Potential conflict of interest: Nothing to report.
Ground glass hepatocytes (GGHs) harboring hepatitis B virus (HBV) pre-S mutants have been recognized as precursor lesions of hepatocellular carcinoma (HCC). Previously, we observed the activation of mammalian target of rapamycin (mTOR) in GGHs and HCCs, together with a decreased expression of HBV surface antigen (HBsAg) in HCC tissues. It is, therefore, hypothesized that the activation of mTOR during HBV tumorigenesis may potentially down-regulate HBsAg expression. In this study, we verified an inverse relationship between the expression of HBsAg and phosphorylated mTOR (p-mTOR) in 13 of 20 paired nontumorous liver and HCC tissues. In vitro, wild-type or mutant pre-S proteins could activate mTOR in the HuH-7 cell line. Interestingly, the up-regulated mTOR, in turn, suppressed HBsAg synthesis at the transcriptional level via the transcription factor, Yin Yang 1 (YY1), which bound to nucleotide 2812-2816 of the pre-S1 promoter. This inhibitory effect by the mTOR signal could be abolished by the knockdown of histone deacetylase 1 (HDAC1). Furthermore, YY1 was physically associated with HDAC1 in a manner dependent on mTOR activation. Collectively, pre-S protein-induced mTOR activation may recruit the YY1-HDAC1 complex to feedback suppress transcription from the pre-S1 promoter. Conclusion: The activation of mTOR signal in GGHs may feedback suppress HBsAg synthesis during HBV tumorigenesis and explain the observed decrease or absence of HBsAg in HCC tissues. Therapy using mTOR inhibitors for HCCs may potentially activate HBV replication in patients with chronic HBV infection. (HEPATOLOGY 2011 )
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Chronic hepatitis B virus (HBV) infection has been recognized as a major risk factor for the development of hepatocellular carcinoma (HCC).1 Several mechanisms have been proposed to explain HBV-related hepatocarcinogenesis, including insertional mutagenesis of HBV genomes, inflammation, regeneration, and transactivating functions of HBV gene products, such as X protein and truncated middle surface protein.2, 3 Previously, we proposed HBV pre-S mutants as viral oncoproteins, which were accumulated in the endoplasmic reticulum (ER) of ground glass hepatocytes (GGHs).4 pre-S mutants can induce ER stress signals, oxidative DNA damages, and transforming capabilities.5 GGHs are, therefore, recognized as the precursor lesions of HCCs.6
One intriguing observation in chronic HBV infection is the low detection rate of HBV surface antigen (HBsAg), usually below 20% of cases in HCC tissues, whereas HBsAg can be detected in almost 100% of cases in paired nontumorous livers.7 The same finding was observed in HBsAg-expressing transgenic mice, which were accompanied by a decreased or absent expression of HBsAg in HCCs.8 These observations indicate that the decreased HBsAg expression is a consistent phenomenon during the process of HBV tumorigenesis. Although the levels of HBV DNA and HBsAg usually decline along with the natural course of chronic HBV infection,9, 10 there exists such a possibility that host cell factors may become activated to inhibit HBsAg expression or HBV replication during HBV tumorigenesis. This speculation gains support from one recent study reporting that the activation of mammalian target of rapamycin (mTOR)-signaling pathway inhibited the transcription of the HBV large surface antigen (LHBs) gene.11 Because mTOR is frequently activated in HCCs,12 the activated mTOR signal may account for the decreased expression of HBsAg in HCC tissues. Previously, we demonstrated that HBV pre-S mutants could activate the mTOR signal in GGHs.13 Therefore, there appears to an inverse relationship between the expression of HBsAg and the activation of mTOR during HBV tumorigenesis.
The transcription of the LHBs gene is under control of the pre-S1 promoter.14 Several transcription factors may contribute to pre-S1 promoter activity, including TATA box-binding protein, hepatocyte nuclear factor 1 and 3, and Sp1.15-17 This study was, therefore, designed to investigate whether the activation of mTOR by HBV pre-S proteins would inhibit HBsAg expression via the recruitment of transcription factor complex in a feedback manner. The inverse relationship between the expression of HBsAg and phosphorylated mTOR (p-mTOR) was confirmed in 20 paired nontumorous liver and HCC tissues. In vitro, wild-type (WT) or mutant pre-S proteins could up-regulate mTOR. Interestingly, the activated mTOR signal could, in turn, feedback suppress LHBs expression via the transcription factor, Yin Yang 1 (YY1),18 which is physically associated with histone deacetylase 1 (HDAC1) to form a complex on the pre-S1 promoter.
cDNA, complementary DNA; Co-IP, coimmunoprecipitation; DAPA, DNA affinity precipitation assay; EMSA, electrophoretic mobility-shift assay; ER, endoplasmic reticulum; GGHs, ground glass hepatocytes; HBV, hepatitis B virus; HBsAg, hepatitis B virus surface antigen; HCC, hepatocellular carcinoma; HDAC1, histone deacetylase 1; LHBs, HBV large surface antigen; mTOR, mammalian target of rapamycin; Mut, mutated; PCR, polymerase chain reaction; p-mTOR, phosphorylated mTOR; RT-PCR, reverse-transcription PCR; SD, standard deviation; shRNA, short-hairpin RNA; WT, wild type; YY1, Yin Yang 1.
Materials and Methods
Clinical Specimen Collection.
Freshly frozen liver tissues were obtained from the Department of Pathology, National Cheng Kung University Hospital (Tainan, Taiwan), from 1995 to 2007, under the approval of the institutional research committee. Pathology was examined by two experienced pathologists (I.J.S. and H.W.T.).
Plasmids p(3A)SAg-WT, p(3A)SAg-ΔS1, and p(3A)SAg-ΔS2 expressing WT LHBs, pre-S1 mutant, and pre-S2 mutant from the pre-S1 promoter (nucleotide 2438-2845; National Center for Biotechnology Iinformation accession no.: AB014370) were constructed as previously described.19 The short-hairpin (sh)RNA expression plasmids were generated by annealing and ligating shRNA oligonucleotides (Supporting Table 1) into the pSUPER vector (Oligoengine, Seattle, WA). The pre-S1 promoter reporter plasmids were constructed by inserting promoter fragments into the pG5luc vector (Promega, Madison, WI), followed by inserting a Renilla luciferase expression cassette, which was generated from the pRL-TK vector (Promega). pre-S1 promoter regions were amplified by polymerase chain reaction (PCR) with primers shown in Supporting Table 2. Mutated reporter plasmids were further generated using the QuikChange Site-Directed Mutagenesis Kit (Strategene, La Jolla, CA), according to the manufacturer's instructions, with primers shown in Supporting Table 3.
Cell Line and Transient Transfection.
The HuH-7 cell line was used in this study. All transfections were performed with the MicroPorator (Invitrogen Life Technologies, Carlsbad, CA), according to the manufacturer's instructions.
Western Blotting Analysis.
Protein lysates were harvested with lysis buffer radioimmunoprecipitation assay (Upstate Biotechnology, Lake Placid, NY), resolved on sodium dodecyl sulfate/polyacrylamide gels, and transferred to polyvinylidene difluoride membranes. Membranes were incubated with primary antibodies, followed by secondary antibodies, then visualized by the extrachemiluminescence Chemiluminescence Kit (PerkinElmer Life Science, Boston, MA), according to the manufacturer's instructions. The primary antibodies used in this study were anti-LHBs,20 anti-mTOR (Cell Signaling Technology, Danvers, MA), anti-p-mTOR (Abcam, Cambridge, UK), anti-YY1, anti-HDAC1, and anti-HDAC2 (Santa Cruz Biotechnology, Santa Cruz, CA), anti-histone H1 (Upstate Biotechnology), anti-β-Actin (Chemicon, Temecula, CA), and anti-α-Tubulin (NeoMarkers, Fremont, CA).
Total RNAs were extracted using the RNeasy Mini Kit (Qiagen Inc., Valencia, CA), according to the manufacturer's instructions, and converted to complementary DNA (cDNA). PCR was then performed with primers shown in Supporting Table 4.
Real-time PCR was performed using the LightCycler reagents and detection system (Roche Applied Science, Indianapolis, IN), according to the manufacturer's instructions, with primers and TaqMan probes shown in Supporting Table 4. Relative RNA levels were calculated using LightCycler software (Roche Applied Science).
Luciferase Reporter Assay.
Luciferase-expressed cells were assayed by the Dual-Luciferase Reporter Assay System (Promega), according to the manufacturer's instructions. Renilla luciferase activities were measured for normalization. Each experiment was independently repeated at least three times, and data represent the mean with standard deviation (SD) error bar of luciferase activities relative to the control reporter plasmid. Prediction web softwares TESS MASTER (“TESS: Transcription Element Search Software on the WWW”; available at: http://www.cbil.upenn.edu/tess, access date: April 1, 2010) and TFSEARCH (“TFSEARCH: Searching Transcription Factor Binding Sites”; available at: http://www.rwcp.or.jp/papia/, access date: April 1, 2010) were used to search for putative transcription factor binding sites in DNA sequences. Rapamycin (Calbiochem, San Diego, CA) and insulin (Sigma-Aldrich, St. Louis, MO) treatments, when required, were started 24 hours before cell lysis.
Extraction of Cytoplasmic and Nuclear Proteins.
Cells were incubated in cytoplasmic lysis buffer, followed by the addition of 10% nonyl phenoxypolyethoxyethanol. After centrifugation, the supernatant represented the cytoplasmic protein fraction. The pellets were next resuspended in nuclear lysis buffer. The resulting supernatant represented the nuclear protein fraction.
Electrophoretic Mobility-Shift Assay.
Electrophoretic mobility-shift assay (EMSA) was performed using the LightShift Chemiluminescent EMSA Kit (Pierce, Rockford, IL), according to the manufacturer's instructions. The following WT and Mut oligonucleotides were unlabeled or labeled with biotin at 5′-termini, then annealed with their complementary strands to generate DNA probes: WT, 5′-CGTAGCGCATCATTTTGCGGGTCAC-3′; Mut, 5′-CGTAGCGCATAATCTTGCGGGTCAC-3′ (underlined are the putative YY1-binding sites corresponding to nucleotide 2812-2816 of the pre-S1 promoter).
DNA Affinity Precipitation Assay.
The DNA affinity precipitation assay (DAPA) was performed by incubating biotin-labeled DNA probes and salmon sperm DNA with nuclear extracts in binding buffer. Subsequently, the streptavidin-agarose beads (Sigma-Aldrich) were added to the reaction mixtures. The beads were collected, and proteins bound to the beads were subjected to western blotting analysis. The DNA probes were the same as those for EMSA.
Total cell extracts were incubated with YY1 antibody in coimmunoprecipitation (Co-IP) buffer, followed by the addition of the protein G-agarose beads (Millipore, Bedford, MA). The beads were precipitated, and proteins bound to the beads were characterized by western blotting analysis.
Data were analyzed by Student's t-test (P < 0.05, P < 0.01, and P < 0.001) and are shown with SD error bar. For all analysis, only P < 0.05 was considered statistically significant.
Decreased LHBs Expression Was Correlated With Increased p-mTOR Expression in Paired Nontumorous Liver and HCC Tissues.
As compared with nontumorous livers, expression levels of LHBs were reduced in 16 of 20 and those of p-mTOR were enhanced in 15 of 20 paired HCC tissues (Fig. 1). In 13 of 20 cases, an inverse relationship was observed between decreased LHBs and enhanced p-mTOR expressions.
Expression of LHBs Was Stepwise Decreased With Concurrent mTOR Activation Over Time in HuH-7 Cells.
As shown in Fig. 2A, expression of WT LHBs, pre-S1 mutant, or pre-S2 mutant could induce mTOR activation, but its expressions were stepwise decreased at both RNA and protein levels over the time course. Because all three types of LHBs showed similar patterns in expression levels, we selected pre-S2 mutant plasmid as the representative construct for studies in the following experiments. As shown in Fig. 2B, pre-S2 mutant-induced mTOR activation occurred at 48 hours with concurrently decreased LHBs RNA expression, followed by the decrease of LHB protein expression at 72 hours after transfection. The results implied that the negative regulation of LHBs by mTOR signal occurred at the transcriptional level.
Activation of mTOR Suppressed LHBs Expression and Secretion.
To verify whether the observed decrease of LHBs expression would be mediated by mTOR activation, we tested this effect using the mTOR inhibitor rapamycin. As shown in Fig. 3, blockage of mTOR activation by rapamycin significantly restored both RNA and protein expression levels of LHBs in the hepatocytes. Importantly, secreted LHBs in culture supernatant showed the same patterns, implying that serum HBsAg level may be concurrently decreased when mTOR becomes activated during HBV tumorigenesis. The negative regulation of LHBs by mTOR signal was further confirmed by RNA interference studies (data not shown).
Activation of mTOR Repressed pre-S1 Promoter Activity.
We next performed the luciferase reporter assay to clarify whether LHBs suppression by mTOR activation would result from transcriptional repression of the pre-S1 promoter. As shown in Fig. 4A, pre-S1 promoter-driven luciferase activities were decreased in pre-S2 mutant-expressed cells, and the reduced activities could be restored by rapamycin. The same results were observed by RNA interference studies (data not shown). The suppression of pre-S1 promoter by mTOR was further tested by using another mTOR activator: insulin. As shown in Fig. 4B, insulin treatment exhibited a dose-dependent repressive effect on pre-S1 promoter activity that could be abolished by rapamycin. Collectively, these results indicate that mTOR activation could suppress LHBs expression through a negative feedback loop that repressed pre-S1 promoter activity.
Nucleotide 2812-2816 of pre-S1 Promoter Was the Candidate Region Responsible for LHBs Suppression by mTOR Signal.
To determine the critical elements responsible for pre-S1 promoter repression by mTOR signal, we constructed three 5′-deletion mutants of pre-S1 promoter reporter plasmids (Fig. 5Aa,b,c). As shown in Fig. 5Ba,b,c, luciferase activities of all these deletion mutants consistently showed that only approximately 50% of the activities were detected in pre-S2 mutant-expressed cells, compared with those in control cells, indicating that nucleotide 2789-2845 of the pre-S1 promoter was the minimal region for mediating mTOR signal-induced transcriptional repression. A computer search further revealed that this region contains three putative transcription factor binding sites at nucleotide 2794-2801 (site 1), 2812-2816 (site 2), and 2820-2825 (site 3) of the pre-S1 promoter. We, therefore, generated pre-S1 promoter constructs with mutations at each one of these three sites (Fig. 5Ad,e,f) for further assays. As shown in Fig. 5Bd,f, activities of pre-S1 promoter-carrying mutations at site 1 or site 3 still showed mTOR activation-dependent repression, indicating that these two sites were not responsive elements for the mTOR signal. Interestingly and unexpectedly, the mutation at site 2 resulted in a relatively low or insignificant luciferase activity of the pre-S1 promoter (Fig. 5Be), suggesting that site 2 was transcriptionally necessary for pre-S1 promoter activity. To further evaluate the role of site 2 in pre-S1 promoter repression by mTOR, we created one additional mutant construct, in which site 2 remains intact, but site 1 and site 3 are mutated (Fig. 5Ag). We observed that this mutant construct could negatively respond to mTOR activation (Fig. 5Bg). Therefore, the 2812-2816 site appeared to be not only transcriptionally important for pre-S1 promoter activity, but also capable of mediating a suppressive effect under the status of mTOR activation.
YY1 Bound to the 2812-2816 Site of pre-S1 Promoter In Vitro.
Because sequence analysis revealed that the 2812-2816 site of pre-S1 promoter is the putative binding site for transcription factor YY1, we performed EMSA and DAPA to confirm the binding of YY1 to this site. In the EMSA experiment, two major shifted bands (a and b) were detected using WT probes containing the 2812-2816 site, one (a band) of which was not detected using the Mut probes, which destroyed the 2812-2816 site (Fig. 6A). Coincubation with competitors abrogated the formation of these two bands. The data suggest that the slower migrating band (a), but not the faster migrating band (b), represented the YY1-probe complex. To further confirm this speculation, we additionally added YY1 antibody to the reaction mixtures and observed specifically reduced intensity of the slower migrating band (a). This data further verified the interaction of YY1 with the 2812-2816 site. The DAPA assay further revealed a result consistent with that of EMSA, indicating that WT probes precipitated YY1, whereas the Mut probes did not (Fig. 6B).
Activation of mTOR Enhanced YY1 Expression and Nuclear Localization.
We next attempted to explore whether the mTOR signal could regulate YY1. As shown in Fig. 7A, YY1 protein expression was increased in pre-S2 mutant-expressed cells, and the up-regulation of YY1 was apparently mediated by mTOR activation, because it could be abolished in the presence of rapamycin. Furthermore, subcellular fractionation analysis showed increased levels of nuclear YY1 accumulation in pre-S2 mutant-expressed cells that could be diminished by rapamycin (Fig. 7B). The results were further confirmed by RNA interference studies (data not shown).
Recruitment of HDAC1 by YY1 Mediated pre-S1 Promoter Repression by mTOR Signal.
Accumulating evidence indicates that YY1 can execute transcriptional repression by complexing with corepressors, among which HDAC1 and HDAC2 are the most relevant.21, 22 Therefore, we hypothesized that mTOR signal-induced pre-S1 promoter repression might be the result of the recruitment of HDACs by YY1. As shown in Fig. 8A, selective knockdown of HDAC1, but not HDAC2, protected pre-S1 promoter activity from repression by pre-S2 mutant-induced mTOR activation, suggesting that it was HDAC1 that might be physically associated with YY1 and contribute to its suppressive activity. We next carried out Co-IP experiments to confirm the possible association between YY1 and HDAC1. As shown in Fig. 8B, YY1 antibody could coimmunoprecipitate higher levels of HDAC1 from pre-S2 mutant-expressed cells than control cells. Furthermore, this increased association of YY1 with HDAC1 was dependent on mTOR activation, because it could be abolished by rapamycin. Unlike HDAC1, HDAC2 showed no interactions with YY1. Experiments using the HDACs inhibitor, suberoylanilide hydroxamic acid, revealed the same findings (data not shown).
This study, for the first time, demonstrated one interesting negative feedback regulation of surface antigen synthesis by the activation of the mTOR signal during the progression of HBV tumorigenesis. The decreased levels of HBsAg and HBV DNA in serum or hepatocytes, therefore, may not necessarily represent a good sign of disease improvement during the natural course of HCC development, but instead, it may indicate a disease progression toward tumorigenesis, especially at the advanced stage of diseases. This finding, together with the detection of pre-S mutations in serum,23-25 should provide an additional hallmark to predict disease progression in the follow-up of patients with chronic HBV infection.
Activation of the mTOR signal plays essential roles in cell growth control by regulating many cellular processes26 and is a major molecular event in HBV tumorigenesis.27 Previously, we demonstrated that HBV pre-S mutants could enhance the expression of vascular endothelial growth factor-A and activation of Akt/mTOR signaling in GGHs.13 In this report, we have proposed that the activated mTOR signal in GGHs could further lead to decreased HBsAg expression in dysplastic GGHs and HCCs. Because mTOR is frequently activated in the absence of HBsAg expression in HCC tissues, as shown in this study, the activation of mTOR in HCCs may be sustained or activated by other molecular events, such as the inactivation of tuberous sclerosis complex.28, 29 Furthermore, the activation of mTOR during HBV tumorigenesis may not be the sole factor responsible for the decrease or complete absence of HBsAg in HCC tissues. Several transcription factors may contribute to pre-S1 promoter activity in a positive or negative manner.15-17 Whether other transcriptional repressors of the pre-S1 promoter exist or there is an unidentified mechanism involved in the regulation of HBsAg in HCC tissues remain to be clarified in the future.
In this study, we further verified nucleotide 2812-2816 of the pre-S1 promoter as the specific binding site for mTOR signal-regulated transcription factor YY1. YY1 is a multifunctional transcription factor that can either activate or repress transcription, depending upon the promoter context in which it binds or specific protein interactions.30 Our results revealed that mTOR activation could enhance YY1 expression and increase its nuclear localization to bind to the pre-S1 promoter. Because mTOR cannot enter the nucleus in HuH-7 cells, we suggest that mTOR may regulate YY1 indirectly through a hitherto unidentified signaling pathway. Furthermore, we found that HDAC1 was physically associated with YY1, depending upon mTOR activation, and contributed to the suppressive effect of YY1 on the pre-S1 promoter. One interesting finding in this study was the greatly reduced luciferase activity in the preS1 promoter construct with mutation at the 2812-2816 site, suggesting that this site was also transcriptionally important besides the mTOR activation-induced suppressive function. Several studies have reported similar findings on the link between YY1 expression levels and its repressive effect on promoters.31, 32
The suppression of HBsAg by mTOR signal is implicated in the regulation of HBV replication. One recent study reported that the activation of the mTOR-signaling pathway could inhibit HBV RNA transcription and DNA replication, and the suppression may, possibly, be mediated by transcriptional regulators that recognize precore/core and pre-S1 promoters.11 Therefore, it will be interesting to clarify whether the inhibition of HBV replication by mTOR activation is through down-regulating pre-S1 promoter activity. Finally, several mTOR inhibitors have been developed at various phases of clinical trials.33 According to our findings in this study, to target mTOR signaling for HBV-related HCC may potentially lead to HBV reactivation. There are increasing reports on the reactivation of HBV replication and hepatitis flare-up in HBV-related HCC patients receiving anticancer treatments.34, 35 Our results provide an explanation for the untoward effect of mTOR inhibitors on HCC patients with chronic HBV infection and emphasize the necessity of combining antiviral agents when mTOR inhibitors are used for HBV-related HCC therapy.
In conclusion, our data reported here propose a novel molecular mechanism to explain the stepwise decrease of HBsAg expression during HBV tumorigenesis. The negative regulation of HBsAg by mTOR signal raises a serious issue regarding the clinical significance of decreased levels of HBV DNA and surface antigens in patients with chronic HBV infection, especially at the advanced stage of diseases. Our current attempt on targeted therapy using mTOR inhibitors may carry a potential risk to activate HBV replication and result in untoward clinical consequences.