Loss of hepatic Sirt7 accelerates diethylnitrosamine (DEN)-induced formation of hepatocellular carcinoma by impairing DNA damage repair

Abstract

The mammalian sirtuin family (SIRT1–SIRT7) has shown diverse biological roles in the regulation and maintenance of genome stability under genotoxic stress. SIRT7, one of the least studied sirtuin, has been demonstrated to be a key factor for DNA damage response (DDR). However, conflicting results have proposed that Sirt7 is an oncogenic factor to promote transformation in cancer cells. To address this inconsistency, we investigated properties of SIRT7 in hepatocellular carcinoma (HCC) regulation under DNA damage and found that loss of hepatic Sirt7 accelerated HCC progression. Specifically, the number, size, and volume of hepatic tumor colonies in diethylnitrosamine (DEN) injected Sirt7-deficient liver were markedly enhanced. Further, levels of HCC progression markers and pro-inflammatory cytokines were significantly elevated in the absence of hepatic Sirt7, unlike those in the control. In chromatin, SIRT7 was stabilized and colocalized to damage site by inhibiting the induction of γH2AX under DNA damage. Together, our findings suggest that SIRT7 is a crucial factor for DNA damage repair and that hepatic loss-of-Sirt7 can promote genomic instability and accelerate HCC development, unlike early studies describing that Sirt7 is an oncogenic factor.

INTRODUCTION

Globally, liver cancer is the sixth most common cancer and the third leading cause of cancer fatalities (1). Hepatocellular carcinoma (HCC), the major histological subtype of primary liver cancers, frequently arises as a result of chronic injury and inflammation that can promote DNA damage and chromosomal instability (2) by initiating signaling cascades known as DNA damage response (DDR), DNA repair, and cell cycle arrest, ultimately leading to cell death or senescence (3-5).

The mammalian sirtuin family (SIRT1–SIRT7) has diverse biological roles. Genome stability regulation under genotoxic stress is a key function of sirtuins (6, 7). SIRT7, the least studied sirtuin, resides in the nucleus along with SIRT1 and SIRT6 (7, 8), wherein it exhibits catalytic activities as a selective histone H3 lysine 18 (H3K18Ac) deacetylase at DNA damage sites towards proteins involved in DNA damage and repair (9-11). In chromatin, SIRT7 also functions as a desuccinylase of histone H3 lysine 122 and a deglutarylase of histone H4 lysine 91 to promote chromatin compaction necessary for DNA damage repair (8, 9).

DDR actively participates in HCC carcinogenesis. Notably, various HCC-associated risk factors can promote DNA damage, formation of DNA adducts, and chromosomal aberrations (10). Hence, alterations in DDR pathways may trigger hepatocarcinogenesis and facilitate advanced HCC progression. Sirt7 is generally considered as an oncogene in HCC as it is upregulated in human HCC samples. Its expression increased with increasing tumor grade (11). However, effects of SIRT7 deficiency on DNA damage-induced HCC development and progression in vivo remain unclear. Related molecular mechanisms have not been elucidated either.

Diethylnitrosamine (DEN) is a DNA-alkylating carcinogen widely used to study HCC progression. It mimics human HCC by stimulating inflammatory responses and hyperproliferation (12). Therefore, the objective of this study was to investigate pathological characteristics of SIRT7 in the regulation of hepatic tumorigenesis driven by DEN-induced DNA damage in a liver-specific Sirt7 null mouse model.

RESULTS

Sirt7 depletion in liver accelerates tumorigenesis in DEN-induced hepatocarcinogenesis

To investigate whether alteration of Sirt7 plays a role in hepatocarcinogenesis and its involvement in HCC progression, we first generated a Sirt7 liver-specific knockout mouse model (Sirt7^hep−/−) and exposed it to chronic DEN treatment (Fig. 1A). Ten weeks after DEN administration, total body and liver weights did not differ between the two genotypes (Fig. 1B). However, the ratio of liver to body weight was slightly heavier in Sirt7^hep−/− mice than in control mice possibly because of increased number and size of HCC lesions. This was further validated by gross anatomy analysis (Fig. 1C and Supplementary Fig. 1A). In addition, observable numbers of tumor nodules increased with tumor volumes in DEN-induced Sirt7^hep−/− mice compared to those in their wild-type (WT) littermates, as demonstrated by liver micro-CT images (Fig. 1D, E and Supplementary Fig. 1B, C). Collectively, these phenotypic changes suggest that deficiency of Sirt7 in the liver can potentiate the promotion of hepatic tumorigenesis driven by DEN-induced DNA damage.

Fig. 1.

Loss of hepatic Sirt7 promotes hepatic tumorigenesis in DEN-injected mice. (A) Schematic diagram showing diethylnitrosamine (DEN)-induced hepatocarcinogenesis mouse model. MGMT: O⁶-Methylguanine-DNA Methyltransferase. (B) Body and normalized liver weights at sacrifice (n = 7 per group). (C) Representative images of livers from 10-month-old Sirt7^hep+/+ and Sirt7^hep−/− mice. (D) Liver micro-CT images captured from 9-month-old Sirt7^hep+/+ and Sirt7^hep−/− mice. (E) Quantitative analysis of hepatic tumors using micro-CT. Data in (B) and (E) represent the mean ± SEM. *P ≤ 0.05, **P ≤ 0.01, unpaired Student’s t-test.

Hepatic Sirt7-knockout mice exhibit hallmarks of HCC progression

To further evaluate effects of Sirt7 deletion on HCC onset and progression, histological analysis was performed for liver tumors. Sirt7^hep−/− mice showed more malignant or pre-malignant nodular in cancerous lesions (Fig. 2A), with increased central-to-portal bridging fibrosis in non-cancerous liver lesions compared to that in WT littermates (Supplementary Fig. 2A, B). In histological analysis of Sirt7^hep−/− livers, levels of Ki-67 (a proliferation marker) and alpha fetoprotein (AFP, an HCC marker) were elevated (Fig. 2B, C). Consistently, serum AFP levels exhibited a significant increase (Fig. 2D). Gene expression levels of hallmarks of HCC progression, including Afp, Kitl, Vegf, Pdgfpb, and Ki67, were also markedly elevated upon Sirt7 deletion (Fig. 2E). Similarly, pro-inflammatory cytokines known to play important roles in HCC development were also increased in Sirt7^hep−/− mice compared to those in WT mice (Supplementary Fig. 2C, D). Furthermore, expression levels of DNA damage related proteins (Trp53 and CDKN1A/p21) and proliferation makers (PCNA and AFP) were higher in livers of Sirt7-depleted mice compared with those in WT mice (Fig. 2F). Taken together, these findings suggest that Sirt7 depletion following DEN-induced DNA damage can lead to pro-tumorigenic inflammation in the liver, eventually accelerating hepatic carcinogenesis.

Fig. 2.

Effects of hepatic loss-of-Sirt7 in liver tumor. (A) The level of tumor development was measured by hematoxylin/eosin (H&E) staining of liver tumor obtained from Sirt7^hep+/+ and Sirt7^hep−/− mice (10-month-old male). (B, C) Ki67 staining and Alpha-fetoprotein (AFP) and in liver tumors of Sirt7^hep+/+ and Sirt7^hep−/− mice (10-month-old male). (D) ELISA showing serum AFP levels in Sirt7^hep+/+ and Sirt7^hep−/− mice (10-month-old male). (E) The qRT-PCR results using tumor tissues showing hepatic expression of hallmarks of HCC progression such as Afp, Kitl, Vegf, Pdgfpβ, and Ki67. (F) Western blot analysis of DNA damage and tumor proliferation-related molecules, including PCNA, AFP, Trp53, and p21, was performed in Sirt7^hep+/+ and Sirt7^hep−/− mice (10-month-old male). The quantification of western blot (right). Data in (D), (E), and (F) represent the mean ± SEM. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001. Unpaired Student’s t-test.

Sirt7 expression is positively correlated with DNA repair gene-set

To clarify the pathological involvement of Sirt7 based on our in vivo findings, we performed gene set enrichment analysis (GSEA; www.broadinstitute.org/gsea) using hepatic transcriptomes of BXD mouse strains (13). Results of GO and KEGG enrichment analysis confirmed that Sirt7 expression was highly involved in the DNA damage pathway (Fig. 3A) and positively correlated with DNA damage repair-related gene sets (Fig. 3B, C and Supplementary Fig. 3A). To bolster these findings, we conducted a re-analysis of the proteomics showing potential SIRT7 interactome (14). In accordance with transcriptomic analysis, functional clustering of protein interactions revealed SIRT7’s role in various biological processes such as RNA metabolism (16.8%), DNA damage and replication (14%), and ribosome and translation (10.5%) (Fig. 3D). Taken together, bioinformatics analysis strongly suggests that Sirt7 is essential for DNA damage repair.

Fig. 3.

Sirt7 expression is positively correlated with the DNA repair gene-set. (A-C) GSEA results of the hepatic transcriptomes of BXD-GRP mice showing a correlation between Sirt7 and DNA repair genes. (A) Bubble chart showing the normalized enrichment score (x-axis), false discovery rate (FDR; y-axis), and nominal P value (size of bubble; −log10). (B) GSEA plots of DNA damage repair gene sets. (C) Heat maps showing the expression of DNA repair genes in relation to Sirt7 expression levels in livers of BXD mice. (D) Pie charts showing functional clustering of protein-protein interactions.

SIRT7 is a crucial factor for DNA damage repair

To determine whether SIRT7 was a crucial factor for DNA damage repair, we blocked de novo protein synthesis using cycloheximide (CHX), a protein synthesis inhibitor, to determine the half-life of SIRT7. SIRT7 degradation was delayed upon treatment with bleomycin and etoposide (Ble + Eto) which induced DNA damage, whereas no difference was detected in the expression of SIRT6 (Fig. 4A, B), a DNA double-strand break (DSB) sensor (15). Interestingly, SIRT7 degradation significantly slowed and its half-life sustained for up to 4 h under DNA damage found mostly in chromatin fractions (Fig. 4C and Supplementary Fig. 4A), indicating that DNA damage might stabilize SIRT7 and shift its localization to the chromatin. We examined whether alterations in SIRT7 had potential to participate in DNA damage repair by examining the level of γH2AX induction under conditions of DNA damage in Sirt7-floxed (Sirt7L2/L2) primary hepatocytes (Fig. 4D). It was found that deletion of SIRT7 using adenovirus-mediated Cre-recombinase (Ad Cre) significantly increased γH2AX expression compared to the control (Ad GFP). However, reconstruction of SIRT7 in SIRT7 deleted primary hepatocytes markedly suppressed γH2AX induction (Fig. 4E). These findings support SIRT7’s crucial role in repairing DNA damage.

Fig. 4.

SIRT7 is a crucial factor in DNA repair. (A, B) Comparison of Protein Stability of SIRT7 and SIRT6 following DNA damage induced by treatment with bleomycin and etoposide for 0, 60, or 120 min. (C) Analysis of SIRT7 localization changes following induction of DNA damage for 4 hours using chromatin fractionation. (D, E) The effect of genetic aberrations, loss and gain of Sirt7, upon DNA damage indicated by γH2AX in isolated Sirt7^flox/flox primary hepatocytes. Adenovirus-mediated deletion of Cre recombinase (Ad Cre) and overexpression of SIRT7 (Ad GFP). Data are represented as mean ± SEM. **P < 0.01. Unpaired Student’s t-test.

DISCUSSION

Early studies have demonstrated that Sirt7 could play a pivotal role in hepatic metabolism (16, 17). The absence of Sirt7 has been associated with the onset of non-alcoholic fatty liver disease, a prevalent chronic liver condition in developed regions and a recognized risk factor for HCC (18). Here, we investigated the effect of hepatic Sirt7 deficiency in a mouse model of liver cancer in which HCC was driven by chronic exposure to DEN-induced DNA damage. Our results revealed a potential protective effect of Sirt7 against genetic aberrations in chronic HCC development. Regardless of the etiology, the majority of cirrhotic livers progress to HCC and over 90% of them show chromosomal abnormalities that are present universally in HCC. In damaged cells, chromosomal defects proliferate to form dysplastic nodules, which usually transform into HCC precursors. Therefore, increased genomic instability is a hallmark of HCC. Persistent DNA damage with defects in DNA repair contributes to genomic instability and eventually accelerates tumorigenesis (19).

To ensure genome stability, DSBs must be repaired. Unrepaired DNA lesions eventually lead to cancer pathologies (20). Several processes of DNA damage repair correspond to mechanisms implicated in HCC, such as stalled DNA replication forks by homologous recombination (HR) (21), base mismatches by mismatch repair (MMR) (22), and DBS, the most serious form of DNA damage (23), by non-homologous end joining (NHEJ) (24). The major pathway for repairing DSBs is NHEJ, which modifies broken DNA ends and ligates them together regardless of homology by deletions or insertions (25). Therefore, NHEJ is intrinsically mutagenic (26). Defects in DDR by NHEJ can lead to chromosomal instability and contribute to HCC (27).

SIRT7 appears to be a crucial player in DDR pathways as it has a histone deacetylase activity at DNA damage sites and additional catalytic activities to modify proteins involved in DNA damage and repair (9, 28). More specifically, SIRT7 displays a protective role in cancer cell survival under genotoxic stress-induced DNA damage by halting cell cycle to ensure DNA repair and by accumulating γH2AX and p53 while attenuating stress-activated protein kinases (p38 and JNK) to maintain genomic integrity (29). Of these, DDR-related genes are key to maintaining genomic stability. Loss of DDR can result in the onset and progression of cancer (30). Of note, Vazquez et al. have found that Sirt7^−/− cells with accumulating DNA mutations and replication stress can result in genome instability, leading to compromised NHEJ (28). Consistent with these findings, we also observed that hepatic Sirt7^−/− mice injected with DEN displayed accelerated HCC development compared to WT littermates while proteins related to DDR were significantly downregulated with liver primary Sirt7^−/− cells showing higher proliferative activities. In chromatin, we witnessed SIRT7 recruitment to form discrete nuclear foci that instantly colocalized with γH2AX, which reflected the induction of DSBs, triggering recruitment of various DNA-damage repair proteins to damaged sites (31) after DNA damage induction. Additionally, HCC progression in Sirt7^hep−/− mice caused the development of more visible numbers of tumor nodules with higher tumor volumes than that in WT mice, strongly suggesting that loss of Sirt7 upon DNA damage could affect tumor growth. This implies a higher susceptibility to carcinogenesis under Sirt7 deletion, which infers a significant contribution of SIRT7 in DSB pathways to protect tumorigenesis.

Inconsistent with such findings, however, increased SIRT7 expression is frequently observed in various human cancers (11, 32). In particular, SIRT7 was significantly upregulated and positively associated with HCC grade in a large cohort of patients (33). Zhao et al. have also observed that both Sirt7 mRNA and protein levels are significantly upregulated in most HCC tissues compared with those in adjacent non-tumoral liver tissues (34). Furthermore, Sirt7 expression is significantly correlated with poor overall survival rate and relatively lower in normal liver and primary human hepatocytes (35). These data suggest that Sirt7 might act as an oncogene during HCC development. Furthermore, SIRT7 is a selective deacetylase of H3K18, which is an epigenetic biomarker of aggressive tumors (36).

In the present study, we described a DNA damage induced Sirt7 deficiency HCC mouse model, in which a gene that participated in DDR probably through the NHEJ pathway upon DEN exposure, caused widespread chromosomal aberrations, disrupted cell cycle checkpoints, and accelerated liver carcinogenesis. The importance of SIRT7 in DNA damage repair, as identified in our study, suggests that this enzyme might function as a tumor suppressor.

To the best of our knowledge, a plausible explanation for these paradoxical results is that SIRT7 might have contradictory effects on cancer initiation and progression. SIRT7 is an important regulator of cancer cell growth and survival. It might act as a molecular bridge between genome stability and cancer progression. However, detailed mechanisms underlying these opposing effects remain largely unknown. They require further investigations. More systematic research is required to outline how SIRT7 function might change across cancer initiation and development at molecular and physiological levels. Such studies will enable us to develop novel therapeutic targets for clinical applications in HCC (37-39).

In conclusion, our findings indicate that Sirt7 plays a dual role in the liver. It appears to actively participate in the control of hepatic energy metabolism by directly influencing the expression of mitochondrial genes and managing endoplasmic reticulum (ER) stress (16, 17). Sirt7 also appears to be involved in preserving genomic stability in the liver, a tissue particularly vulnerable to the impact of diverse toxicants. However, further in-depth mechanistic studies are necessary.

MATERIALS AND METHODS

Animal experiments

All animal experiments were performed in accordance with the Swiss and EU ethical guidelines and were approved by the local animal experimentation committee of Canton de Vaud under license #2868 and also by the Institutional Animal Care and Use Committee (IACUC) of Sungkyunkwan University School of Medicine (SUSM) (code/SKKUIACUC 2021-07-47-1) and Gwangju Institute of Science and Technology (GIST-2023-027). Details of the experimental protocols are provided in the Supplementary Materials.

Generation of Sirt7^hep−/− mice and tumorigenesis

Liver-specific Sirt7 knockout mice (Sirt7^hep−/−) were generated as previously specified (17). To initiate the development of liver tumors, male mice of the C57BL/6 strain, specifically Sirt7^hep−/− and Sirt7^hep+/+ littermates, were subjected to intraperitoneal injections of DEN (Sigma) at a dose of 25 mg/kg body weight on day 14 postpartum as described (40, 41).

All mice were monitored weekly for changes in weight and activity. Nine months following the injection of DEN, liver tumor volumes and numbers induced by DEN were assessed utilizing micro-CT analysis. Ten months following DEN injection, all mice were humanely euthanized, and liver and blood tissues were collected for further analysis.

Primary mouse hepatocyte culture

In accordance with established procedures (17), primary liver cells from wild type C57BL/6J, S mice aged seven to nine weeks were procured. Subsequent to isolation, these cells were cultured in Media 199 with a composition of 4.5 g/L glucose, 10% fetal bovine serum, 0.1 mM NEAA, 25 mM HEPES, 100 nM dexamethasone, and 50 μg/ml gentamicin. The cultured cells were maintained at 37°C in a 5% CO₂ environment, as outlined in a previously published study. Adenoviruses were prepared according to the methods described in previous studies (17, 42), and administered to primary hepatocytes six hours after they were seeded.

Western blots

Western blot analyses were conducted using 50-100 μg of protein from whole cell lysates and 5-30 μg of protein from subcellular fractions, following established procedures detailed in previous publications (40, 41). The antibodies used in these analyses are listed in Supplementary Table 2.

Histology

Freshly harvested liver tissues were fixed, embedded, and sliced into 5 μm thick sections. Briefly, paraffin sections were first deparaffinized and then hydrated to develop in diaminobenzidine solution under a microscope and counterstained with hematoxylin and eosin (H&E) (17, 43) and immunostaining was performed using Ki67 (1:100) and AFP (1:100) primary monoclonal antibodies.

Total RNA extraction and real-time PCR

As described (42), total RNA was extracted using Trizol reagent (Invitrogen, USA) and was reverse-transcribed into complementary DNA with random hexamer primer (Thermo Fisher Scientific, USA). The qRT-PCR was performed in triplicates using QuantStudio 6 Flex Real-Time PCR System (Applied Biosystems, USA) according to the manufacturer’s instructions. The primers used for qRT-PCR were listed in supplementary data.

Bioinformatics analysis

All bioinformatics analysis and visualization were conducted as previously described (44, 45). All raw transcriptomic data are publicly available on the Gene Expression Omnibus (GEO; www.ncbi.nlm.nih.gov/geo) under the accession numbers GSE48452 and GSE24031, or GeneNetwork (www.genenetwork.org). Heat maps were constructed using GENE-E software (The Broad Institute, www.broadinstitute.org/cancer/software/GENE-E/). Principal component analysis and construction of corrgrams and correlation networks were performed using functions and packages in R (www.r-project.org). The depth of shading in the correlation matrices indicates the magnitude of the correlation (Pearson’s r). Positive and negative correlations within the corrgram and correlation network are represented in blue and red, respectively. Only correlations with P < 1E-05 are displayed in the correlation network.

Statistical analysis

Statistical analyses were performed using Prism (version 8.0; GraphPad Software Inc.) and R software. The significance of the differences between the two groups was determined using unpaired two-tailed Student’s t-tests. To compare multiple groups, we applied a one-way analysis of variance (ANOVA) with a post-hoc Bonferroni test. The results are presented as mean ± standard error of the mean (SEM). Statistical significance was set at P < 0.05.

Funding Statement

ACKNOWLEDGEMENTS This study was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF), funded by the Ministry of Science and ICT (NRF-2021R1A6A3A13044725 to Y.K.; 2020R1A2C2010964, 2022K2A9A1A06091879, 2023R1A2C3006220, RS-2023-00283539, and RS-2023-00261370 to D.R.). D.R. was supported by a “GIST Research Institute (GRI) IIBR” grant funded by the GIST in 2023.