Chemical-induced degradation of PreS2 mutant surface antigen reverses HBV-mediated hepatocarcinogenesis

Naturally evolved immune-escape PreS2 mutant is an oncogenic caveat of liver cirrhosis and hepatocellular carcinoma (HCC) during chronic hepatitis B virus (HBV) infection. Notably, PreS2 mutants is prevalent in above 50% of patients with HCC. Intrahepatic expression of PreS2 mutant large surface antigen (PreS2-LHBS) induces endoplasmic reticulum stress, mitochondria dysfunction, cytokinesis failure and subsequent chromosome hyperploidy. In this study, we ask if long-term inhibition of PreS2-LHBS may act to reverse HBV-mediated hepatocarcinogenesis. We set up a stability reporter platform and identified ABT199 as an inhibitor of PreS2-LHBS from a library of 1068 FDA-approved drugs. Treatment of ABT199 induced PreS2-LHBS degradation without affecting the general cell viability, as shown in hepatoma and immortalized hepatocyte cell lines. We found that ABT199 induced the recruitment of PreS2-LHBS to ring-shaped structures in close proximity to lysosomal marker Lamp1 and multivesicular body marker Rab7. Simultaneously, inhibitions of lysosomal degradation or microautophagy restored the expression of PreS2-LHBS. Specifically, a 24-hr treatment of ABT199 reduced DNA damages and cytokinesis failure induced by PreS2-LHBS. Persistent treatment of ABT199 for 3 weeks reversed chromosome hyperploidy in PreS2-LHBS cells and suppressed anchorage-independent growth of HBV-positive hepatoma cell line. Together, we showed that ABT199 provoked selective degradation of PreS2-LHBS through the induction of microautophagy, and a long-term treatment of ABT199 reversed oncogenic mechanisms induced by HBV. Our results indicate that long-term degradation of PreS2-LHBS may serve as a novel therapeutic strategy to constrain HBV-mediated hepatocarcinogenesis. Graphical abstract Author summary HBV PreS2 mutant large surface antigen (PreS2-LHBS) is an oncoprotein that induces liver cirrhosis and liver cancer. This study identified ABT199 as a potential inhibitor of PreS2-LHBS. We found that ABT199 could trigger the degradation of PreS2-LHBS through the induction of microautophagy. Moreover, a long-term treatment of ABT199 significantly reversed LHBS-induced oncogenic events including DNA damage, mitotic failure, chromosome hyperploidy, and anchorage-independent growth. This study not only identifies a specific inhibitor of PreS2-LHBS but also highlights a plausible strategy to constrain HBV-mediated hepatocarcinogenesis through targeting the degradation of PreS2-LHBS.


Graphical abstract 1
Introduction 1 Chronic hepatitis B virus (HBV) infection is the most prevalent cause of liver diseases-3.5% 2 of people were chronically infected worldwide; 20-30% of patients progress to liver cancer 3 (hepatocellular carcinoma, HCC) and liver cirrhosis; 25% of patients eventually died in the 4 diseases attributed to chronic HBV infection [1,2]. Although a prophylactic vaccine and antiviral 5 drug exist, no cure at this stage [1,2]. 6 HBV S gene encodes three segments (PreS1, PreS2, S) that translate three proteins differing in 7 length but sharing the same C terminal: large (LHBS), middle (MHBS), and small surface antigen 8 (SHBS). All three surface antigens together constitute HBV surface antigens (HBsAg) in serology. 9 Surface antigen plays a role in mediating disease progression in CHB patients. A high level of 10 HBsAg has been found to suppress CD8 positive T cell function, leading to T cell exhaustion and 11 incomplete virus clearance [3,4]. In addition, viral surface antigen has been identified as an 12 oncoprotein associated with HCC and cirrhosis progression. Transgenic mice overexpressing 13 surface antigen have been found to develop HCC [5]. In several studies, patients with a high 14 HBsAg level ( >1000 IU/mL) showed increased risk of HCC even in patients with low HBV DNA 15 level ( <2000 IU/mL) [6][7][8]. We recently also showed that LHBS might overwrite hepatocyte 16 G2/M checkpoint, increase DNA damage, induce cytokinesis failure, and consequently provoke 17 hepatocyte hyperploidy [9,10]. 18 During the chronic hepatitis B (CHB) progression, surface antigen accumulates in ER, leading 19 to flat and hazy cell morphology, called ground-glass hepatocyte (GGH). According to 20 immunohistochemistry staining of surface antigen, GGH can be classified as type I and type II 21 GGH. Type I GGH displays a homogenous cytoplasmic distribution of surface antigen; type II 22 GGH displays a marginal distribution of surface antigens. In CHB patients, the prevalence of type 23 II GGH increases upon disease progression. This advanced stage of GGH usually bears a mutant 24 LHBS with PreS2 region deletion (PreS2-LHBS, predominate deletion in 2-55 or 4-57) [11]. 25 Studies have shown that type II GGH carrying PreS2-LHBS has a significant association with 26 HCC, cirrhosis, and the decreased survival rate [12,13]. In addition, compared to wild-type LHBS, 27 PreS2-LHBS shows higher ER retention that induces a substantial ER stress, leading to calcium 28 accumulation, mitochondrial dysfunction, and genomic instability [14,15]. Notably, the PreS2 29 deletion region corresponded to the epitope of CD8 T cell, suggesting the immune-escape attribute 30 of this mutant [11]. Furthermore, transgenic mice expressing PreS2-LHBS are sufficient to 31 develop liver dysplasia and HCC [16]. Notice that the increasing prevalence of integrated HBV 1 genome, a driving force of HCC, particularly supports the consistent expression of intracellular 2 surface antigen [17]. These studies indicate that intrahepatic PreS2-LHBS is not only a histological 3 hallmark but a caveat of disease progression to liver cirrhosis and HCC. Thus, targeting PreS2-4 LHBS is a therapeutic option to constrain HBV-induced hepatocarcinogenesis. 5 Therapeutic options to eliminate HBV are not available at this stage. The ideal treatment 6 endpoint, called "functional cure", is determined by the complete loss of HBsAg in the circulation. 7 According to the AASLD hepatitis B guideline in 2018, approximately only 1~ 8% of HBeAg-8 positive patients achieved HBsAg loss with current treatment interferon and nucleotide analog 9 (NA). The chance of HBsAg loss was further down to below 1% in HBeAg-negative patients [18]. 10 Note that even though NA significantly reduced the viral marker in serum, the expression of 11 intrahepatic virus HBsAg was not be affected, especially the type II GGH carrying viral preS2 12 mutant, which was linked to decreased overall survival of HCC [19]. Coordinately, PreS2-LHBS 13 was detected in above 60% of HCC patients [11]. Accordingly, intrahepatic PreS2-LHBS is a risk 14 factor for hepatocarcinogenesis, and yet no conventional therapeutic can eliminate such oncogenic 15 caveat in the liver at this stage. 16 As a result, there is an unmet medical need for the reduction of intrahepatic PreS2-LHBS in 17 CHB patients. We ask if long-term inhibition of PreS2-LHBS may reverse the existing oncogenic 18 effect. Therefore, we set up a protein stability platform to screen for chemical inhibitors that induce 19 the degradation of intracellular PreS2-LHBS. From an FDA-approved drug library, we identified 20 ABT199 as a candidate inhibitor of PreS2-LHBS. We showed that long-term degradation of 21 PreS2-LHBS was achieved by treating with ABT199, which subsequently reversed oncogenic 22 effects induced by PreS2-LHBS. These results indicate that the long-term degradation of 23 intrahepatic PreS2-LHBS may serve as a novel therapeutic approach to constrain the development 24 of HBV-mediated HCC. 25

Result 1
Repurposing FDA-approved drug to target PreS2-LHBS through protein degradation 2 To target PreS2-LHBS, we decided to explore the existing chemicals that reduce PreS2-LHBS 3 through protein degradation. Compared to other chemicals, FDA-approved drugs have been well 4 characterized in toxicity and efficacy, making them more appreciated for drug repurposing; thus, 5 we determined to screen for the chemicals from this library. To perform the drug screening, we 6 first established a cell-based screening platform to measure the abundance of PreS2-LHBS. 7 Specifically, we adopted a protein stability reporter system in 293T cells [20]. This system 8 evaluated the PreS2-LHBS abundance by two fluorescent signals: EGFP, tagged with PreS2-9 LHBS; DsRed, an internal control, synthesized from the same mRNA of EGFP ( Figure 1A). The 10 established reporter stable cell line expressing PreS2-LHBS in 93.4% of the cell population was 11 used in this study ( Figure S1). Next, we individually challenged cells with one chemical from the 12 FDA-approved drug library (1068 drugs) and assessed fluorescent signals and cell viability. 13 According to Chebyshev's inequality, around 90% of data will locate within 3-fold of standard 14 deviation (SD) around the mean without any model assumption. Therefore, we considered the drug 15 significantly reduced PreS2-LHBS when the fluorescent signal was beyond 3 SD from the mean 16 ( Figure 1A). We selected drugs that significantly reduced fluorescent signal without affecting cell 17 viability as our initial candidates. From a collection of 1068 drugs, 11 drugs hit this criterion 18 ( Figure 1B). These drugs reduced around 40% to 60% of fluorescence signal but maintained over 19 80% cell viability ( Figure 1C). In addition, hierarchical clustering analysis revealed that screen 20 hits were in the same cluster, indicating that we selected a group of drugs with a similar mechanism 21 of action on PreS2-LHBS, strengthening our selection criterion. ( Figure 1D). We treated SNAP-22 tagged PreS2-LHBS stable hepatocyte cell line (NeHep-S2) with selected drugs and evaluated the 23 expression level of PreS2-LHBS by immunoblotting. We confirmed that ABT199, Rapamycin, 24 Bupivacaine, and Deferoxamine reduced intracellular PreS2-LHBS ( Figure 2A). 25

ABT199 reduces intracellular expression of PreS2-LHBS 27
To assess the efficacy of repurposed drugs on reducing PreS2-LHBS, we evaluated the 50% 28 effective concentration (EC50) and the maximal reduction percentage (Rmax). To this end, NeHep-29 S2 cells were treated with each drug in serial dilution. We found that ABT199, Bupivacaine, and 30 Rapamycin, but not Deferoxamine, reduced intracellular PreS2-LHBS in a dose-dependent manner. 31 Of three effective drugs, ABT199 showed the lowest EC50; ABT199 and Bupivacaine showed Rmax 1 higher than 80% ( Figure 2B). Notably, we observed that the reduction of PreS2-LHBS (>30%) 2 occurred upon treatment with all three drugs in nanomolar concentration. Furthermore, these drugs 3 showed no cytotoxicity (~100% cell viability) in their effective concentration ( Figure 2C), 4 indicating that the efficacy of reducing PreS2-LHBS was not due to cell death. Because ABT199 5 was the most effective drug and can be taken orally, we focused on ABT199 in the following study. 6 To better understand how ABT199 reduced PreS2-LHBS, we examined the kinetics of PreS2-7 LHBS reduction upon treatment of ABT199 over time. NeHep-S2 cells were treated with ABT199 8 and were followed the expression by immunoblotting. We found that ABT199 significantly 9 reduced PreS2-LHBS within 6hr (>50%), whereas the internal control, BCL2, was not affected 10 ( Figure 3A). Next, we tested if ABT199 affects the PreS2-LHBS protein half-life, a hallmark for 11 estimating intracellular protein stability. To measure the protein half-life, we blocked the protein 12 synthesis by cycloheximide in the presence or absence of ABT199. Strikingly, fitting the 13 expression with non-linear regression showed that ABT199 reduced the PreS2-LHBS protein half-14 life by 54%, shirking from 5.7hr to 2.6hr ( Figure 3B), suggesting PreS2-LHBS tends to be 15 degraded in the presence of ABT199. Finally, we verified the efficacy of ABT199 in HuH7 and 16 HepG2 infected with adenovirus carrying a complete wild-type HBV genome (AdHBV) [21]. As 17 expected, ABT199 reduced the intracellular LHBS in HuH7 and HepG2 by around 40% (Figure  18 3C). Accordingly, our data indicate that ABT199 effectively reduces either wild-type or PreS2-19 LHBS in the context of the complete HBV genome. 20

ABT199 induces PreS2-LHBS degradation through microautophagy 22
From the previous section, we hypothesized that ABT199 reduced PreS2-LHBS by activating 23 cellular pathways; thus, we aimed to identify the cellular pathways that control the protein level of 24 PreS2-LHBS. To study whether PreS2-LHBS transcription is affected by ABT199, NeHep-S2 25 cells were treated with ABT199 or solvent control (DMSO), and the mRNA level was analyzed 26 by quantitative real-time PCR. We found no significant difference in PreS2-LHBS mRNA level 27 ( Figure 4A). Thus, ABT199 does not affect the transcription of PreS2-LHBS. Next, we tested 28 whether ABT199 directly interacts with LHBS to trigger the degradation. To this end, we 29 incubated purified LHBS protein with ABT199 or solvent control (DMSO) and measured the 30 interaction by microscale thermophoresis (MST). Compared to solvent control, the interaction 31 between LHBS and ABT199 showed no correlation with ABT199 concentration ( Figure S2), 1 suggesting ABT199 does not directly interact with LHBS. Thus, we sought to explore the 2 degradation pathway activated by ABT199. 3 ABT199 is a BH3 domain analog that selectively attaches to BCL2 and inhibits its activity. 4 Previous studies have shown that dissociation of Beclin-1 and BCL2 can initiate macroautophagy, 5 a well-studied protein degradation pathway [22]. Therefore, we tested whether ABT199 may 6 activate macroautophagy. We examined the macroautophagy signal, the localization between 7 PreS2-LHBS and autolysosome, and the phenomena after depleting BCL2 and/or Beclin1 in 8 ABT199 treated cells ( Figure 4A, S3, S4); however, we found no evidence supporting the 9 activation of macroautophagy. Thus, we turned to investigate all possible degradation pathways. 10 The protein degradation pathway can be classified into: proteasomal degradation and lysosomal 11 degradation, where lysosomal degradation can be further classified into: macroautophagy, 12 microautophagy, and chaperone-mediated autophagy. To identify key degradation pathways, we 13 treated cells with each pathway inhibitor in the presence or absence of ABT199 and examined the 14 expression of PreS2-LHBS. We found that inhibition of either lysosomal degradation or 15 microautophagy abolished ABT199-induced degradation, indicating that microautophagy is 16 involved in regulating the intracellular stability of PreS2-LHBS ( Figure 4C). Notably, ABT199-17 induced degradation was also abolished upon depletion of NPC1, which inhibited 18 microautophagy's dynamic ( Figure 4E). Immunofluorescent staining showed that PreS2-LHBS 19 was recruited to a ring-shaped structure, and the distribution was in close proximity of the 20 lysosomal marker Lamp1 and MVB marker Rab7 upon treatment of ABT199 ( Figure 4D). In 21 addition, we performed a purification screening following mass spectrometry analysis and 22 identified HSC70, an essential chaperone mediating microautophagy [23], significantly interacted 23 with PreS2-LHBS ( Figure S5). Taken together, our data suggest that ABT199 activated 24 microautophagy to degrade PreS2-LHBS. 25 26

Long-term inhibition of PreS2-LHBS by ABT199 reverses oncogenic effect in PreS2-LHBS 27 expressing hepatocytes 28
To understand the long-term impact of reducing PreS2-LHBS, we tested if LHBS-induced 29 oncogenic events such as increased DNA damage, mitotic defect, chromosome hyperploidy, and 30 oncogenesis [9] can be reversed by ABT199. To evaluate the DNA damage, we stained ABT199-31 treated and untreated NeHep-S2 cells for phosphorylated H2AX (pH2AX), a DNA damage marker, 1 and calculated the percentage of pH2AX positive cells in the population. Strikingly, compared to 2 control, we found that ABT199 significantly reduced DNA damage in PreS2-LHBS cells by 2.1-3 fold ( Figure 5A). In addition, long-term treatment (3weeks) of ABT199 not only significantly 4 reduced DNA damage but showed a lower variation of positive percentage, suggesting that 5 persistent suppression of PreS2-LHBS can restore DNA damage much effectively. Next, to 6 monitor mitotic failure, we followed the mitotic progression of NeHep-S2 cells by time-lapse 7 imagining and classified mitotic events according to their division results ( Figure 5B, left panel). 8 Compared to control cells, we found that ABT199 significantly reduced 45% of mitotic failure in 9 NeHep-S2 cells ( Figure 5B, right panel). Furthermore, analysis of mitotic failure revealed that the 10 reduction came from the decrease of cytokinesis failure ( Figure 5B, right panel), which is identical 11 to the observation while depleting LHBS in our previous study [9]. 12 We next examined if ABT199 can reduce LHBS-induced hepatocyte hyperploidy. We treated 13 cells with ABT199 for 3weeks and calculated the chromosome numbers by chromosome spread. transcription), or co-treated with ABT199 and ETV, respectively. Interestingly, only treatment of 26 ABT199 significantly inhibited colony formation ( Figure 6B). Compared to ABT199, ETV did 27 not significantly inhibit colony formation; however, the colony number and size were further 28 reduced by co-treatment with ABT199, suggesting the oncogenesis can be constrained by targeting 29 LHBS. In summary, these data show that persistently reducing PreS2-LHBS by ABT199 could 30 consequently revere the oncogenic effects caused by PreS2-LHBS. 31

Discussion 1
In this study, we repurposed several new drug candidates to target intracellular PreS2-LHBS. 2 According to HBV drug watch, most of the developing anti-HBV drugs take action on HBV viral 3 polymerase and capsid, whereas only two chemicals target extracellular HBsAg in clinical 4 development [26]. To our best knowledge, inhibitors that specifically target intracellular surface 5 antigens are not available. As such, these candidates represent the first chemical inhibitors of 6 intracellular PreS2-LHBS. 7 Targeting viral surface antigen faces several technical difficulties. First, surface antigen is an 8 envelope structural protein with no enzyme activity. Therefore, it is not possible to develop an in 9 vitro functional enzyme assay for the selection of chemical inhibitors. Second, surface antigen can 10 be translated from either HBV cccDNA or integrated genome, and the integration of HBV DNA 11 occurs within three days post-infection [27], making it challenging to eliminate surface antigen at 12 the DNA level. In addition, RNA interference is likely inefficient to eliminate surface antigen in 13 the context of a whole liver. A recent study showed that siRNA, designed using a complete HBV 14 sequence, could not efficiently reduce surface antigens in HBeAg-negative patients due to the 15 increased integrated DNA and viral sequence mutation following CHB progression [17]. As an 16 alternative approach, we tested whether any inhibitor can effectively target the PreS2-LHBS at the 17 protein level. Activating the cellular degradation pathway to promote the clearance of toxic protein 18 has been described in neurodegenerative diseases, such as Alzheimer's disease [28]. These 19 neurodegenerative diseases share a common feature: accumulation of non-function protein, similar 20 to CHB that accumulates PreS2-LHBS in hepatocytes. With this central ideal, we adopted the 21 protein stability platform and identified Rapamycin, Bupivacaine, and ABT199 that induced 22 selected degradation of intracellular PreS2-LHBS. Interestingly, all these screen hits are linked to 23 different processes of autophagy. Rapamycin is a well-studied mTOR inhibitor that triggers 24 autophagy and ubiquitin-proteasome system [29]; Bupivacaine has been reported to induce protein 25 degradation in muscle cells [30,31] and also induce autophagy in lung cancer cells [32]; ABT199 26 is a BCL2 inhibitor that has the potency to regulate apoptosis and autophagy pathway [33]. These 27 data imply that manipulating protein-specific degradation is a plausible approach to inhibit PreS2-28

LHBS. 29
In this study, we found that microautophagy is activated by ABT199. Endosomal 30 microautophagy occurs during the formation of MVB to deliver the protein into the late 31 endosome/MVB vesicle and degrade the content either directly in the late endosome or by the 1 fusion with the lysosome. This process requires the ESCRT component and HSC70, an essential 2 chaperone, to sequester protein into the endosome membrane [23]. Notably, we also identified that 3 HSC70 interacted with PreS2-LHBS from purification screening upon ABT199 treatment (Figure  4 S6). How exactly ABT199 activates microautophagy is not clear. By searching the 5 SwissTargetPrediction [34], mTOR is listed as a potential interaction partner of ABT199 with a 6 high prediction probability. mTOR has been reported to regulate microautophagy [35,36], 7 suggesting a possible role of microautophagy in the stability control of PreS2-LHBS. We suggest 8 the interaction with HSC70, and the implication of endosomal transport may contribute 9 synergistically to the degradation of PreS2-LHBS in the presence of ABT199. 10 Whether the side effect of ABT199 is acceptable for CHB treatment? ABT199 is a BH3 domain 11 analog, mainly used to treat chronic lymphocytic leukemia by inhibiting BCL2 function [37]. 12 Tumor lysis syndrome, caused by a large number of tumor cell death, is the major precautions side 13 effect in treated patients [38]. In this study, the EC50 of ABT199 in inhibiting PreS2-LHBS was 14 much lower than the IC50 of cell viability, suggesting that ABT199 may inhibit PreS2-LHBS 15 without affecting general cell viability. Previous studies have shown that the inhibition of BCL2 16 can disrupt HBV-induced calcium imbalance, drug resistance, and viral replication [39,40]. We 17 also did not detect cytotoxicity while depleting BCL2 in PreS2-LHBS expressing cells ( Figure S3). 18 These data imply that (1) depletion of BCL2 is not cytotoxic to HBV-positive hepatocytes and (2) 19 inhibition of BCL2 may block HBV through multiple intracellular pathways, including the 20 induction of microautophagy as shown in this study, and the reversion of calcium imbalance, drug 21 resistance, and viral replication. 22 Finally, we showed that long-term inhibition of PreS2-LHBS by ABT199 can significantly 23 decrease DNA damage, cytokinesis failure, hyperploidy, and oncogenesis. Surface antigen has 24 been proposed to induce ER stress, chromosome instability, and HCC [15,41,42]. Our results 25 highlight the possibility of reversing these oncogenic effects by reducing surface antigen. 26 Interestingly, hyperploidy population seems to be an inflexible consequence but also can be 27 reduced by long-term inhibiting PreS2-LHBS. A study has shown that hepatocytes can undergo 28 ploidy reversal to divide into different genome copies containing hepatocytes [43]. Our result 29 suggests that PreS2-LHBS is required for maintaining the hyperploidy population. Due to the time 30 required for ploidy reversal [43,44], we notice that long-term inhibition of Pres2-LHBS is the key 31 to restricting oncogenesis. Current methods such as CRISPR/Cas9 gene editing may be applied to 1 long-term inhibition of PreS2-LHBS. Studies have shown that both integrated and cccDNA can 2 be targeted by CRISPR/Cas9 system, resulting in the inhibition of viral product [45-47]. However, 3 CRISPR/Cas9 system requires the pre-modified cells expressing Cas9 protein, which might be 4 challenging for clinical application. In addition, eliminating all the HBV genomes in the infected 5 hepatocytes is also tricky. Although the long-term efficacy of ABT199 needs further investigation 6 in different advanced models, and there might be other methods to achieve long-term inhibition of 7 PreS2-LHBS, ABT199 is an on-broad method for targeting PreS2-LHBS. Importantly, our result 8 proves the concept that long-term reduction of PreS2-LHBS is a plausible therapeutic option to 9 prevent HBV-induced oncogenesis. Whether such an approach may advance the complete 10 elimination of HBV in patients is still awaiting further examinations. 11

Molecular cloning and cell culture 2
To produce PreS2-LHBS reporter construct, PreS2-LHBS sequence (deletion nucleotides 2-55) 3 was cloned into the protein stability construct after the c-terminal of EGFP by In-Fusion cloning 4 (Takara). The protein stability construct was a kind gift from Dr. Hsueh-Chi Sherry Yen, Academia 5 Sinica, Taiwan [20]. To generate protein stability reporter cell line, 293T cells were transfected 6 with the protein stability construct carrying PreS2-LHBS. Transfected cells were selected by 2 7 µg/ml puromycin for two weeks and further selected by cell sorting according to their protein 8 stability signal. Selected cells were limit diluted into 96 well plates to obtain the single clone.

FDA-approved drug library screening and cell viability assay 25
The FDA-approved drug library was purchased from TargetMol (No. L1010). PreS2-LHBS 26 reporter stable cells were seeded in 96 well plated and treated with 10uM of one drug for 24hr. 27 The culture medium was removed, and the cells were washed with PBS two times. The EGFP and 28  Images were acquired by epifluorescence microscopy (Leica DMI6000). 12

Live cell microscopy 13
Cells were grown in chamber slides (Ibidi) at 80% confluent. Time-lapse images were acquired 14 at 15 min intervals for 24 h by epifluorescence microscopy (Leica DMI6000) equipped with Andor 15 Luca R EMCCD camera (Andor Technology) and a motorized stage supplied with 5% CO2 at 37 16 ˚C. Mitotic cells were followed once the cells rounded up and were classified by their destination. The ploidy conveyor of mature hepatocytes as a source of genetic variation. Nature. function fitted to the concentration against the expression. The expression was normalized to 10 solvent control (DMSO) and loading control (GAPDH). 11 (C) The cytotoxicity of the repurposed drug was determined by cell viability. Six thousand NeHep-1 S2 cells were seeded in 96 plates and subjected to repurposed drug with indicated concentration 2 for 24hr to measure the cell viability. The relative IC50 was calculated by a three-parameter 3 logistic function fitted to the concentration against the viability. The viability was normalized to 4 solvent control (DMSO) treated cells. The vertical line shows the EC50 of each drug, and the 5 heretical line shows 50% cell viability. 6 Error bar show SEM. Error bar showed SEM. **P < 0.005, *P < 0.05, determined by two-tailed, unpaired Student's t-4 test. Error bar shows SEM. **P < 0.005, *P < 0.05, n.s, non-significant determined by two-tailed, 17 unpaired Student's t-test. Error bar showed SEM. **P < 0.005, *P < 0.05, n.s., non-significant determined by two-tailed, 3 unpaired Student's t-test. size in each treatment.

12
Error bar showed SEM. **P < 0.005, *P < 0.05, n.s., non-significant determined by two-tailed, 13 unpaired Student's t-test. 14 15 Supporting information 1 2 Figure S1. Expression profile of PreS2-LHBS reporter stable cell line 3 Ten thousand of 293T cells stably expressing PreS2-LHBS reporter were analyzed by flow 4 cytometry. The population percentage located in each block was shown in the corner. 5 1 Figure S2. Investigation of the interaction between HBsAg and ABT199 2 500nM of HBsAg protein were labeled with HisDye and were incubated with the indicated 3 concentration of ABT199 and solvent control (DMSO). The graph shows the interaction signal 4 determined by MST. 5 1 Figure S3. The effect of depleting BCL2 and BECN1 2 NeHep-S2 cells were knockdown by transfection with a BECN, BCL2, or control specific siRNA. 3 The protein expression was then analyzed by immunoblotting. GAPDH served as a loading control. 4 1 Figure S4. localization of macroautophagy and PreS2-LHBS upon ABT199 treatment 2 NeHep-S2 cells were stained for PreS2-LHBS and autolysosome after treating with ABT199 for showed a significantly increased number of different proteins. 5 (C) Functional enrichment analysis of the different proteins in ABT199 treatment. 6 Table S1. FDA-approved drug screening result, related to Figure 1 1 2 Table S2. Identified protein in purification screening upon ABT199 treatment, related to 3 Figure S5  4