspase 3/7 activity of HepG2, Huh7 and Chang Liver cell lines was measured post virosomal delivery of AFPEn-Pr + 2myc or AFPEn-Pr + 2myc Scr by using caspase 3/7 assay kit (Promega, USA) as per manufacturer's protocol.

Statistical analysis
All experiments including dual luciferase assay, cell survival assays and RT-PCR was repeated thrice and performed in triplicates. Western blotting, virosome fluorescence dequenching assay, Flow cytometric analysis, Bisulfite PCR, ChIP assay and capase 3/7 assay were repeated at least twice. Student's t-test was utilized to calculate the significance in all experiments and p < 0.05 was considered significant whereas p < 0.001 as highly significant. The data are shown as mean ± SD.

AFP promoter/enhancer mediated expression is hepatocarcinoma specific
The generated luciferase constructs were transfected in both transformed and untransformed cell lines and their proficiency was determined by dual luciferase assay after 48 hours. In the transformed HCC cells, HepG2 and Huh7, the luciferase activity was highest with AFPEn-Pr + 25luc followed by NFκBEn-Pr + 25luc and lastly by AFPPr + 25luc, indicating the relative activity of the AFPEn-Pr + 25 in the transformed cells is significantly higher than SV40 promoter ( Figure 1B and C). However, in the untransformed Chang Liver and nonhepatic CHO cells, significant activity was observed only with SV40luc and not in case of AFP promoter/ enhancer constructs ( Figure 1D and E).

Decrease in c-Myc level by TGS inducing shRNA
Various c-Myc shRNA constructs, against c-Myc P2 promoter (Figure 2A), were generated as described in methods ( Figure 2B). AFPEn-Pr + 2myc and AFPEn-Pr + 2myc Scr were transfected in HepG2 cells and fall in the expression of c-Myc was evaluated consecutively for 6 days by RT-PCR ( Figure 2C). The decrease in c-Myc mRNA level was significant at all-time points (p < 0.05) with respect to its control and was maximum on the 5 th day. Slight apparent increase on the 6 th day when compared to that of 5 th day was insignificant (p = 0.25). Similarly, fall in the c-Myc expression, by other shRNA constructs was also evaluated 5 days post transfection in HepG2 cells ( Figure 2D). Similar results were observed for Huh7 cells ( Figure 2E). However, the absolute levels of c-Myc were higher in HepG2 as compared to Huh7. No significant decrease in c-Myc was observed in the Chang Liver and CHO cells (p > 0.05 for both; Figure 2F and G). The levels of c-Myc protein ( Figure 2H) corroborated with mRNA data but the tissue non-specific CMV promoter driven c-Myc shRNA (CMVPrmyc) decreased the level of c-Myc even in Chang Liver and CHO cells (p < 0.001 for both; Figure 2F and G).

TGS of c-Myc reduced cell survival and increased apoptosis
To examine whether decrease in the expression of c-Myc, by TGS, affects cell growth, both cell survival and apoptosis were evaluated. MTT assay, on the 6 th day post shRNA transfection, revealed decrease in cell survival of the transformed cell line HepG2 and Huh7 (p < 0.05 for both; Figure 3A and Additional file 4: Figure S8), however, Huh7 cells were less responsive to c-Myc knockdown. No such inhibitory effects were observed in the case of untransformed Chang Liver cell line. On the other hand, CMVPrmyc showed significant cell killing and suppression even in Chang Liver cells due to its non-specific nature (p = 0.019; Figure 3B). Cell survival, of HepG2, Huh7 and Chang Liver cells, was further estimated by trypan blue staining followed by cell counting, which corroborated with the MTT data ( Figure 4 and Additional file 4: Figure S9; p < 0.05). Any molecular analysis beyond 6 days was not possible due to complete detachment of HepG2 cells treated with the AFPEn-Pr + 2myc when compared to the scrambled control. Flow cytometric studies by PI staining showed that the percentage of apoptotic cells (sub G1 proportion) in HepG2 was in concordance with the strength of the AFP promoter/enhancer constructs driving the shRNA expression ( Figure 5A). Similar trend was observed in the case of Huh7 cells but to a lesser degree (Additional file 4: Figure S10). Significant apoptosis in Chang Liver was seen only by CMVPrmyc and not by any of the AFP promoter/enhancer mediated c-Myc shRNA constructs ( Figure 5B). c-Myc suppressed cells (HepG2 and Huh7), in addition to apoptosis (subG1 proportion), were found to be within the G0-G1 phase with decreased S and G2M phase. Suppression of c-Myc, by TGS, had a profound effect on the cell survival and apoptosis of HepG2 cells when compared with that of Huh7.

Specific binding of Sendai F-virosomes to cells of liver origin
Once the specificity of c-Myc suppression in HCC cell lines was established, we aimed to increase the level of specificity further by packaging the AFP promoter/ enhancer shRNA constructs within the Sendai virosomes for liver specific delivery. Real time fusion kinetics by fluorescence dequenching assay revealed that Sendai Fvirosomes bind specifically to hepatic cells (HepG2, Huh7 and Chang Liver) and not with control non hepatic cell line CHO. Virosomes with inactivated F-proteins (HC: Heat control), displayed poor fusion even with HepG2 cells, confirming the specific fusion via F-protein and ASGPR of the hepatocytes ( Figure 6A). The difference in the fusion observed might be dependent upon the number of ASGPRs expressed by various cell types.
Once significant fusion was confirmed, the generated constructs were packaged and delivered by Sendai Fvirosomes to both transformed and untransformed liver cells. Time dependent fall in the c-Myc level post virosomal delivery in HepG2 cells ( Figure 6B) was highly comparable to that by conventional method ( Figure 2B). Maximum suppression of c-Myc was observed on the 5 th day with AFPEn-Pr + 2myc and slight increase on the 6 th day when compared to the 5 th was insignificant (p = 0.41). Significant fall in the expression of c-Myc mRNA was seen both in HepG2 and Huh7 by other AFP promoter/enhancer constructs, (p < 0.05 for both; Figure 6C and D). Even though the fluorescence dequenching experiments demonstrated fusion of F-virosomes with Chang Liver, TGS was not effective in these cells due to inactivation of AFP promoter/enhancer system ( Figure 6E). Decrease in c-Myc protein levels were in concordance with its mRNA levels ( Figure 6F).

No interferon response is mounted by c-Myc shRNA
Entry of dsRNA into the cell might lead to non-specific interferon (IFN) responses [50] which involves the activation of the PKR/RNase L pathway ultimately inducing an IFN marker 2,5-oligoadenylate synthetase 1 (OAS1) [51]. There was no significant induction of OAS1 in HepG2, Huh7 and Chang Liver cells (p > 0.05 at all points) post 5 days of shRNA delivery through F-virosomes; indicating the absence of an IFN response ( Figure 7A). Furthermore, no significant increase in the levels of OAS1 was observed in 24, 48, 72 and 96 hours (p > 0.05 at all points) after similar treatment of HepG2 cells (Additional file 4: Figure S11), ruling out IFN response being generated even at earlier time points following F-virosomal delivery of the entrapped shRNA constructs.

c-Myc inactivation caused down-regulation of other proliferative genes
c-Myc regulates growth and proliferation by regulating various genes [26]. Cyclin D3 as well as human telomerase reverse transcriptase (hTERT) were studied in HepG2 cells at both mRNA and protein level. Fall in c-Myc by F-virosomes loaded with AFPEn-Pr + 2myc led to significant decrease in Cyclin D3 and hTERT both at mRNA (p = 0.0022 and p < 0.001) and protein levels, suggesting the down-regulation of c-Myc effector molecules ( Figure 7B and C).

Increase in caspase 3/7 activity following TGS of c-Myc
To validate the activation of apoptosis after c-Myc suppression by chimeric AFP promoter driven c-Myc shRNA, caspase 3/7 activity was evaluated in HepG2, Huh7 and Chang Liver cell lines, 5 days after virosomal delivery of AFPEn-Pr + 2myc ( Figure 7D). The increase in caspase activity was in agreement with the magnitude of chimeric AFP promoter driving the shRNA. In HepG2 cells, significant increase in caspase 3/7 activity was observed (p = 0.005) as compared to its scrambled control, however, activation of caspase 3/7 was to a lesser degree in Huh7 (p = 0.035). No increase in the activity was seen in Chang Liver cells (p = 0.38).

shRNA induced TGS by chromatin condensation and CpG methylation of c-Myc P2 promoter
To evaluate the mechanism by which shRNA acted on the target region, the chromatin status of the c-Myc P2 promoter was evaluated by ChIP assay, on the 6 th day, post virosomal delivery of the AFPEn-Pr + 2myc construct in HepG2 cells. ChIP followed by quantitative RT-PCR revealed that c-Myc shRNA mediated TGS was associated with H3K9 dimethylation and H3K27 trimethylation. Cells pre-treated with HDAC inhibitor TSA showed reduced enrichment of histone chromatin marks even in the presence of AFPEn-Pr + 2myc. This indicated the likely involvement of HDACs in gene silencing of c-Myc ( Figure 8A). Similarly, we checked the acetylation status of the target region following AFPEn-Pr + 2myc transfection, by utilizing anti-histone 3 acetylated antibodies. Significant decrease in the acetylation level was observed post c-Myc suppression on day 6 ( Figure 8B; p = 0.016). However, in the presence of TSA, no decrease was observed (p > 0.05) as the shRNA failed to recruit HDACs.
Furthermore, the methylation status of CpG islands was checked by bisulfite PCR followed by DNA sequencing. Methylation of CpG 8, 9 and 10, when compared to scrambled control, was observed in the test shRNA treated cells ( Figure 8C and D). Moreover, such effect was abrogated by pre-treatment of HepG2 cells with DNMT inhibitor AZA, confirming the possible recruitment of DNMTs, by shRNA, to the target site ( Figure 8E).
We also determined the effect of TSA/AZA or both in combination on c-Myc transcription in HepG2 cells by RT-PCR. Cells pre-treated with both AZA and TSA showed no significant decrease in c-Myc levels by AFPEn-Pr + 2myc on the 6 th day after treatment. Additionally, when the cells were pretreated with AZA or TSA individually, AFPEn-Pr + 2myc down-regulated c-Myc levels significantly, indicating that both HDACs and DNMTs are involved in gene silencing of c-Myc (Additional file 5: Figure S12).
It is known that TGS can continue for a significant number of days after transfection [39,52,53]. In this study, we performed real time PCR to study the dynamics of c-Myc mRNA as well as shRNA expression after transient transfection of various shRNA constructs in HepG2 cells. For AFPEn-Pr + 2myc construct, shRNA was maximally expressed after 48 hours while declining to around 18% of the maximum on day 6 (Additional file 5:   Figure S13). On day 6, c-Myc mRNA was continued to be suppressed ( Figure 2C and 6B) and all the molecular markers of TGS were present (Figure 8). On day 7, almost all the cells detached from the culture plate due to extensive cell death, making it impossible to do any mRNA/ shRNA quantitation. This supports the possibility that TGS continues even after the reduction of shRNA, even though because of cell death, we were unable to reach zero expression.

Discussion
Specificity is the cornerstone of cancer therapy and a considerable part of the current research on cancer therapeutics tries to address this issue in the context of efficacy. In this study, we have tried to combine modalities for achieving specificity at two levelsthat of the delivery system as well as the transcription of its cargo. This approach has been utilized for the expression of shRNA for inducing the suppression of c-Myc by TGS. Although majority of the c-Myc transcripts are P2 promoter driven [26], targeting approaches are hindered by the lack of specificity. Since c-Myc is required for normal growth and proliferation, its non specific suppression might lead to hazarduous effects [54].
Sendai virosomes are naturally hepatotropic in nature because of their internalization through the ASGPRs of hepatocytes [14]. One of us has earlier described their properties both in vitro and in vivo and has used this system for gene delivery to hepatocytes in the Gunn rat model with good efficacy [15]. Sendai virosomes were shown to have high degree of direct cytoplasmic delivery with low immunogenicity [15][16][17].
At the second level of specificity we have tried to use liver tumour specific AFP promoter based fusion constructs. The AFP promoter has been used earlier to drive specific genes, mostly apoptotic or pro-drug metabolizing enzymes in hepatoma cells [55][56][57][58][59]. However, in our study, we have taken the minimal AFP promoter and added upstream enhancer regions from the AFP gene itself and, in another construct, the NFκB response element. This was done to increase the extent of promoter specific gene expression. Our studies showed that the AFP promoter fused with AFP enhancer (AFPEn-Pr + 25), had the strongest and specific expression in HCC cells.
As demonstrated by Dual Luciferase Assay, various AFP promoter based enhancer sytems specifically and optimally expressed luciferase in hepatoma models HepG2 and Huh7 but not in untransformed Chang Liver and non liver CHO cells ( Figure 1B-E). Only the positive control construct (SV40luc) expressed luciferase in both Chang Liver and CHO cells because of its nonspecific nature ( Figure 1D and E).
The specially designed AFP promoter/enhancer driven c-Myc shRNA encompassing ME1a1 site upstream of c-Myc P2 promoter resulted in reduced c-Myc expression only in transformed hepatocarcinoma cells ( Figure 2D and E). However, due to its universal nature, CMVPrmyc decreased the level of c-Myc even in Chang Liver and CHO cells ( Figure 2F and G). The suppression of c-Myc in transformed cells was in concordance with the strength of each construct ( Figure 2D-G). The AFP Enhancer -AFP promoter construct was equivalent in strength to the known constitutive viral promoter CMV and stronger than SV40, while retaining specificity for HCC cells. However, Huh7 having lower basal level of c-Myc compared to HepG2 was less responsive to c-Myc suppression. Previous studies have shown that c-Myc could abrogate the p53-induced cell-cycle arrest [60], and it is possible that HepG2 cells, which contain wild-type p53 compared to mutant p53 in Huh7 [61], were more sensitive to c-Myc suppression. Additionally, increased activity of Wnt/β-catenin pathway in HepG2 than Huh7, which is a direct regulator of c-Myc [62], also might add on to the greater c-Myc level and its consequent implications in HepG2.
c-Myc suppressed cells showed decreased cell survival and increased apoptosis, as evaluated by MTT Assay and Flow Cytometric analysis respectively (Figures 3 and 5). Moreover, cell survival estimated by trypan blue cell counting corroborated with the MTT data ( Figure 4). This was concordant with the strength of promoter/ enhancer construct driving shRNA expression. The effect on HepG2 cells (Figures 3A, 4A and 5A) were more pronounced than that of Huh7 cells (Additional file 4: Figure S8, S9 and S10). However, no decrease in cell viability was observed in the case of Chang Liver cells as the AFP promoter based system was inactive in these cells (Figures 3B and 4B). Moreover, the specificity for transformed hepatocytes was clear as c-Myc shRNA under the CMV promoter induced apoptosis even in Chang Liver cells ( Figure 5B). Due to c-Myc suppression, via TGS, majority of the transformed cells were found to be present within the subG1 phase followed by G0-G1 phase.
Since the use of antisense oligonucleotides or siRNA/ shRNA is potentially limited by ineffective delivery into cancer cells [63], to ensure specific and substantial level of therapeutic entry, shRNA constructs were packaged and delivered to various cell lines through Sendai F-virosomal system. Post virosomal delivery, the reduction in the level of c-Myc was significantly comparable to that by conventional transfection reagent (Figures 2 and 6).
c-Myc shRNA did not induce IFN response since there was no significant increase in the level of IFN marker OAS1 in HepG2, Huh7 and Chang Liver cells, post 5 days of virosomal delivery ( Figure 7A) as well as at earlier time points (up to 4 days; Additional file 4: Figure S11 ). Some of the c-Myc effector molecules are hTERT [64] and Cyclin D3 [65]. Although hTERT is not oncogenic per se, the activation of hTERT is essential for maintaining neoplastic transformation [66]. Following virosomal delivery of AFPEn-Pr + 2myc, significant decrease in hTERT and Cyclin D3 mRNA and protein was observed in HepG2 cells following c-Myc suppression ( Figure 7B and C). Furthermore, the more pronounced increase in caspase 3/7 activity in HepG2 and not in Chang Liver was in agreement with Flow cytometric studies ( Figure 7D).
Earlier reports of TGS have shown that silencing occurs through histone modifications [67][68][69], CpG methylation [70,71] or interference of RNA polymerase binding [52]. In our case, we could demonstrate the induction of TGS by both heterochromatization and DNA methylation.
Previously, other groups have targeted different regions of c-Myc promoter by siRNAs. siRNA against c-Myc transcription start site has shown promising results in suppressing prostate cancer cells, for a longer duration, by interfering with the binding of RNA polymerase [52]. The same group has recently shown an effective strategy in suppressing prostate cancer stem cells, with good efficacy, both in culture and in mouse model through the promoter directed siRNAs [72]. Small molecule inhibitor of c-Myc has proved useful in suppressing as well as chemo sensitizing HepG2 cells towards conventional drugs [73]. Additionally, several reports have demonstrated that suppression in c-Myc levels induces shrinkage in tumour volume [30][31][32].
In published literature, it is indicated that while PTGS would require sustained presence of the effector siRNA molecule, TGS would be long lasting, by virtue of its capability to induce heritable epigenetic changes [37,74]. Hence PTGS would also work in this cell specific promoter/delivery system albeit possibly for a shorter duration. However, we have not demonstrated the same experimentally. There is a report that after 7 days of continuous induction by siRNA against human ubiquitin c gene's (UbC) promoter, TGS persisted for over a month [39]. In a recent study from our lab, TGS of HIV clade C LTR was shown to be effective for at least 21 days after siRNA transfection [53].
In our study, we could follow the expression of c-Myc mRNA and shRNA for only 6 days after transfection ( Figures 2C and 6B and Additional file 5: Figure S13). Extensive cell death, of HepG2 cells, prevented us from quantifying mRNA and shRNA levels on day 7 and beyond. While on day 6, shRNA levels were around 18% of the maximum (on day 2), the molecular markers of TGS were observed to be sustained (Figure 8). This indicates the possibility of a long term sustainability of TGS, even when shRNA levels have declined, although the persistence of TGS in the absolute absence of shRNA could not be determined because of the extensive cell death on day 7. In this study we observed that TGS could result in the reduction of c-Myc for up to 6 days after single transfection.
By ChIP assay and bisulfite PCR/DNA sequencing, we demonstrated that the shRNA induces both histone and DNA methylation in HepG2 cells, which is accompanied by reduced c-Myc promoter acetylation (Figure 8). This was also confirmed by RT-PCR, since the test shRNA failed to decrease c-Myc transcript levels significantly in cells pretreated with both AZA and TSA (Additional file 5: Figure S12). In our earlier report, we were successfully able to induce TGS in glioma cell line U87 and this was shown to be by DNA methylation [34]. The current study is based on HCC cells and involves both heterochromatization and DNA methylation. It is possible that the variation in HDAC involvement is related to the cell type. As the primary message in the paper is related to the internalization of cargo via the ASGPRs, we have not explored the subtle differences in the mechanism of c-Myc TGS in this study.
Here we have demonstrated two levels of specificity by combining a liver cell specific delivery system with a hepatocarcinoma specific promoter/enhancer system. The effector arm of the system is the shRNA inducing TGS of c-Myc. With this we have been able to demonstrate silencing of the c-Myc, specifically in transformed liver cells, leading to extensive cell death. It is expected that combined cell delivery/transformation specific gene expression system, would be a prototype for therapeutic gene delivery in transformed cells. The shRNA inducing TGS of c-Myc, would also serve as an effective mechanism for inducing cell death in the targeted cells.

Conclusions
The dual specificity resulting from Sendai F-virosomal delivery and tumour specific activation offers a novel mode of targeting HCC at two levels, first by targeted liver cell specific delivery and secondly by promoter/enhancer driven expression only in transformed hepatocarcinoma cells. Such approaches might also be utilized for other therapeutic modalities that are based on specific gene transcription e.g. Gene dependent enzyme prodrug therapy (GDEPT). shRNA induced suppression of c-Myc expression by TGS is a possible gene therapy modality that could be utilized in such a delivery system. In the long run, such a targeting system may also be considered for introducing specific genes for expression in the embryonic liver or putting a check on recalcitrant cancer cells with deregulated c-Myc.   Figure S3. Sequence of c-Myc P2 promoter with siRNA target site and CpG islands. GAACGGAGGGAGGGATCGCGCT is the siRNA target site for P2 promoter of c-myc and contain CpG sites 8, 9 and 10. CpG sites are highlighted in red. TATAAAAG represents the TATA box. Figure S4. AFP promoterc-myc shRNA (AFPPr + 2myc) clone. (A) Schematic representation for cloning of AFPPr + 2myc construct. (B) Annealed c-myc sense and antisense oligos. (C) AFPPr + 2myc clone confirmation by digestion with EcoRI and HindIII restriction enzymes. Figure S5. NFκBc-myc shRNA (NFκBEn-Pr + 2myc) clone. (A) Cloning strategy followed for the generation of this construct. (B) NFκB conjugated c-myc shRNA was confirmed by EcoRI and HindIII digestion. Figure S6. AFP enhancer -AFP promoterc-myc shRNA (AFPEn-Pr + 2myc) clone. (A) Cloning strategy followed for the generation of this construct. (B) AFP enhancer and promoter conjugated c-myc shRNA was confirmed by EcoRI and HindIII digestion.
Additional file 3: Table S1. Sequence of the primers used in the study.
Additional file 4: Figure S7. Sequences of AFP promoter, enhancer and NFκB response element used in the study. (A) AFP Promoter sequence from -230 to +25 bp. (B) AFP Enhancer. (C) Sequence of NFκB responsive element (4 x 10 copies). Figure S8. Cell survival of Huh7 cells, by MTT assay, following TGS of c-Myc. Following c-Myc suppression, Huh7 cells showed decreased cell survival but to a lesser degree when compared to that of HepG2. Figure S9. Cell survival of Huh7 cells, by Trypan Blue based cell counting, post c-Myc shRNA treatment. On the 6th day post transfection of all AFP promoter/enhancer driven c-Myc shRNA constructs, the decrease in cell survival of Huh7 corroborated with the MTT assay (p < 0.05). Figure S10. Evaluation of apoptosis in Huh7 cells by flow cytometry. Percentage of apoptotic cells, after c-Myc suppression via TGS, was dependent upon the strength of each construct driving shRNA expression. Figure S11. Evaluation of Interferon response, in HepG2 cells, at various time point post F-virosomal delivery of c-Myc shRNA constructs. No significant increase in the levels of OAS1 was observed in 24, 48, 72 and 96 hours post virosomal delivery of the entrapped shRNA plasmids (p > 0.05 at all points).
Additional file 5: Figure S12. Evaluation of c-Myc levels in HepG2 cells, pretreated with AZA/TSA or both in combination, followed by c-Myc shRNA transfection. HepG2 cells pretreated with TSA/AZA or both simultaneously were transfected with AFPEn -Pr + 2myc and AFPEn-Pr + 2myc Scr. On the 6 th day, real time PCR was done to evaluate the c-Myc transcript levels. Significant decrease in the c-Myc levels were observed in both AZA + AFPEn -Pr + 2myc and TSA + AFPEn -Pr + 2myc treated HepG2 cells (p < 0.05 for both). Combined treatment of both AZA + TSA along with AFPEn -Pr + 2myc showed no decrease in c-Myc levels (p > 0.05). This confirmed that shRNA induces recruitment of both HDACs and DNMTs which play their part in c-Myc down-regulation. Figure S13. Determination of shRNA expression in HepG2 cells at various time intervals by RT-PCR. c-Myc shRNA expression level was determined at various time points post transfection of c-Myc shRNA constructs. The expression of shRNA, by AFPEn -Pr + 2myc, was found to be maximum in 48 hours. The expression decreased significantly with time and was the lowest on day 6 (18% of the maximum on day 2; p < 0.05). shRNA, against luciferase mRNA, driven by CMV promoter (CMVPrluc shRNA) was utilized as a control on day 1.