This study started with a project involving Sirt1, a major research focus in our lab. Sirt1 is a metabolic sensor known to be activated by calorie restriction, and methionine restriction has been shown to mimic calorie restriction to improve metabolic outcomes. We hypothesized that Sirt1 may be involved in methionine restriction-mediated health benefits in the liver, where half of dietary methionine is metabolized.
To test this possibility, we took advantage of existing human liver cancer cell lines and treated them with methionine/cystine-depleted medium (MCR). We were surprised to find that different liver cancer cell lines had drastically different responses to MCR, with some lines were more resistant to MCR-induced cell death than other lines.
To better understand the molecular difference between the sensitive and resistant lines, we performed a bioinformatic clustering analysis of a public cancer cell line database that provides mRNA profiles for multiple liver cancer cell lines. We assessed the expression patterns of genes that might be involved in the sensitivity to MCR. Since these cell lines are derived from liver tumors, we evaluated the epithelial-mesenchymal transition (EMT) markers indicative of tumor progression and drug resistance, and liver-specific markers including HNF4α, the master regulator of liver functions and known to suppress EMT. Key enzymes for methionine metabolism, broadly sulfur amino acid (SAA) metabolism were also included for evaluation, because these enzymes are reported downregulated in liver tumors, and dysregulation of these enzymes might play a part in the differential response to MCR.
As a result of the clustering analysis, we found that these cell lines were clustered into an epithelial group and a mesenchymal group. Sensitive lines fell into the epithelial group with high HNF4α, while resistant cell lines fell into the mesenchymal group with low HNF4α, suggesting a role of HNF4α in the sensitivity to MCR. In addition, SAA enzymes were enriched in the epithelial group with HNF4α.
Given the positive correlation of SAA enzymes and HNF4α, a transcription factor that regulates numerous hepatic genes, we hypothesized a transcriptional link between the two. Indeed, these enzymes all possess multiple HNF4α-binding sites, which was confirmed by our ChIP and luciferase reporter experiments.
We subsequently interrogated the biological consequences of HNF4α deficiency. We found that knocking down HNF4α reduced the expression of SAA enzymes, altered the SAA metabolism, promoted EMT, and increased resistance to MCR-induce cell death. Importantly, knocking down individual SAA enzymes induced similar outcomes, indicating an integral role of SAA metabolism in HNF4α-mediated MCR sensitivity.
Based on our findings in this study, we postulate that dietary methionine restriction provides a feasible therapy for liver cancer, either alone or in combination with available chemotherapy such as sorafenib. Clinically, HNF4α could potentially serve as a biomarker for predicting the sensitivity to methionine restriction and selecting prospective patients. In addition, the finding that restoration of SAA metabolism could alleviate HNF4α deficiency-induced outcomes suggests that supplement of SAA metabolites, such as H2S donors, might boost the efficacy of methionine restriction.