ABSTRACT
-
Background/Aims
Hepatocellular carcinoma (HCC) is one of the leading causes of cancer-related mortality worldwide and is driven by metabolic reprogramming that supports tumor growth and progression. A common missense genetic variant (rs2642438, p.A165T) in mitochondrial amidoxime reducing component 1 (MTARC1), identified as protective against liver disease, has been recently associated with lower prevalence of steatosis, cirrhosis, and HCC. However, the mechanistic role of MTARC1 in HCC is unclear. Therefore, we sought to decipher the role of MTARC1 in HCC.
-
Methods
We investigated the role of MTARC1 in HCC by performing siRNA-mediated knockdown across human immortalized HCC cell lines (Hep3B2, HuH7, HepG2 and HepaRG) homozygous for the risk allele (p.A165) and by generating stable CRISPR-Cas9 knockout (KO) models. Next, we assessed the effect of MTARC1 loss on cell proliferation, migration, lipid metabolism, and fatty acid oxidation in vitro, as well as tumor aggressiveness in a subcutaneous xenograft mouse model. Additionally, we performed global proteomics in both in vitro and xenograft models.
-
Results
Transient knockdown of MTARC1 p.A165 reduced proliferation in HCC cell lines. CRISPR-Cas9-mediated stable MTARC1 p.A165 KO in Hep3B2 cells led to decreased neutral lipid intracellular accumulation, enhanced β-oxidation and reduced cell migration. An MTARC1 KO xenograft model had reduced tumor volume. Proteomic analyses of both in vitro HCC cells and xenograft tumors revealed inhibition of oncogenic pathways and activation of anti-proliferative proteins.
-
Conclusions
Downregulation of MTARC1 p.A165 inhibits lipid accumulation, dampens tumor-promoting pathways and restricts tumor growth, highlighting MTARC1 as a promising therapeutic target for HCC.
-
Keywords: MTARC1; Subcutaneous xenograft model; Hypoxia; EMT; Fatty acid degradation
Study Highlights
• Downregulation of the MTARC1 p.A165 risk allele reduces cell proliferation, intracellular neutral lipid accumulation, and migration, while enhancing β-oxidation in Hep3B2 cells.
• Transient knockdown of CPT1A, a key protein involved in β-oxidation, partially reverses the beneficial effects observed upon MTARC1 p.A165 knockout.
• MTARC1 p.A165 knockout markedly reduces tumor growth and proliferation in a subcutaneous HCC xenograft model.
• These findings identify MTARC1 as a potential therapeutic target in hepatocellular carcinoma.
Graphical Abstract
INTRODUCTION
Hepatocellular carcinoma (HCC) is the most common form of liver cancer and the third leading cause of cancerrelated mortality worldwide [
1]. The incidence of HCC continues to rise, mainly due to increasing prevalence of chronic liver diseases, such as metabolic dysfunction-associated steatotic liver disease (MASLD) and alcohol-related liver disease [
2]. Despite recent advances in its treatment regimen, including immune checkpoints and tyrosine kinase inhibitors, the overall prognosis for HCC remains poor, with a median survival of less than two years, indicating an un-met medical need [
3,
4].
Metabolic reprogramming is a classical hallmark of HCC. HCC cells adapt their metabolic state to increase proliferation, evade cell death and survive under hypoxic conditions [
5]. This adaptation involves lipid accumulation through increased synthesis and impaired export, while altered fatty acid oxidation supports energy production and survival under stress [
5,
6]. In this context, β-oxidation represents a double-edged metabolic pathway. On the one hand, enhanced β-oxidation may promote tumorigenic activity by providing energy and supporting survival [
6], but on the other, it may exert anti-tumorigenic effects by limiting cell proliferation, such as in HCC, prostate and breast cancer, underscoring a context-dependent role of lipid metabolism in tumor progression [
6]. Understanding the molecular mediators of these processes is essential for identifying novel therapeutic targets.
Mitochondrial amidoxime reducing component 1 (MTARC1, also referred to as MARC1) is a molybdenum-containing enzyme present on the outer mitochondrial membrane [
7]. Initially characterized for its role in the reductive metabolism of xenobiotics, MTARC1 has subsequently been implicated in lipid metabolism and redox homeostasis [
7,
8]. Genome-wide association studies (GWASs) have identified a common protective
MTARC1 variant (p.A165T) in which the minor allele is associated with lower hepatic triglyceride content, as well as a lower prevalence of cirrhosis and of HCC [
8-
10] suggesting a relevant role of MTARC1 in cancer biology. However, the mechanistic contribution of MTARC1 to HCC biology has not been elucidated.
To address this gap, we investigated the role of MTARC1 in HCC using in vitro and in vivo models. We found that loss of MTARC1 p.A165 inhibits HCC cell proliferation and disrupts key oncogenic signaling pathways. Further, in a subcutaneous xenograft model of MTARC1 p.A165 knockout (KO) cells in mice we found reduced tumor volume and weight compared with xenografts derived from MTARC1 control (mock) cells. Together, these findings establish that downregulation of hepatic MTARC1 p.A165 may represent a promising strategy to treat HCC.
MATERIALS AND METHODS
Details for each experiment are provided in the Supporting Information.
RESULTS
Transient downregulation of the MTARC1 p.A165 risk allele reduces cellular proliferation in HCC cell lines
To investigate the role of
MTARC1 in HCC, we transiently downregulated its expression using siRNA in four human HCC cell lines homozygous for the p.A165 risk allele, namely, Hep3B2, Huh7, HepG2, and HepaRG.
MTARC1 p.A165 knockdown efficiency was confirmed by qPCR in all cell lines, with mRNA levels lower by more than 80% at 24 hours, 48 hours and 72 hours compared with scrambled siRNA control (siSCR) (
Fig. 1, left panels;
P<0.001 for all time points).
MTARC1 p.A165 downregulation lowered the cell count by 34% in Hep3B2, 24% in Huh7 and 24% in HepG2 cells, respectively, averaged across the three time points (
Fig. 1A–
C, right panels;
P≤0.01). In HepaRG cells, proliferation was reduced by 16% at 48 hours (
P≤0.01) and at 72 hours (
P<0.001) (
Fig. 1D, right panel).
These findings demonstrate that transient downregulation of MTARC1 p.A165 exerts an anti-proliferative effect.
Stable KO of MTARC1 p.A165 reduces intracellular neutral lipid content, cell proliferation and migration in vitro
To further characterize the role of MTARC1 in HCC, we generated a stable
MTARC1 p.A165 KO cell line using the CRISPR-Cas9 method in Hep3B2 cells and tested the effect of the KO on intracellular neutral lipid content, cell proliferation and migration in vitro. By immunoblot analysis we found a marked reduction of MTARC1 protein expression in the KO cell lysates (
Fig. 2A, upper panel), and densitometric quantification normalized to calnexin (CNX) revealed a more than 90% reduction in MTARC1 protein levels (
P≤0.01).
We found that
MTARC1 p.A165 KO cells had lower intracellular neutral lipid content, as visualized by Oil-Red-O staining (
Fig. 2B;
P<0.001) compared with mock cells. Additionally, via a [³H]-palmitate oxidation assay we found higher β-oxidation activity in KO cells compared with mock cells (
P≤0.01), indicating higher fatty acid oxidation (
Fig. 2D).
When we analyzed cell proliferation, we found that
MTARC1 p.A165 KO had lower proliferation compared with mock cells at 24 hours (35% reduction,
P<0.001), 48 hours (41.0% reduction,
P≤0.01) and 72 hours (43% reduction,
P<0.001) post-seeding (
Fig. 2C).
Next, we examined cell migration using a wound healing assay. Compared with mock cells,
MTARC1 p.A165 KO showed a markedly lower wound closure at 24 hours (73% inhibition,
P<0.001), 48 hours (55% inhibition,
P≤0.01) and 72 hours (60% inhibition,
P≤0.01) (
Fig. 2E, lower panel).
These data were virtually identical in two different
MTARC1 mock and KO cell clones (
Supplementary Fig. 1).
To gain further insight into the molecular mechanisms of MTARC1 in HCC, we performed whole-cell proteomics on mock and
MTARC1 p.A165 KO Hep3B2 cells. We detected a total of 9,193 proteins, of which 238 were upregulated and 190 were downregulated (FDR adjusted, |log₂FC|≥±0.58) in
MTARC1 p.A165 KO cells compared with mock cells. To decipher the biological significance of these differentially expressed proteins, we next performed pathway enrichment analysis and found that
MTARC1 p.A165 KO results in a profound downregulation of epithelial-mesenchymal transition (EMT), hypoxia, epidermal growth factor receptor 1 (EGFR1) signaling and TGF-β regulation of extracellular matrix, as well as upregulation of fatty acid metabolism and p53 signaling (
Fig. 3). After analysis of the differentially expressed proteins, five proteins were randomly selected whose expression levels were unchanged between Hep3B2 mock and Hep3B2
MTARC1 p.A165 KO cells (
Supplementary Fig. 2A).
EMT enables cancer cells to acquire migratory and invasive properties [
11]. We found that levels of CTHRC1, THBS1, CDH6 and DKK1 were lower in
MTARC1 p.A165 KO cells compared with mock cells, resulting in impaired mesenchymal transition. This suggests that the loss of
MTARC1 p.A165 hinders the epithelial-to-mesenchymal plasticity of HCC cells (
Fig. 3A).
Hypoxia is a major hallmark of HCC, driving tumor aggressiveness, metabolic reprogramming and extracellular matrix remodeling [
12]. We found that the expression levels of proteins, such as BGN, NDRG1 and ALDOA, were markedly lower in the
MTARC1 p.A165 KO cells compared with mock cells (
Fig. 3B).
The EGFR pathway is a critical driver of HCC progression [
13] and
MTARC1 p.A165 KO led to lower levels of various EGFR-associated regulators, including ITGA2, TAGLN and TGFB2, compared with mock cells (
Fig. 3C).
Fatty acid degradation is a cellular metabolic pathway that involves the breakdown of fatty acids to generate metabolic energy [
6]. By proteomic profiling we found higher levels of proteins involved in fatty acid degradation, including HMGCS1, GPX1, FMO1 and CPT1A, in
MTARC1 p.A165 KO cells compared with mock cells (
Fig. 3D).
The TGF-β pathway is known to promote cancer progression, metastasis and extracellular matrix remodeling [
14]. We found that the levels of TGFB2 and RRAS were lower, whereas SKI was higher, in
MTARC1 p.A165 KO cells compared with mock cells, indicating suppression of the TGF-β pathway in the KO cells (
Fig. 3E).
Although Hep3B2 cells are null for
TP53 (the gene encoding p53) [
15], pathway enrichment analyses highlighted higher p53-related signaling, likely reflecting a p53-independent response converging on similar downstream targets. Notably, the tumor suppressor CDKN2A was higher in
MTARC1 p.A165 KO cells compared with mock cells, consistent with the induction of an anti-proliferative pathway (
Fig. 3F).
Our findings demonstrate that loss of MTARC1 p.A165 in Hep3B2 cells leads to suppression of EMT, hypoxia adaptation and EGFR signaling while promoting fatty acid degradation.
We also performed whole-cell proteomics on
MTARC1 p.A165 KO and
MTARC1 mock HepG2 cells. We detected a total of 7,000 proteins, of which 299 were upregulated and 296 were downregulated in the
MTARC1 p.A165 KO cells compared wi th the mock cel ls (FDR adjusted, |log₂FC|≥±0.58). After analysis of the differentially expressed proteins, five proteins were randomly selected whose expression levels were unchanged between HepG2 mock and HepG2
MTARC1 p.A165 KO cells (
Supplementary Fig. 2C). Heatmaps of differentially expressed proteins between mock and
MTARC1 p.A165 KO are shown for each pathway (
Supplementary Fig. 3). Loss of
MTARC1 p.A165 in HepG2 cells led to suppression of the mTORC1 pathway, while promoting fatty acid metabolism. Certain EMT markers were increased; however, this should be interpreted cautiously, as TGF-β–associated proteins were decreased (
Supplementary Fig. 3), indicating that the observed changes may represent a feedback response rather than bona fide EMT activation.
To assess the causal relationship between
MTARC1 p.A165 KO and β-oxidation on proliferation, we transiently knocked down
CPT1A in Hep3B2 mock and
MTARC1 p.A165 KO cells. In mock cells, si
CPT1A induced upregulation of
MTARC1 mRNA, suggesting a feedback response to lower fatty acid availability in the mitochondria (
P<0.01,
Fig. 4A). Consistent with these results,
CPT1A knockdown lowered β-oxidation by 30% (
P<0.001) in mock cells and by 50% (
P<0.001) in
MTARC1 p.A165 KO cells compared with their respective siSCR-transfected cells (
Fig. 4B).
As a consequence of lower β-oxidation,
CPT1A knockdown was associated with higher neutral lipid content in both backgrounds. si
CPT1A knockdown resulted in 59% (
P<0.001) higher neutral lipid content compared with mock siSCR.
MTARC1 p.A165 KO siSCR exhibited 42% (
P<0.001) lower neutral lipid content than mock siSCR. The double perturbation (
MTARC1 p.A165 KO+si
CPT1A) showed 41% (
P<0.001) higher neutral lipid content relative to
MTARC1 p.A165 KO siSCR but remained 18% (
P<0.001) lower than mock siSCR, indicating a partial rescue of the KO phenotype (
Fig. 4C).
Similarly,
CPT1A knockdown enhanced cell proliferation. si
CPT1A knockdown resulted in a 22% (
P<0.01) higher number of cells than mock siSCR.
MTARC1 p.A165 KO siSCR exhibited 49% (
P<0.01) lower cell numbers than mock siSCR cells. The double perturbation (
MTARC1 p.A165 KO+si
CPT1A) showed 65% (
P<0.01) higher cell numbers relative to
MTARC1 p.A165 KO siSCR but remained 16% (
P<0.01) lower than mock siSCR, again indicating a partial rescue of the KO phenotype (
Fig. 4D).
Together, these findings demonstrate that CPT1A knockdown reduces β-oxidation and partially reverses, the beneficial effect observed in MTARC1 p.A165 KO cells, demonstrating a causal role for β-oxidation in this phenotype.
KO of MTARC1 p.A165 reduces tumor growth in a subcutaneous xenograft mouse model
To determine whether
MTARC1 p.A165 reduction influences tumorigenicity in vivo,
MTARC1 p.A165 KO and Hep3B2 mock cells were injected subcutaneously into the flanks of 5-week-old BALB/c nude mice (
Supplementary Fig. 4). Tumor growth was monitored after injection over a period of 48 days, with tumor volumes measured every 3 days (
Supplementary Fig. 4).
Tumors derived from mock cells grew more rapidly, whereas those from
MTARC1 p.A165 KO cells showed markedly lower growth (
Fig. 5A, top and middle panels). At the end point, gross examination revealed smaller tumors in the
MTARC1 p.A165 KO group, which also appeared less red and less vascularized compared with the mock tumors (
Fig. 5A, bottom panel).
Longitudinal measurements of tumor volume revealed a sustained difference between groups throughout the study, with a lower volume observed from day 36 after injection onwards, until the end of the experiment (day 48). By day 48, tumor volume in the
MTARC1 p.A165 KO group was 45% lower compared to the mock controls, indicating a reduction in tumor burden among the KO group (
Fig. 5B;
P<0.001). Consistent with these results, tumor weight was also 38% lower in the
MTARC1 p.A165 KO group compared to the mock group (
Fig. 4C;
P<0.05). Additionally, we found that the expression of Ki-67, a marker of cell proliferation, was 42% lower in the
MTARC1 p.A165 KO tumors compared with mock tumors, as visualized by immunofluorescence (
Fig. 5C;
P<0.05).
These results demonstrate that MTARC1 p.A165 KO inhibits tumor growth in a subcutaneous xenograft mouse model.
In vivo tumor proteomics reveals that MTARC1 p.A165 KO downregulates oncogenic pathways in HCC
To investigate the molecular consequences of MTARC1 p.A165 KO in Hep3B2 xenograft tumors, we performed whole-cell proteomic analysis comparing protein expression profiles between Hep3B2 mock and Hep3B2 MTARC1 p.A165 KO xenografts in nude mice. We detected 8,381 proteins, of which 375 were upregulated and 228 were downregulated in Hep3B2 MTARC1 p.A165 KO xenografts compared with the mock control group (FDR adjusted, |log₂FC| ≥±0.58).
In Hep3B2 2D proteomics (mock vs. MTARC1 p.A165 KO), we detected 428 differentially expressed proteins, whereas in xenografted tumors in nude mice (mock vs. MTARC1 p.A165 KO), we detected 603 differentially expressed proteins. Notably, 111 proteins were commonly differentially expressed across both models. Among these shared proteins, 102 proteins exhibited concordant expression changes while 9 displayed discordant regulation. The inverse regulation may reflect adaptation of tumor cells to the in vivo microenvironment and contributions from hosttumor interactions within the xenograft setting.
The overall expression patterns demonstrated a strong positive correlation in protein expression between Hep3B2 mock and
MTARC1 p.A165 KO cells and their corresponding xenografted tumors, with a Log
2FC concordance yielding a Pearson correlation coefficient of r=0.76 (
P=5.7×10⁻²²) (
Fig. 6A).
Pathway enrichment analysis revealed changes in erythrocytic proteins, hypoxia signaling, myc repressed pathway, G2-M checkpoint and p53 signaling. Heatmaps of differentially expressed proteins between mock tumors and KO tumors are shown in each pathway (
Fig. 6). After analysis of the differentially expressed proteins, five proteins were randomly selected whose expression levels did not differ between Hep3B2 mock and Hep3B2
MTARC1 p.A165 KO xenograft tumors (
Supplementary Fig. 2B).
We observed lower expression of several erythrocytic markers, including HBA1, HBB and SPTB, in
MTARC1 p.A165 KO tumors compared with mock tumors (
Fig. 6B).
Hypoxia pathway analysis revealed lower expression of hypoxia-related proteins, including PGAM2, EFNA1 and GALK1, in
MTARC1 p.A165 KO tumors compared with mock tumors. This indicates that loss of
MTARC1 reduces hypoxic stress within tumors. The combined effect of reduced erythrocytic and hypoxia markers likely contributes to the pale appearance of
MTARC1 p.A165 KO tumors, reflecting lower vascularization compared with mock tumors (
Fig. 6C).
Despite the fact that Hep3B2 cells are
TP53 null [
15],
MTARC1 KO exhibited higher levels of p53-associated proteins, including PRKAB1, HEXIM1, and ZFP36L1. These proteins act as tumor suppressors in various cancers, and their increased expression suggests that
MTARC1 loss can restrict tumor growth even in the absence of p53 (
Fig. 6D).
We also found increases in several G2/M checkpoints proteins in
MTARC1 p.A165 KO tumors compared with mock tumors, which can lead to cell cycle arrest in response to stress. Notably, CDKN1B was higher in the KO tumors, potentially blocking cyclin–CDK activity and ceasing their progression into mitosis (
Fig. 6E).
Notably, in
MTARC1 p.A165 KO tumors we observed an increase in the proteins that inhibit the Myc pathway, such as CDKN1B, FOXO3, and SMAD2, suggesting a role for
MTARC1 in this pathway (
Fig. 6F).
DISCUSSION
The main finding of this study is that MARC1 p.A165 ablation reduces HCC aggressiveness by metabolic reprogramming.
We began by assessing proliferation in human HCC cell lines homozygous for the
MTARC1 risk allele (p.A165). Transient siRNA-mediated knockdown of
MTARC1 p.A165 led to lower proliferation in four HCC cell lines (Hep3B2, HuH7, HepG2 and HepaRG). These results were validated in CRISPR-Cas9-mediated stable
MTARC1 p.A165 KO Hep3B2 cells, in which we found a larger effect size compared to the transient knockdown approach, likely reflecting the stable nature of CRISPR-Cas9-mediated KO. In addition to the effects on proliferation and migration, Hep3B2
MTARC1 p.A165 KO displayed reduced intracellular neutral lipid content and increased fatty acid β-oxidation compared to mock KO control cells, consistent with our previous work with human primary hepatocytes [
7]. We should note that transient knockdown of
CPT1A partially reversed the beneficial effect of the
MTARC1 p.A165 KO, indicating that the increase in β-oxidation induced by the KO underlies this beneficial effect.
To further explore the molecular mechanisms underlying the role of MTARC1 in HCC, we performed proteomic profiling in vitro using Hep3B2 mock and
MTARC1 KO cells. Loss of
MTARC1 p.A165 led to suppression of key oncogenic pathways, including EMT, EGFR signaling, TGF-β-mediated extracellular matrix remodeling and hypoxia responses. Downregulation of EMT markers (CTHRC1, THBS1) and hypoxia-related proteins (NDRG1, ALDOA) indicates impaired cellular plasticity and adaptation to hypoxic stress, both of which are critical for HCC aggressiveness [
16-
19]. Interestingly, we found that
MTARC1 p.A165 KO increased levels of tumor suppressor-associated proteins (CDKN2A, ZBTB16), suggesting activation of anti-proliferative pathways, even in Hep3B2 cells, which are p53 deficient. All together, these findings indicate that MTARC1 contributes to tumor progression by orchestrating oncogenic signaling cascades. In vivo, we found that
MTARC1 p.A165 KO had profound effects on tumor growth. Xenograft tumors derived from
MTARC1 p.A165 KO Hep3B2 cells exhibited slower growth, reduced vascularization and smaller size compared with Hep3B2 mock cells. We also found a 42% reduction in Ki-67 expression, a well-known proliferation marker upregulated in HCC [
20], in
MTARC1 p.A165 KO tumors. Consistent with these observations, erythrocyte markers and hypoxia-related proteins were lower in
MTARC1 p.A165 KO tumors, supporting the notion that MTARC1 influences both tumor metabolism and its microenvironment. Interestingly, by proteomics, we found 30% more differentially expressed proteins in xenograft tumors compared with 2D cell culture, highlighting the complexity of host-tumor interactions and involvement of additional signaling pathways in vivo.
The missense variant p.A165T (rs2642438) is associated with a small protective effect on HCC risk with an odds ratio (OR) of ~0.83, which is attenuated when HCC cases are compared specifically with cirrhosis without HCC (OR 0.91; 95% CI 0.84–1.00), indicating that the observed effect of the variant is shared with progression to cirrhosis rather than tumor development per se [
10].
In contrast, a recent genome-wide meta-analysis identified a separate
MTARC1 signal, an intronic variant (rs2642442), which reaches genome-wide significance and is associated with a modest increase in risk (β ~0.17, corresponding to an OR of ~1.2) [
21]. The genetic effect correlates strongly with genetic determinants of hepatic steatosis (r²=0.75) and cirrhosis (r²=0.69). These data support a model in which this variant influences HCC risk primarily through hepatic lipid accumulation and progression of chronic liver disease. Notably, this intronic variant is compatible with a gain-of-function and is associated with an increased risk of MASLD (OR of 1.15) [
22].
Overall, both studies are consistent with typical common germline variants, supporting a role for MTARC1 as a modifier of risk rather than a strong oncogenic driver. Also, our study focuses on germline variation and metabolic regulation rather than somatic mutagenesis.
The modest effect size associated with the
MTARC1 p.A165 variant in humans contrasts with the much larger effects observed following
MTARC1 KO in cancer cell lines. There are at least three plausible explanations for the observed discrepancy: (1) in our hands, the
MTARC1 p.A165 variant reduces protein levels by approximately 50% [
23,
24], whereas CRISPR-mediated KO virtually eliminates the protein, resulting in a stronger perturbation of the system; (2) people born with the
MTARC1 protective genetic variant may develop compensatory mechanisms as a consequence of lifelong reduction in
MTARC1 function, whereas acute ablation of
MTARC1 on a wild-type background would not allow for such adaptation and may therefore produce a larger effect size. An illustrative example is congenital analbuminemia, a largely benign condition characterized by oedematous swelling, in contrast to acquired severe hypoalbuminemia, which is a life-threatening condition [
25]; and (3) in carriers of the genetic variant,
MTARC1 loss of function occurs at the whole-body level (including the immune system), whereas our experiments are limited to hepatocytes.
Our study has some limitations. All experiments were conducted in HCC cells carrying the
MTARC1 risk variant (p.A165) and not in cells expressing the protective p.T165 allele. We have previously demonstrated that the protective allele leads to a loss of MTARC1 function through proteasomal degradation [
23]. Based on these findings, silencing the protective allele would be expected to have little or no impact on cell proliferation and metabolic reprogramming. Moreover, our in vivo studies were performed in nude mice, which do not reflect the interaction with the immune system. Further studies in orthotopic models or humanized mouse systems will be valuable to confirm these results. Finally, our proteomic analyses revealed suppression of various oncogenic pathways. The causal relationship between
MTARC1 ablation and these signaling cascades remains to be defined.
In conclusion, our results demonstrate that MTARC1 p.A165 ablation results in metabolic reprogramming attenuating HCC aggressiveness by downregulating oncogenic and upregulating tumor suppressor signals. These data suggest that downregulation of MTARC1 p.A165 may be a potential strategy to treat HCC.
FOOTNOTES
-
Authors’ contributions
LK contributed to study design, data collection, data analysis, data interpretation, and manuscript drafting; JZ and FGM were involved in data collection and visualization; TD was involved in data interpretation and manuscript drafting; XG and BA were involved in data collection and visualization; OJ was involved in data interpretation; PI and EB were involved in visualization and manuscript drafting; MM was involved in data collection and visualization; RMM contributed to study design, supervision, discussion, and manuscript writing; SR led the study with conceptualization, funding acquisition, supervision, data interpretation, methodology, and manuscript writing. All authors read and approved the final version for publication.
-
Acknowledgements
Proteomic analysis was performed at the Proteomics Core Facility, Sahlgrenska academy, Gothenburg University, with financial support from SciLifeLab and BioMS. We would like to thank Annika Thorsell and Evelin Berger for their contributions to the proteomic experiments and writing the methods.
LK and BA were supported by a grant from the K. and A. Wallenberg Foundation via the Wallenberg Centre for Molecular and Translational Medicine, Gothenburg.
S.R. was supported by the Swedish Cancerfonden (22 2270 Pj), the Swedish Research Council (Vetenskapsradet (VR), 2023-02079), the Swedish state under the Agreement between the Swedish government and the county councils (the ALF agreement, ALFGBG-965360), the Swedish Heart Lung Foundation (20220334), the Novonordisk Distinguished Investigator Grant - Endocrinology and Metabolism (NNF23OC0082114), the Novonordisk Project grants in Endocrinology and Metabolism (NNF24OC0091535), and a Novo Nordisk donation to the Karolinska Institutet in connection with the professor appointment.
-
Conflicts of Interest
In the last 5 years S.R. received research grants from Novonordisk and AstraZeneca for basic science research on steatotic liver disease, and has been consulting for AstraZeneca, GSK, Celgene Corporation, Ribocure AB¸ Madrigal, Ultragenyx, Amgen, Sanofi, Wave Life Sciences, Lipigon, Novartis, Profluent, Aina, Echosense and Chiesi in the last 5 years; declares equity from Heptabio; and is inventor on a Patent with title “Method for treating fatty liver disease”, on PSD3, US application number 17,480266 filed on 21st September 2021. All other authors have none to declare.
SUPPLEMENTARY MATERIAL
Supplementary material is available at Clinical and Molecular Hepatology website (
http://www.e-cmh.org).
Supplementary Figure 1.
Stable knockout of MTARC1 p.A165 in Hep3B2 cells reduces proliferation, lipid content and migration. (A, B) Immunoblotting was performed to assess MTARC1 KO efficiency. Cell lysates were run in SDS-PAGE and probed against MTARC1 and CNX. Band intensities were normalized to CNX and quantified using ImageLab software. Statistical analysis was performed using an unpaired Student’s t-test. Hep3B2 mock or MTARC1 p.A165 KO cells were seeded and fixed after 48 hours in 4% PFA, followed by Oil Red O (ORO) and DAPI staining. The ORO stained area was quantified using ImageJ. Objective: 20×, DAPI: Blue, ORO: Red. Scale bar=100 µm. P-values were calculated by Mann–Whitney non-parametric test. Hep3B2 mock or MTARC1 p.A165 KO cells were collected at 24 hours, 48 hours, and 72 hours, and cell counts were performed. P-values were calculated using an unpaired Student’s t-test. A scratch wound assay was performed in Hep3B2 mock cells or MTARC1 p.A165 KO cells. After 24 hours of seeding, a scratch was made using pipette tips and images were captured at 24 hours, 48 hours, and 72 hours. Objective: 20×. Scale bar=100 µm. Data are normalized to the mock control at each respective time point (set as 100%). Wound closure was calculated in percentage. P-values were calculated using an unpaired Student’s t-test. Data are shown as mean±SD for three independent experiments. Statistical analysis was performed using GraphPad Prism version 9 (San Diego, CA, USA). **P<0.01, ***P<0.001. CNX, calnexin; KO, knockout; MTARC1, mitochondrial amidoxime-reducing component 1; PFA, Paraformaldehyde.
Supplementary Figure 2.
Control proteins in various whole cell proteomics data whose expression did not change between groups. (A) Hep3B2 mock and MTARC1 p.A165 KO cells, (B) Hep3B2 mock and MTARC1 p.A165 KO cells xenografted in nude mice and (C) HepG2 mock and MTARC1 p.A165 KO cells. KO, knockout; MTARC1, mitochondrial amidoxime-reducing component 1.
Supplementary Figure 3.
Comprehensive proteomics analysis of MTARC1 p.A165 in HepG2 cells. Whole-cell proteomics was performed on three technical replicates of HepG2 mock and MTARC1 KO. Samples were processed and proteins were quantified using MS/MS, liquid chromatography-tandem mass spectrometry. The protein abundance ratios were log2-transformed, and statistical significance was assessed using Benjamini-Hochberg test, with an adjusted P-value threshold of <0.05. A |log2FC|≥±0.58 cutoff was used to define upregulation or downregulation. The colour intensity in the heatmap represents the Log2FC values. Pathway enrichment analysis was performed using Enrichr, and the following pathways were enriched in MTARC1 KO cells: (A) mTORC1 pathway, (B) fatty acid metabolism, (C) epithelial-mesenchymal transition, (D) apoptosis, (E) p53 pathway, and (F) TGF-β regulation of extracellular matrix. KO, knockout; MS/MS, tandem mass spectrometry; MTARC1, mitochondrial amidoxime-reducing component 1; TGF-β, transforming growth factor-beta.
Supplementary Figure 4.
Hep3B2 mock and MTARC1 p.A165 KO cells were injected subcutaneously into the upper right flanks (mock)
and upper left flanks (KO) of 5-week-old male BALB/c nude mice. KO, knockout; MTARC1, mitochondrial amidoxime-reducing component 1.
Figure 1.
MTARC1 p.A165 risk allele downregulation reduces cell proliferation. (A–D) mRNA expression levels of MTARC1 (left panels) and cell count per mL (right panels) assessed in (A) Hep3B2, (B) HuH7, (C) HepG2 and (D) HepaRG cells after siRNA mediated knockdown of MTARC1. Cells were transfected with either siSCR (control) or siMARC1, and both mRNA expression and cell count was measured at 24 hours, 48 hours, and 72 hours post-transfection. P-values were calculated using unpaired Student’s t-test. Data are shown as mean±SD for three independent experiments. Statistical analysis was performed using GraphPad Prism version 9 (San Diego, CA, USA). **P<0.01, ***P<0.001. MTARC1, mitochondrial amidoxime-reducing component 1; siSCR, scrambled siRNA.
Figure 2.Stable knockout (KO) of MTARC1 p.A165 in Hep3B2 cells reduces their proliferation, lipid content and migration. (A) Immunoblotting was performed to assess MTARC1 KO efficiency. Cell lysates were run in SDS-PAGE and probed against MTARC1 and CNX. Band intensities were normalized to CNX and quantified using ImageLab software. Statistical analysis was performed using unpaired Student’s t-test. (B) Hep3B2 mock or MTARC1 p.A165 KO cells were seeded and fixed after 48 hours in 4% PFA, followed by Oil Red O (ORO) and DAPI staining. The ORO stained area was quantified using ImageJ. Objective: 20×, DAPI: Blue, ORO: Red. Scale bar=100 µm. P-values were calculated by Mann–Whitney non-parametric test. (C) Hep3B2 mock or MTARC1 p.A165 KO cells were collected at 24 hours, 48 hours, and 72 hours, and cell counts were performed. P-values were calculated using an unpaired Student’s t-test. (D) Cells were incubated with 8.3 μCi/mL [3H] palmitate and 110 µM palmitate for 2 hours. Media were collected, and radioactivity was measured using a scintillation counter. Results were expressed as disintegrations per minute (DPM) and normalized to cell number. P-values were calculated by Mann–Whitney non-parametric test. (E) A scratch wound assay was performed in Hep3B2 mock cells and MTARC1 KO cells. After 24 hours of seeding, a scratch was made using pipette tips and images were captured at 24 hours, 48 hours, and 72 hours. Objective: 20×. Scale bar=100 µm. Data are normalized to the mock control at each respective time point (set as 100%). Wound closure was calculated in percentage. P-values were calculated using an unpaired Student’s t-test. Data are shown as mean±SD for three independent experiments. Statistical analysis was performed using GraphPad Prism version 9 (San Diego, CA, USA). **P<0.01, ***P<0.001. CNX, calnexin; MTARC1, mitochondrial amidoxime-reducing component 1; PFA, Paraformaldehyde.
Figure 3.Proteomics analysis of MTARC1 p.A165 KO in Hep3B2 cells. Whole-cell proteomics was performed on six technical replicates of Hep3B2 mock and MTARC1 KO cells. Samples were processed and proteins were quantified using MS/MS, liquid chromatographytandem mass spectrometry. The protein abundance ratios were log2-transformed, and statistical significance was assessed using Benjamini-Hochberg test, with an adjusted P-value threshold of <0.05. A |log₂FC|≥±0.58 cutoff was used to define upregulation or downregulation. The colour intensity in the heatmap represents the Log2FC values. Pathway enrichment analysis was performed using Enrichr, and the following pathways were enriched in MTARC1 KO cells: (A) EMT, (B) Hypoxia, (C) EGFR1, (D) Fatty acid metabolism, (E) TGF-β regulation of extracellular matrix and (F) p53 pathway. EGFR1, epidermal growth factor receptor 1; EMT, epithelial–mesenchymal transition; KO, knockout; MS/MS, tandem mass spectrometry; MTARC1, mitochondrial amidoxime-reducing component 1; TGF-β, transforming growth factor-beta.
Figure 4.
CPT1A knockdown reduces β-oxidation and partially rescues the phenotype observed in MTARC1 p.A165 knockout (KO) cells. Hep3B2 mock or MTARC1 p.A165 KO were seeded and, after 24 hours, transfected with scrambled siRNA (siSCR) or CPT1A targeting siRNA (siCPT1A). Cells were subsequently collected for the indicated analyses. (A) Relative mRNA expression of MTARC1 (left) and CPT1A (right) determined by qRT-PCR in mock and MTARC1 p.A165 KO following siSCR or siCPT1a transfection. Data are expressed as relative units (RU). (B) β-oxidation measured by incubation with 8.3 μCi/mL [³H]-palmitate and 110 µM palmitate for 2 hours. Media were collected and the radioactivity was quantified using a scintillation counter. Results are expressed as disintegrations per minute (DPM) normalized to cell number. (C) Cells were fixed 48 hours post-transfection with 4% PFA and stained with Oil Red O (ORO) and DAPI. Lipid accumulation was quantified as ORO-positive area per cell using ImageJ. Representative images are shown (objective 20×; DAPI, blue; ORO, red). Scale bar=100 µm. (D) Cell counts were performed 48 hours post-transfection. Data are shown as mean±SD from six independent experiments. Statistical significance was assessed using the Mann–Whitney non-parametric test and analyzed with GraphPad Prism version 9 (San Diego, CA, USA). **P<0.01, and ***P<0.001. MTARC1, mitochondrial amidoxime-reducing component 1; PFA, Paraformaldehyde.
Figure 5.
MTARC1 p.A165 knockout (KO) reduces tumor volume and weight in a subcutaneous xenograft mouse model. (A) Representative images of mice (top) and xenograft tumors (bottom) at the day of termination. Scale bar for mouse: 1.8 mm and scale bar for tumor: 0.38 mm. (B) Tumor volume was measured every 3 days. Statistical significance was determined using a linear mixed model. Linear mixed-effects model was fitted using R. (C) Tumor weight was measured on the day of termination. The P-value was calculated by Mann–Whitney non-parametric test. (D) Representative tumor sections were processed for immunofluorescence with an anti-Ki-67 antibody, and the nuclei were stained with DAPI (blue). Objective: 20×, DAPI: Blue, Ki-67: Green. Scale bar=100 µm. P-values were calculated by Mann–Whitney non-parametric test. Statistical analysis was performed using GraphPad Prism version 9 (San Diego, CA, USA). *P<0.05, **P<0.01, and ***P<0.001. MTARC1, mitochondrial amidoxime-reducing component 1.
Figure 6.Proteomic profiling of xenografted tumors reveals downregulation of oncogenic pathways in MTARC1 p.A165 knockout (KO) tumors. Whole-cell proteomics was performed on 8 Hep3B2 mock and 7 MTARC1 p.A165 KO xenografted tumors. Samples were processed and proteins were quantified using MS/MS, liquid chromatography-tandem mass spectrometry. The protein abundance ratios were log2-transformed, and statistical significance was assessed using Benjamini-Hochberg test, with an adjusted P-value threshold of <0.05. A |log₂FC|≥±0.58 cutoff was used to define upregulation or downregulation. (A) Correlation of log2 FC in protein expression between Hep3B2 mock and MTARC1 KO cells and their corresponding xenografted tumors in nude mice. Red colour indicates upregulated proteins while blue colour indicates downregulated proteins in both cell lines and tumors. The correlation highlights the similarities and differences in protein expression profiles between in vitro and in vivo settings. Pearson’s correlation coefficient was calculated in R. The colour intensity in the heatmap represents the Log2 FC values. Pathway enrichment analysis was performed using Enrichr, and the following pathways were enriched in MTARC1 KO tumors: (B) Erythrocytic proteins, (C) Hypoxia, (D) p53 pathway, (E) G2-M checkpoint, and (F) Myc repressed pathway. MTARC1, mitochondrial amidoxime-reducing component 1; MS/MS, tandem mass spectrometry.
Abbreviations
cyclin-dependent kinase inhibitor 1B
cyclin-dependent kinase inhibitor 2A
collision-induced dissociation (MS fragmentation)
carnitine palmitoyltransferase 1A
collagen triple helix repeat containing 1
Dickkopf-related protein 1
epidermal growth factor receptor
epithelial-mesenchymal transition
field asymmetric ion mobility spectrometry
genome-wide association studies
higher-energy collision dissociation (MS fragmentation)
human embryonic kidney 293T cells
4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid
hepatocyte-related progenitor-derived RG cells
hexamethylene bis-acetamide inducible 1
liquid chromatography coupled to triple-stage mass spectrometry
mitochondrial amidoxime reducing component 1
metabolically dysfunction associated steatotic liver disease
N-myc down regulated Gene1
phosphate-buffered saline
phosphoglycerate mutase 2
protein kinase AMP-activated non-catalytic subunit beta 1
mothers against decapentaplegic homolog 2
transforming growth factor beta
tandem mass tag (isobaric labeling for proteomics)
REFERENCES
- 1. Mauro E, de Castro T, Zeitlhoefler M, Sung MW, Villanueva A, Mazzaferro V, et al. Hepatocellular carcinoma: epidemiology, diagnosis and treatment. JHEP Rep 2025;7:101571.
- 2. Singh SP, Madke T, Chand P. Global epidemiology of hepatocellular carcinoma. J Clin Exp Hepatol 2025;15:102446.
- 3. Melendez-Torres J, Singal AG. Early detection of hepatocellular carcinoma: roadmap for improvement. Expert Rev Anticancer Ther 2022;22:621-632.
- 4. Zheng S, Chan SW, Liu F, Liu J, Chow PKH, Toh HC, et al. Hepatocellular carcinoma: current drug therapeutic status, advances and challenges. Cancers (Basel) 2024;16:1582.
- 5. Wang H, Lun Y, Xu D, Jiang H, Yan Y, Yang X, et al. Research progress and therapeutic strategies in hepatocellular carcinoma metabolic reprogramming. J Adv Res 2025 Aug 22;doi: 10.1016/j.jare.2025.08.023.
- 6. Ma J, Wang S, Zhang P, Zheng S, Li X, Li J, et al. Emerging roles for fatty acid oxidation in cancer. Genes Dis 2025;12:101491.
- 7. Ciociola E, Dutta T, Sasidharan K, Kovooru L, Noto FR, Pennisi G, et al. Downregulation of the MARC1 p.A165 risk allele reduces hepatocyte lipid content by increasing beta-oxidation. Clin Mol Hepatol 2025;31:445-459.
- 8. Mancina RM, Valenti L, Romeo S. Human genetics of steatotic liver disease: insights into insulin resistance and lipid metabolism. Nat Metab 2025;7:2199-2211.
- 9. Emdin CA, Haas ME, Khera AV, Aragam K, Chaffin M, Klarin D, et al. A missense variant in Mitochondrial Amidoxime Reducing Component 1 gene and protection against liver disease. PLoS Genet 2020;16:e1008629.
- 10. Innes H, Nischalke HD, Guha IN, Weiss KH, Irving W, Gotthardt D, et al. The rs429358 locus in apolipoprotein E is associated with hepatocellular carcinoma in patients with cirrhosis. Hepatol Commun 2022;6:1213-1226.
- 11. Ribatti D, Tamma R, Annese T. Epithelial-mesenchymal transition in cancer: a historical overview. Transl Oncol 2020;13:100773.
- 12. Bao MH, Wong CC. Hypoxia, metabolic reprogramming, and drug resistance in liver cancer. Cells 2021;10:1715.
- 13. Al-Awadhi SSA, Patil P, Shetty P, Shetty PK, Shetty RA, Shetty VV, et al. Potential role of epidermal growth factor receptors (EGFR) signaling in the pathogenesis and management of hepatocellular carcinoma. Bioimpacts 2025;15:30905.
- 14. Cecerska-Heryć E, Jerzyk A, Goszka M, Polikowska A, Rachwalska J, Serwin N, et al. TGF-β signaling in cancer: mechanisms of progression and therapeutic targets. Int J Mol Sci 2025;26:7326.
- 15. Gomes AR, Abrantes AM, Brito AF, Laranjo M, Casalta-Lopes JE, Gonçalves AC, et al. Influence of P53 on the radiotherapy response of hepatocellular carcinoma. Clin Mol Hepatol 2015;21:257-267.
- 16. Liu YJ, Du J, Li J, Tan XP, Zhang Q. CTHRC1, a novel gene with multiple functions in physiology, disease and solid tumors (Review). Oncol Lett 2023;25:266.
- 17. Wang X, Sun R, Che N, Zhang D, Li Y, Zhao N, et al. Overexpression of NDRG1 leads to poor prognosis in hepatocellular carcinoma through mediating immune infiltration and EMT. Dig Liver Dis 2024;56:1382-1399.
- 18. Snaebjornsson MT, Poeller P, Komkova D, Röhrig F, Schlicker L, Winkelkotte AM, et al. Targeting aldolase A in hepatocellular carcinoma leads to imbalanced glycolysis and energy stress due to uncontrolled FBP accumulation. Nat Metab 2025;7:348-366.
- 19. Fu PY, Hu B, Ma XL, Tang WG, Yang ZF, Sun HX, et al. Far upstream element-binding protein 1 facilitates hepatocellular carcinoma invasion and metastasis. Carcinogenesis 2020;41:950-960.
- 20. Ramos-Santillan V, Oshi M, Nelson E, Endo I, Takabe K. High Ki67 gene expression is associated with aggressive phenotype in hepatocellular carcinoma. World J Oncol 2024;15:257-267.
- 21. Ghouse J, Gellert-Kristensen H, O’Rourke CJ, Seidelin AS, Thorleifsson G, Sveinbjörnsson G, et al. Genome-wide meta-analysis identifies nine loci associated with higher risk of hepatocellular carcinoma development. JHEP Rep 2025;7:101485.
- 22. Fairfield CJ, Drake TM, Pius R, Bretherick AD, Campbell A, Clark DW, et al. Genome-wide association study of NAFLD using electronic health records. Hepatol Commun 2022;6:297-308.
- 23. Dutta T, Sasidharan K, Ciociola E, Pennisi G, Noto FR, Kovooru L, et al. Mitochondrial amidoxime-reducing component 1 p.Ala165Thr increases protein degradation mediated by the proteasome. Liver Int 2024;44:1219-1232.
- 24. Caridi G, Dagnino M, Erdeve O, Di Duca M, Yildiz D, Alan S, et al. Congenital analbuminemia caused by a novel aberrant splicing in the albumin gene. Biochem Med (Zagreb) 2014;24:151-158.
- 25. Campagna F, Fioretti F, Burattin M, Romeo S, Sentinelli F, Bifolco M, et al. Congenital analbuminemia attributable to compound heterozygosity for novel mutations in the albumin gene. Clin Chem 2005;51:1256-1258.
