J Biosci (2021)46:53 Ó Indian Academy of Sciences

DOI: 10.1007/s12038-021-00176-0 (0123456789().,-volV)(0123456789().,-volV)

Hypomethylation of LIMD1 and P16 by downregulation of DNMT1 results in restriction of liver carcinogenesis by amarogentin treatment

1 2 1 3 DEBOLINA PAL ,SUBHAYAN SUR ,RITUPARNA ROY ,SUVRA MANDAL and 1 CHINMAY KUMAR PANDA * 1Department Oncogene Regulation, Chittaranjan National Cancer Institute, Kolkata 700 026, India 2Department of Pathology, Saint Louis University, Saint Louis, USA 3Department of Chemistry, National Research Institute for Ayurvedic Drug Development, Kolkata, India

*Corresponding author (Email, [email protected]) MS received 28 December 2020; accepted 17 May 2021

Amarogentin (active component of Chirata) was found to prevent CCl4/NDEA-induced liver carcinogenesis at mild dysplastic stage through modulation of cell cycle, apoptosis, self-renewal pathways. The cell cycle regulatory LIMD1, P16 and RBSP3 were found to be upregulated in restricted liver lesions. To understand the mechanism of upregulation during restriction of cacinogenesis, the effect of amarogentin on epigenetic modification was evaluated in this study. It was also validated in vitro. Hypermethylation of LIMD1 and P16 was seen in mouse hepatocellular carcinoma (30th week carcinogen control mice); however, hypomethylation of these genes was seen in amarogentin-treated liver. In the case of RBSP3, no such change was seen. DNMT1 expression (mRNA/) was significantly increased in later stages of carcinogenesis, whereas its expression was comparable to normal liver in the case of amarogentin treatment. No significant change in expression (mRNA/protein) of HDAC1/2 was observed irrespective of treatment. Amarogentin treatment upregulated the expression (mRNA/protein) of LIMD1, P16 and RBSP3 in the HepG2 cell line. Here also treated cells showed LIMD1 and P16 hypomethylation with DNMT1 downregulation. Increased expres- sion of LIMD1, P16 and RBSP3 after treating cells with demethylating agent 5-aza-2-deoxycytidine indicated epigenetic modulation by amarogentin treatment.

Keywords. Amarogentin; liver carcinogenesis; epigenetics; LIMD1; P16; DNMT1

1. Introduction several cellular pathways like cell cycle, apoptosis and Wnt/Hedgehog self-renewal (Pal et al. 2012; Sur et al. Amarogentin, the bitterest plant glycoside known, is 2016). During restriction of the liver carcinogenesis the active compound of Swertia chirata (chirata) which amarogentin could upregulate the expression of several had widely been used in Ayurvedic medicines since antagonists and downregulate several agonists of ancient times (Medda et al. 1999). The anticarcino- growth promoting pathways. To the best of our genic activity of this compound was first evident by its knowledge, mechanism of this expression modulation anti-proliferative and pro-apoptotic activity in a mouse by amarogentin is not yet known. skin carcinogenesis model (Saha et al. 2004). It was Thus, in this study attempts have been made to evident in our previous studies that amarogentin could understand, the mechanism of upregulation of cell restrict progression of experimental mouse liver car- cycle G1/S phase associated genes like LIMD1, P16 cinogenesis at mild dysplastic stage by modulating and RBSP3 in CCl4/NDEA-induced liver cancer mouse Supplementary Information: The online version contains supplementary material available at https://doi.org/10.1007/ s12038-021-00176-0. http://www.ias.ac.in/jbiosci 53 Page 2 of 11 Debolina Pal et al. model as well as in liver cancer cell line HepG2. It as mentioned in our previous study (Pal et al. 2012; Sur seems that amarogentin could modulate the epigenetic et al. 2016). Dose of the amarogentin for animal pathways like DNA methylation, histone deacetylation, experiment was 0.2 mg/kg body weight (supplemen- etc., as seen in case of other natural products, viz. tary table 1). EGCG, valproic acid, curcumin, etc., in their chemo- For in vitro experiment two doses of amarogentin preventive and anti-carcinogenic activities (Carlos- (i) 75 lg/ml (AM1) (30% cell death) and (ii) 100 lg/ Reyes et al. 2019; Reuter et al. 2011; Khan et al. 2015; ml (AM2) (50% cell death) were selected on the basis Lee et al. 2012; Jiang et al. 2015). of cytotoxicity assay using Sulforhodamine B (SRB) For this reason, in this study at first promoter dye (supplementary table 1) (Sur et al. 2016). methylation of LIMD1, P16 and RBSP3 genes were analyzed in CCl4/NDEA-induced mouse liver lesions with/without amarogentin treatment. Then, expression 2.3 Experimental groups for in vivo study of DNMT1 and HDACs were analyzed in the liver lesions. The in vivo results were further validated Group I (control group): Mice without any treatment. in vitro using human hepatocellular carcinoma cell line Group II (carcinogen control group): Mice were HepG2. injected intraperitoneally (i.p.) with Carbon tetrachlo- Finally, our data showed that reduced expression of ride (CCl4) successively for 4 days followed by DNMT1 might lead to upregulation of the cell cycle N-nitrosodiethylamine (NDEA) i.p. injection weekly inhibitors through promoter hypo-methylation during 75 mg/kg body weight for three successive weeks the restriction of liver carcinogenesis by amarogentin. followed by 100 mg/kg body weight for another 3 successive weeks (supplementary figure 1). Group III (pre-initiation group): Amarogentin twice a 2. Materials and methods week was administered orally for 15 days before 1st carcinogen application. Then, amarogentin treatment 2.1 Animal care was continued once a week during carcinogen appli- cation, which was continued for the rest of the exper- Adult (6 weeks) inbred female Swiss albino mice imental period. Treatment schedule of CCl4 and NDEA (average weight 25 g) were obtained from animal was same as Group II (supplementary figure 1). house of Chittaranjan National Cancer Institute, Group IV (post-initiation group): In this group Kolkata, India. Animals were maintained at 25 ± 5°C amarogentin treatment started after completion of car- temperature under alternating 12 h light/dark cycle with cinogen schedule and continued orally once a week 45–55% humid conditions. Food pellets (Lipton India throughout experimental procedure. Application and Ltd.) and drinking water was provided. Animal han- dose of amarogentin is same as Group III. Carcinogen dling and experimental protocol were undertaken as per schedule was same as Group II (supplementary guidelines of the Institutional Ethical Committee figure 1). (Registration No. 1774/GO/RBi/S/14/CPCSEA) regis- Number of mice in each group was 12. Mice from tered under CPCSEA, India (Committee for the pur- different experimental groups were sacrificed at 10th, pose of control and supervision of the experiments on 20th and 30th weeks of first carcinogen application and animals). Mice were observed regularly for their at each time point three mice were sacrificed from each wellbeing, body weight, toxicity, and survival. group. Liver tissues were dissected out and used for different experiment. Animals for experimental pur- poses were approved by institutional animal ethics 2.2 Drug solution preparation and concentrations committee. used for in vivo and in vitro studies

Amarogentin 1 mg/ml stock solutions were prepared in 2.4 Sample collection 10% ethanol in water. Extraction and purification of amarogentin ([99% pure in high-performance liquid Mice from different experimental groups were sacri- chromatography analysis) were done at the National ficed at 10th, 20th and 30th week. Liver was dissected Research Institute for Ayurvedic Drug Development, out and affected liver lesions were collected and divi- Kolkata, India. The amarogentin doses for in vivo and ded for histopathological analysis, RNA, DNA and in vitro experiments were selected by toxicity analysis protein isolation. Amarogentin restricted liver carcinogenesis by epigenetic modification Page 3 of 11 53 2.5 Histopathology orthovanadate]. Equal amount of was sepa- rated by 10–12% SDS–PAGE and then transferred to The formalin fixed tissue samples were processed polyvinylidine difluoride (PVDF) membrane (Milli- conventionally to prepare paraffin block. Tissue sec- pore, MA). Membranes were incubated with 3–5% tions (3–4 lm) were taken for hematoxylin-eosin (HE) nonfat dry milk for 1–2 h at room temperature followed staining and histopathological evaluation. by overnight incubation at 4°C with desired primary antibodies (Supplementary document) (1:500–1:1000 in 1% nonfat dry milk) and corresponding HRP-con- 2.6 Cell line and Amarogentin treatment condition jugated secondary antibodies (Supplementary docu- ment) (1:2000–1:5000 in 1% nonfat dry milk) for 2 h at Human liver cancer cell line HepG2 was obtained from room temperature (Pal et al. 2012). The protein bands National Centre for Cell Sciences, Pune, India. The cell were then visualized using luminol reagent and line was maintained in Dulbecco’s Modified Eagle’s autoradiographed on X-ray film (Kodak, Rochester, Medium (DMEM) supplemented with 10% FBS, 2 mM NY). All the immuno-blotting experiments were per- glutamine, 100 U Penicillin and 100 lg/ml of strepto- formed in three replicates. The band intensities were mycin. Cells were grown at 37°C in a humidified CO2 quantified using densitometric scanner (Bio Rad GS- incubator with 5% CO2. Re-feeding with fresh growth 800, Hercules, California). Peak densities of the pro- medium and subculturing (using 0.05% trypsin- teins of interest were normalized using peak density of EDTA) of cells was done as per requirement. loading control a-tubulin for both mice and human.

2.7.3 Immunocytochemical analysis: HepG2 cells 2.7 Expression analysis (20,000 cells/cover slip) were grown on sterile cov- erslip for 24 h. Cells were treated with amarogentin 2.7.1 mRNA expression analysis: Total RNA was (at IC50 and IC30 dose) for 48 hrs. Coverslips were extracted from liver lesions and HepG2 cells (19106) fixed with methanol at -208C. Before staining cov- with or without amarogentin treatment using TRIzol erslips were rehydrated and washed with phosphate reagent according to the standard protocol (Pal et al. buffer saline (PBS) for 30 min. Endogenous peroxi- 2012). cDNA was synthesized from the 10 lg of total dase was blocked by 0.3% H2O2 in 1X PBS for RNA with Super Script III Reverse Transcriptase (In- 5–10 min. Cells were then permeabilized with 0.1% vitrogen/Life technology) according to standard pro- Triton-X for 15 min. The non-specific binding sites tocol (Pal et al. 2012). expression was carried out were blocked by 3–5% BSA for 1–2 h followed by with power SYBR Green PCR Master Mix (Applied overnight incubation with primary antibodies (sup- Biosystems/Life technology, USA) in real-time PCR plementary tables) (1:80–1:100 in 1% BSA) and (ABI Prism 7500, Applied Biosystems, Foster City, FITC conjugated secondary antibody (supplementary CA) using specific primers (supplementary table 2). tables) (1:500 in 1% BSA) for 1 h. Cells were then Relative gene expression data were analyzed using the treated with nuclear staining dye DAPI (1lg/ml) for 2-DDCT method (Livak and Schmittgen 2001). Relative 1 min and washed (Sur et al. 2016). Finally, cover expression analysis was performed in three replicates. slips were glycerol mounted on a clean glass slides b2-microglobulin gene (B2M) was used as endogenous and photographed using fluorescence microscope control for both mice and human. Relative expression (Leica DM 4000 B, Germany). was graphically represented against each experimental group. 2.8 DNA isolation and promoter methylation 2.7.2 Protein extraction and western blot analysis: analysis Protein was extracted from liver lesions and HepG2 cells (1 9 106) with or without amarogentin treatment. High molecular-weight DNA was extracted from liver Cell lysate were prepared by sonication with RIPA lesions and HepG2 cells (1 9 106 cells) treated with or buffer [25 mM Tris-HCl (pH 7.6), 150 mM NaCl, 0.1% without amarogentin [at IC 30 (AM1) and IC50 (AM2) Triton-X 100, 1mM EDTA, 0.1% SDS, 1% sodium dose] by proteinase-K digestion, followed by phenol/ deoxycholate, 1l/mL aprotinin, 1l/mL leupeptin, chloroform extraction method according to standard 1mM PMSF, 10mM NaF, and 1mM sodium procedure (Dasgupta et al. 2002). 53 Page 4 of 11 Debolina Pal et al.

Figure 1. Methylation analysis of LIMD1, P16 and RBSP3 of mouse liver lesion by methylation sensitive restriction digestion (MSRA) method (A) In the left panel, schematic diagram showing promoter region of set II primer position (horizontal arrows), numbers of SmaI restriction site (small down arrows), PCR product size, transcription start site (?1) and translation start site (ATG) (table 2) (supplementary table 2). In the right panel, corresponding gel images showing band patterns of undigested (U) and digested (D) samples of each group, Group I, II, III and IV at 30th week only. pUC 19 plasmid Msp1 digested is loaded as marker (M) in the left most lane, band sizes are mentioned. (B) Methylation analysis of LIMD1, RBSP3 and P16 in mouse liver lesion by real-time PCR. Percentage of hypomethylation [Ct [digested]/Ct [undigested] 9100] is plotted for every sample and for specific gene. Samples run in triplicate and mean ± SD (standard deviation) plotted.

Promoter methylation status of the genes was analyzed table 2), the digestion sites in each set of primers for each in CpG rich islands in the promoter region of the enzyme are marked (table 2). PCR products were ana- respective genes by both qualitative and quantitative lyzed on 2% agarose gel, visualized under gel docu- methods (Dasgupta et al. 2019). The qualitative methy- mentation system (Bio Rad) and photographed. lation analysis was done by PCR-based methylation Methylation quantitation was done using real time sensitive restriction analysis (MSRA) method using PCR (ABI prism 7500, Applied Biosystems, Foster methylation sensitive restriction enzyme HpaII (Roche City, CA) amplification (30 cycles) of digested and Diagnostics, Germany) and SmaI (Sibenzyme, Russia). mock digested samples using Power SYBR green Mock digestion was done with each sample without any (Applied Biosystems) (Thomassin et al. 2004; Das- restriction enzyme. After digestion both digested and gupta et al. 2019). Each sample was run in triplicate. mock digested samples were alcohol precipitated and The percentage of hypomethylation was calculated by dissolved with tris EDTA pH 7.4. These samples were taking the ratio of the cycle threshold (Ct) values of the PCR amplified (30 cycles) using specific primers (sup- digested DNA sample with its respective undigested plementary table 2). Two sets of primers were used in DNA, presuming the undigested DNA as 100% case of DNA from mouse liver lesions (supplementary methylated. Amarogentin restricted liver carcinogenesis by epigenetic modification Page 5 of 11 53 Table 1. Methylation status of LIMD1, P16 and RBSP3in different experimental groups with different methylation sensitive restriction enzymes

LIMD1 (Digested with) P16 (Digested with) RBSP3 (Digested with) Experimental Experimental Numbers of mice groups time points in each group Hpa II Sma I Hpa II Sma I Hpa II Sma I Group I 10th week 2 ------Group II 3 ------Group III 3 ------Group IV 3 ------Group I 20th week 2 ------Group II 3 ------Group III 3 ------Group IV 3 ------Group I 30th week 2 ------Group II 3 -?-?-- Group III 3 ------Group IV 3 ------

‘‘ ?’’ Methylated; ‘‘-’’ Unmethylated.

The percentage of hypomethylation of a particular revealed larger cell size, nuclear size, and derangement gene was calculated using the following formula: (Ct in tissue architecture, indicating moderate dysplasia in [digested]/Ct [undigested] 9100). Group II mice at 10th week. Severe dysplastic changes Higher the percentage of hypomethylation, lower the were evident at 20th week in this group. Hepatocellular frequency of methylation and vice versa. The methy- carcinoma was evident at 30th week with some lation status of the genes in HepG2 cells was measured necrotic changes (supplementary figure 2). The liver by MSRA method with conventional PCR and using lesions of Group III mice were restricted to mild dys- one restriction enzyme HpaII and specific primers plastic changes up to 30th week (supplementary fig- (Supplementary document). Here primers specific for ure 2). In Group IV moderate dysplastic liver lesions b-3A-ADAPTIN gene (K1) (445 bp) and RAR b2 gene were evident at 10th week, whereas mild dysplastic (K2) (229 bp) were used as digestion and integrity liver lesions were evident at 20th week and 30th week controls respectively (Loginov et al. 2004). (supplementary figure 2).

2.9 Statistical analysis 3.2 Effect of amarogentin on promoter methylation of LIMD1, P16 and RBSP3 during liver Results are expressed as mean ± standard deviation. carcinogenesis at pre- and post-initiation stages Differences between groups were determined by Chi- square trends. P value \0.05 and \0.001 were con- In our previous study (Pal et al. 2012), it was evident sidered as statistically significant. Treated groups were that expression (RNA/protein) of LIMD1, P16 and compared with control group. RBSP3 were gradually downregulated in liver lesions of Group II (supplementary figure 3). On the contrary, the expression of the genes was found to be upregu- 3. Results lated in the liver lesions of Group III and IV (supple- mentary figure 3) (Pal et al. 2012). To understand the 3.1 Restriction of liver carcinogenesis mechanism of upregulation of the genes after treatment by amarogentin in mouse model system of amarogentin, their promoter methylation status was analyzed by MSRA method at first. The promoter As seen in our previous study (Pal et al. 2012), regions of LIMD1 and P16 showed methylation at histopathological analysis (supplementary figure 2) SmaI restriction sites in the liver lesions of 30th week 53 Page 6 of 11 Debolina Pal et al. of Group II (carcinogen control) (figure 1A; table 1). Group I (figure 2A). On the contrary, comparable m-RNA No such change was seen in RBSP3 gene (table 1) expression of DNMT1 was seen in the liver lesions of (figure 1A). However, absence of promoter methylation Group III and IV with respect to liver samples of Group I. in the genes was seen in the liver lesions of Gr. III and In case of HDAC1 and HDAC2, no significant chan- IV (table 1; figure 1A). ges in their m-RNA expression were seen in the liver This result was further validated by quantitative lesions of Group II, III and IV with respect to liver promoter methylation analysis. Significant decrease in samples of Group I (figure 2A). In western blot analysis hypo-methylation frequencies, i.e. increase in methy- of the samples at 30th week, the expression of DNMT1, lation frequencies, of LIMD1 and P16 were seen in the HDAC1 and HDAC2 proteins showed concordance liver lesions of Group II at 30th week with respect to with respective mRNA expression (figure 2B). the liver samples of Group I (figure 1B). On the other Thus, it seems that DNMT1 upregulation might have hand, hypo-methylation frequencies of LIMD1 and P16 important role in promoter hypermethylation of LIMD1 were significantly increased in the liver lesions at 30th and P16. week of Group III and Group IV (figure 1B). No such change was evident in RBSP3 (figure 1B). 3.4 Validation of promoter methylation status of LIMD1, RBSP3 and P16 in liver cancer cell line 3.3 Effect of amarogentin on expression HepG2 in presence of amarogentin of DNMT1, HDAC1 and HDAC2 during liver carcinogenesis at pre- and post-initiation stage For validation of promoter hypomethylation of LIMD1, RBSP3 and P16, HepG2 cells were treated with To understand the effect of amarogentin on epigenetic amarogentin at IC 30 (AM1) and IC50 (AM2) con- regulation during restriction of liver carcinogenesis, if centrations. In HepG2 cells, promoter regions of any, the expression of different epigenetic modulatory LIMD1 and P16 were methylated, but no such change genes, e.g. DNMT1, HDAC1 and HDAC2 were analyzed was seen in RBSP3 gene promoter (figure 3A). How- in the liver lesions at different time points. Gradual ever, after amarogentin treatment the promoter regions upregulation in expression of DNMT1 was seen in Group of LIMD1 and P16 were seen to be hypomethylated II with significant increase at 30th week with respect to (figure 3A). As a result, there was significant increase

Figure 2. Expression analysis of DNMT1, HDAC1 and HDAC2 in mice liver lesion. (A) mRNA expression analysis of the above genes of all experimental groups. Y axis denotes relative expression. ** p value \ 0.01. (B) Western blot analysis showing expression of corresponding proteins of 30th week samples only along with loading control a-tubulin. Peak densities are marked below the autoradiographs Amarogentin restricted liver carcinogenesis by epigenetic modification Page 7 of 11 53

Figure 3. Methylation analysis of LIMD1, P16 and RBSP3 of liver cancer cell line HepG2 by methylation sensitive restriction digestion (MSRA) method. (A) In the left panel, gel images showing band patterns of undigested [U] and digested [D] samples of each experimental group, HepG2, Lower IC50 value of amarogentin (AM1); IC50 value of amarogentin (AM2) for LIMD1, P16 and RBSP3 respectively. pUC 19 plasmid Msp1 digested is loaded as marker [M] in the left most lane, band sizes are 501/489 bp; 404 bp; 331 bp; 242 bp; 190 bp; 147 bp; 111/110 bp. (B) mRNA analysis of the above genes in treated groups compared with untreated control. Y axis denotes relative expression. **denotes the p value \ 0.01. (C) Western blot analysis showing expression of LIMD1, P16, RBSP3 proteins along with ppRB (Ser 807/811) and (Ser 567) in different groups. a-Tubulin was used for loading control. (D) mRNA analysis of LIMD1, RBSP3, and P16 after treatment of HepG2 cells with DNMT1 inhibitor 5-aza-2-deoxycytidine at three different concentrations 5, 10 and 20 lM. Different concentrations of 5-aza-2-deoxycytidine and relative expression are plotted in X and Y axis. (E) mRNA Expression analysis of DNMT1, HDAC1 and HDAC2 in HepG2 cell line in amarogentin-treated groups compared with untreated control. Relative expression is plotted in Y axis. (F) Western blot analysis showing expression of corresponding proteins (DNMT1, HDAC1 and HDAC2) along with loading control a-tubulin. 53 Page 8 of 11 Debolina Pal et al. in expression (mRNA/protein) of LIMD1 and P16 after methylated in the liver lesions at 30th week, i.e. amarogentin treatment in dose dependent manner hepatocellular carcinoma. Similar phenomenon has (figure 3B, C). In addition, RBSP3 expression (mRNA/ also been seen in human liver cancer cell line HepG2. protein) was also increased after treatment with Downregulation of LIMD1 due to promoter hyper- amarogentin (figure 3B, C). The increased expression methylation was reported in many cancers including of these proteins in HepG2 cells after amarogentin head and neck, lungs, breast, cervix etc. (Ghosh et al. treatment was also confirmed by reduced phosphory- 2010a, b; Foxler et al. 2011; Sharp et al. 2008; lation of RB at Ser 807/811 and Ser 567 (figure 3C). In Chakraborty et al. 2018). However, in liver cancer the immunocytochemical analysis, the cytoplasmic/nuclear promoter methylation profile of LIMD1 has not yet expression of LIMD1, P16 and RBSP3 were increased been studied. In several studies, P16 hypermethylation after amarogentin treatment (supplementary figure 4). and subsequent downregulation were found to be To confirm the role of amarogentin as epigenetic associated with hepatocellular carcinomas (Jin et al. modulator, HepG2 cells were treated with a standard 2000; Zang et al. 2011). Downreglation of RBSP3 due demethylating agent 5-aza-2-deoxycytidine at different to promoter hypermethylation was found to be concentrations. Like amarogentin treatment, gradual associated with breast, cervical and head and neck increase in mRNA expression of LIMD1 and P16 were carcinogenesis (Sinha et al. 2008;Mitraet al. 2010; seen with increase in concentrations of 5-aza-2- Ghosh et al. 2010a, b). However, its promoter deoxycytidine in HepG2 cells with slight increase in methylation profile has not yet been studied in liver RBSP3 expression (figure 3D) in HepG2 cells. Thus, it cancer. So, the absence of methylation in the pro- seems that amarogentin is an epigenetic modulator like moter region of RBSP3 as seen in our study needs 5-aza-2-deoxycytidine. to be confirmed by another primer set or restriction enzyme. In the restricted liver lesions of both pre (Group III) 3.5 Expression analysis of DNMT1, HDAC1 and post (Group IV) initiation stages by amarogentin, and HDAC2 in vitro after treatment expression (mRNA/protein) of LIMD1, P16 and with amarogentin RBSP3 genes were seen to be upregulated (Pal et al. 2012). In MSRA, absence of promoter methylation of To understand the mechanism of epigenetic modulation LIMD1 and P16 genes were seen in the liver lesions at of amarogentin the expression (mRNA/protein) of 30th week after the amarogentin treatment. However, DNMT1, HDAC1 and HDAC2 were analyzed in in quantitative methylation analysis comparable HepG2 cells. Significant decrease in mRNA expression methylation frequencies of LIMD1 and P16 genes were of DNMT1 was seen after amarogentin treatment seen with respect to normal liver (Group I), indicating (figure 3F). The protein expression also showed restoration of normal function of liver compare to downregulation after amarogentin treatment in a con- carcinogen control (Group II). Similarly, amarogentin centration dependent manner (figure 3G). However, the could also upregulate the expression (mRNA/protein) expression (mRNA/protein) of HDAC1 and HDAC2 of LIMD1, P16 and RBSP3 genes in HepG2 cells, did not change considerably (figure 3F, G). Thus, it probably through promoter hypo-methylation (LIMD1/ indicates that amarogentin could downregulate DNMT1 P16) resulting downregulation of ppRBSer 807/811 expression. and ppRB Ser 567 expression as seen in our study. Like amarogentin, upregulation of the genes has been seen in the HepG2 cells after treatment with demethylating 4. Discussion agent/DNMT1 inhibitor 5-aza-2-deoxycytidine. Thus, it seems that amarogentin has comparable function as In this study, the mechanism of upregulation of the 5-aza-2-deoxycytidine. Interestingly, amarogentin cell cycle inhibitors, viz. LIMD1, P16 and RBSP3, could downregulate the expression (mRNA/protein) of during restriction of CCl4/NDEA-induced mouse liver DNMT1 in HepG2 cells, indicating it as potent DNA carcinogenesis by amarogentin treatment was evalu- methylation modifier like other natural products like ated. It was evident that the expression (mRNA/pro- EGCG, Genistein, curcumin, etc. (Ratovitski 2017; tein) of the genes was downregulated during the Mirza et al. 2013; Fang et al. 2003). In addition, the CCl4/NDEA-induced mouse liver carcinogenesis insignificant changes in expression of HDAC1/2 in the (Group II) (Pal et al. 2012). Among the genes, the liver lesions indicate that histone deacetylation might promoter regions of LIMD1 and P16 were hyper- not be the primary mechanism in restriction of liver Amarogentin restricted liver carcinogenesis by epigenetic modification Page 9 of 11 53

Figure 4. Schematic diagram showing the role of amarogentin in epigenetic regulation during mice experimental liver carcinogenesis model and human liver cancer cell line HepG2.

Table 2. Recognition sites, amount of DNA and digestion conditions are mentioned here for two methylation sensitive restriction enzymes used in this study using two primer sets (I and II)

Total No. Total No. of sites in of sites in PCR each amplified promoter region region

Name of Set Set Set Set the genes Restriction enzyme HpaII I II Restriction enzyme SmaI I II LIMD1 Recognition DNA Digested at 2 7 Recognition DNA Digested at –2 P16 site CCGG amount 378C 112site amount 258C for 4 –1 RBSP3 100 ng overnight 36CCCGGG 100 ng h –1 53 Page 10 of 11 Debolina Pal et al. carcinogenesis by amarogentin. Thus, further work is Dasgupta S, Mukherjee N, Roy S, Roy A, Sengupta A, et al. needed to understand the role of HDACs during 2002 Mapping of the candidate tumor suppressor genes’ amarogentin treatment (Eckschlager et al. 2017). loci on human 3 in head and neck squamous Thus, our data suggest that the upregulation of cell carcinoma of an Indian patient population. Oral the LIMD1, P16 and RBSP3 genes during restric- Oncol. 38 6–15 tion of liver carcinogenesis and HepG2 cells by Eckschlager T, Plch J, Stiborova M and Hrabeta J 2017 Histone deacetylase inhibitors as anticancer drugs. Int. amarogentin treatment might be due to promoter J. Mol. Sci. 18 E1414 hypomethylation through downregulation of Fang MZ, Wang Y, Ai N, Hou Z, Sun Y, et al. 2003 Tea DNMT1 expression (figure 4). However, impor- polyphenol (-)-epigallocatechin-3-gallate inhibits DNA tance of amarogentin on other epigenetic regulators methyltransferase and reactivates methylationsilenced during restriction of the liver carcinogenesis needs genes in cancer cell lines. Cancer Res. 63 7563–7570 to be studied to understand its role as epigenetic Foxler DE, James V,Shelton SJ, Vallim TQ, Shaw PE and Sharp modulator in detail. TV 2011 PU.1 is a major transcriptional activator of the tumour suppressor gene LIMD1. FEBS Lett. 585 1089–1096 Ghosh A, Ghosh S, Maiti GP, Sabbir MG, Zabarovsky ER, 5. Conclusion et al. 2010 Frequent alterations of the candidate genes hMLH1, ITGA9 and RBSP3 in early dysplastic lesions of Our data suggest that DNA hypomethylation might be head and neck: clinical and prognostic significance. one of the mechanisms of restriction of liver carcino- Cancer Sci. 101 1511–1520 genesis by amarogentin (figure 4). In addition, the Ghosh S, Ghosh A, Maiti GP, Mukherjee N, Dutta S, et al. evaluation of acetylation profile of histones in the 2010 LIMD1 is more frequently altered than RB1 in head restricted liver lesions needs to be studied to under- and neck squamous cell carcinoma: clinical and prognos- stand the epigenetic modulatory role of amarogentin in tic implications. Mol. Cancer 9 9–58 Jiang A, Wang X, Shan X, Li Y, Wang P, et al. 2015 Curcumin detail. Reactivates Silenced Tumor Suppressor Gene RARbeta by Reducing DNA Methylation. Phytother. Res. 29 1237–1245 Jin M, Piao Z, Kim NG, Park C, Shin EC, et al. 2000 p16 is Acknowledgements a major inactivation target in hepatocellular carcinoma. Cancer 89 60–68 Authors are grateful to the Director, Chittaranjan Khan MA, Hussain A, Sundaram MK, Alalami U, Gunasekera National Cancer Institute, Kolkata, for his kind support D, et al. 2015 (-)-Epigallocatechin-3-gallate reverses the for the work. Financial support for this work was expression of various tumor-suppressor genes by inhibiting provided by grant 09/030(0074)/2014 EMR-I from DNA methyltransferases and histone deacetylases in CSIR, New Delhi, to Debolina Pal and NASI Senior human cervical cancer cells. Oncol. Rep. 33 1976–1984 Scientist Platinum Jubilee Fellowship (2020) to Dr. Lee YH, Kwak J, Choi HK, Choi KC, Kim S, et al. 2012 Chinmay Kumar Panda. EGCG suppresses prostate cancer cell growth modulating acetylation of androgen receptor by anti-histone acetyl- transferase activity. Int. J. Mol. Med. 30 69–74 Livak KJ and Schmittgen TD 2001 Analysis of relative gene References expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) method. Methods 25 402–408 Carlos-Reyes A´ ,Lo´pez-Gonza´lez JS, Meneses-Flores M, Medda S, Mukhopadhyay S and Basu MK 1999 Evaluation Gallardo-Rinco´n D, Ruı´z-Garcı´a E, et al. 2019 Dietary of the in-vivo activity and toxicity of amarogentin, an Compounds as Epigenetic Modulating Agents in Cancer. antileishmanial agent, in both liposomal and niosomal Front Genet. 10 79 forms. J. Antimicrob. Chemother. 44 791–794 Chakraborty C, Mitra S, Roychowdhury A, Samadder S, Mirza S, Sharma G, Parshad R, Gupta SD, Pandya P, et al. Dutta S, et al. 2018 Deregulation of LIMD1-VHL- 2013 Expression of DNA methyltransferases in breast HIF1alpha-VEGF pathway is associated with different cancer patients and to analyze the effect of natural stages of cervical cancer. Biochem. J. 475 1793–1806 compounds on DNA methyltransferases and associated Dasgupta H, Islam S, Alam N, Roy A, Roychoudhury S and proteins. J. Breast Cancer 16 23–31 Panda CK 2019 Hypomethylation of mismatch repair Mitra S, Mazumder Indra D, Bhattacharya N, Singh RK, genes MLH1 and MSH2 is associated with chemotoler- Basu PS, et al. 2010 RBSP3 is frequently altered in ance of breast carcinoma: Clinical significance. J. Surg. premalignant cervical lesions: clinical and prognostic Oncol. 119 88–100 significance. Genes, Cancer 49 155–170 Amarogentin restricted liver carcinogenesis by epigenetic modification Page 11 of 11 53 Pal D, Sur S, Mandal S, Das A, Roy A, et al. 2012 Sinha S, Singh RK, Alam N, Roy A, Roychoudhury S, et al. Prevention of liver carcinogenesis by amarogentin 2008 Frequent alterations of hMLH1 and RBSP3/HYA22 through modulation of G1/S cell cycle check point and at chromosomal 3p22.3 region in early and late-onset induction of apoptosis. Carcinogenesis 33 2424–2431 breast carcinoma: clinical and prognostic significance. Ratovitski EA 2017 Anticancer Natural Compounds: Molec- Cancer Sci. 99 1984–1991 ular Mechanisms and Functions. Part I. Curr. Genomics 18 2 Sur S, Pal D, Banerjee K, Mandal S, Das A, et al. 2016 Reuter S, Gupta SC, Park B, Goel A and Aggarwal BB 2011 Amarogentin regulates self renewal pathways to restrict Epigenetic changes induced by curcumin and other liver carcinogenesis in experimental mouse model. Mol. natural compounds. Genes Nutr. 6 93–108 carcinogenesis. 55 1138–1149 Saha P, Mandal S, Das A, Das PC and Das S 2004 Evaluation Thomassin H, Kress C and Grange T 2004 MethylQuant: of the anticarcinogenic activity of Swertia chirata Buch. a sensitive method for quantifying methylation of Ham, an Indian medicinal plant, on DMBA-induced mouse specific cytosines within the genome. Nucleic Acids skin carcinogenesis model. Phytother Res. 18 373–378 Res. 32 e168 Sharp TV,Al-Attar A, Foxler DE, Ding L, Thomas QDA, et al. Zang JJ, Xie F, Xu JF, Qin YY, Shen RX, et al. 2011 P16 2008 The chromosome 3p21.3-encoded gene, LIMD1, is a gene hypermethylation and hepatocellular carcinoma: a critical tumor suppressor involved in human lung cancer systematic review and meta-analysis. World J. Gastroen- development. Proc. Natl. Acad. Sci. USA 105 19932–19937 terol. 17 3043–3048

Corresponding editor: B J RAO