Proteomic analysis of human hepatoma cells expressing methionine adenosyltransferase I/III: Characterization of DDX3X as a target of S-adenosylmethionine

Abstract

Methionine adenosyltransferase I/III (MATI/III) synthesizes S-adenosylmethionine (SAM) in quiescent hepatocytes. Its activity is compromised in most liver diseases including liver cancer. Since SAM is a driver of hepatocytes fate we have studied the effect of re-expressing MAT1A in hepatoma Huh7 cells using proteomics. MAT1A expression leads to SAM levels close to those found in quiescent hepatocytes and induced apoptosis. Normalization of intracellular SAM induced alteration of 128 proteins identified by 2D-DIGE and gel-free methods, accounting for deregulation of central cellular functions including apoptosis, cell proliferation and survival. Human Dead-box protein 3 (DDX3X), a RNA helicase regulating RNA splicing, export, transcription and translation was down-regulated upon MAT1A expression. Our data support the regulation of DDX3X levels by SAM in a concentration and time dependent manner. Consistently, DDX3X arises as a primary target of SAM and a principal intermediate of its antitumoral effect. Based on the parallelism between SAM and DDX3X along the progression of liver disorders, and the results reported here, it is tempting to suggest that reduced SAM in the liver may lead to DDX3X up-regulation contributing to the pathogenic process and that replenishment of SAM might prove to have beneficial effects, at least in part by reducing DDX3X levels. This article is part of a Special Issue entitled: Proteomics: The clinical link.

Fig. 1

Effect of MAT1A overexpression in Huh7 cells. Western blotting and densitometric measurement for MATI/III protein estimation (A) and intracellular SAM levels (B) in pMAT1A and pEGFP transfected cells. Experiments were performed at different time intervals after transfection. MAT and SAM levels are expressed as % of control condition (100% SAM was 0.3286±0.036 pmol/μg protein) at time 0 hours of transfection. Mean ± SEM from three different experiments is shown. Equal protein loading in SDS-PAGE gels was assessed using an antibody against β-ACTIN. Representative blot from three independent experiments is shown (A).

Fig. 2

Differential proteins in MAT1A expressing Huh7 cells as evidenced by DIGE analysis. Representative 2-D gels from cytosolic and microsomal proteomes of MAT1A expressing Huh7 cells are shown. Differential spots which were subsequently identified by nano-LC-ESI-MS/MS are indicated.

Fig. 3

Functional clustering of differential proteins of MAT1A expressing Huh7 cells using Ingenuity. Data from both proteomic approaches was grouped according to Ingenuity interaction parameters. The best network with a score of 53 is represented in (A). Symbols representing the nodal proteins in nodes refer to their functional classification (red and green are up- and down-regulated, respectively) (see legend in Supplemental Fig. 3). Nodes are interconnected by specific links according to the type of interaction. A cluster of NFkB interacting proteins displayed differential expression patterns in MAT1A expressing Huh7 cells. Activation of NFkB was further confirmed by the increase of phospho-NFkB p65. The second top-rank network with a score of 44 is represented in (B). Ras homolog displayed differential expression patterns in MAT1A expressing Huh7 cells. Inhibition of Ras homolog was further suggested by the increase of the inhibitor molecule phospho-C-RAF Ser259 and the parallel decrease of activator molecule phospho-C-RAF Ser338 as evidenced by Western blot analysis (C). ERK/MAPK and PI3-K/AKT impairment was suggested by the differential expression of some of their target proteins (B), as well as by their decreased phosphate content (C). In agreement, increased apoptosis in MAT1A expressing Huh7 cells was assessed by Western blot (C). Equal protein loading in SDS-PAGE gels was demonstrated using an antibody against β-ACTIN. Representative blot from three independent experiments is shown (C).

Fig. 4

Downregulation of DDX3X in MAT1A expressing Huh7 cells. DDX3X levels were measured at different time intervals on MAT1A expressing Huh7 cells. Western blotting and their densitometric measurements (A), expressed as % of control at time 0. Mean ± SEM from three different experiments is shown. Subcellular localization of DDX3X protein was assessed by subcellular fractionation; a significant decrease in the cytosolic fraction and an increase in the nuclear fraction (B) were observed. Equal protein loading in SDS-PAGE gels was demonstrated using an antibody against β-ACTIN. Representative blot from three independent experiments is shown (B).

Fig. 5

Effect of SAM on DDX3X levels on human hepatoma cell lines. DDX3X and SAM levels are expressed as % of control condition (100% SAM=0.3585±0.021 pmol/μg protein) (A). Mean ± SEM from three different experiments is shown. The sulfate-p-toluensulfonate salt of SAM was used. Western blot of DDX3X protein in Huh7 cells is shown in (A), whereas DDX3X content in HepG2 and Hep3B human hepatoma cell lines is shown in (B). A representative blot is shown from three independent experiments (A, B). Equal protein loading in SDS-PAGE gels was assessed using an antibody against β-ACTIN.