Transcriptomic profiling of hepatic tissues for drug metabolism genes in nonalcoholic fatty liver disease: A study of human and animals

Abstract

Background: Drug metabolism genes are involved in the in vivo metabolic processing of drugs. In previous research, we found that a high-fat diet affected the transcript levels of mouse hepatic genes responsible for drug metabolism.

Aims: Our research intends to discover the drug metabolism genes that are dysregulated at the transcriptome level in nonalcoholic fatty liver disease (NAFLD).

Methods: We analyzed the transcriptome for drug metabolism genes of 35 human liver tissues obtained during laparoscopic cholecystectomy. Additionally, we imported transcriptome data from mice fed a high-fat diet in previous research and two open-access Gene Expression Omnibus (GEO) datasets (GSE63067 and GSE89632). Then, using quantitative real-time polymerase chain reaction (qRT-PCR), we cross-linked the differentially expressed genes (DEGs) in clinical and animal samples and validated the common genes.

Results: In this study, we identified 35 DEGs, of which 33 were up-regulated and two were down-regulated. Moreover, we found 71 DEGs (39 up- and 32 down-regulated), 276 DEGs (157 up- and 119 down-regulated), and 158 DEGs (117 up- and 41 down-regulated) in the GSE63067, GSE89632, and high-fat diet mice, respectively. Of the 35 DEGs, nine co-regulated DEGs were found in the Venn diagram (CYP20A1, CYP2U1, SLC9A6, SLC26A6, SLC31A1, SLC46A1, SLC46A3, SULT1B1, and UGT2A3).

Conclusion: Nine significant drug metabolism genes were identified in NAFLD. Future research should investigate the impacts of these genes on drug dose adjustment in patients with NAFLD.

Clinical trial registration: http://www.chictr.org.cn, identifier ChiCTR2100041714.

Keywords: GEO datasets; drug metabolism genes; high-fat diet; nonalcoholic fatty liver disease; transcriptome.

Figure 1

Flow diagram of the study design. The four cohorts were the human liver samples collected, the GEO dataset GSE63067, the GEO dataset GSE89632, and our previous mouse data. PCA, principal component analyses; NAFLD, nonalcoholic fatty liver disease; HFD, high-fat diet; DEGs, differentially expressed genes; qRT-PCR, quantitative real-time polymerase chain reaction.

Figure 2

Heat map visualization. (A–D) Hierarchical clustering of drug metabolism genes was used by heat map. (A–C) were the analysis results in the human liver samples collected, the GEO dataset GSE63067, and the GEO dataset GSE89632, respectively. (D) belonged to the analysis results of mouse data in our previous study. NAFLD, nonalcoholic fatty liver disease; HC, healthy control; ND, normal diet; HFD, high-fat diet.

Figure 3

PCA and cluster analysis of the four merged samples after batch correction. (A) PCA before batch correction; (B) PCA after batch correction. (C) Cluster analysis of samples from the four merged after batch correction. A scattered point of PCA represented a sample and included the calculated values of all 476 drug metabolism genes for each sample. The ellipse indicated the 95% confidence interval within a group for the PCA. In A and B, the HC group was in red, and the NAFLD group was in blue. In C, Branch length represents the distance between samples, with shorter branches for more closely related samples. Samples in the same color belong to the same group. PCA, principal component analyses. NAFLD, nonalcoholic fatty liver disease; HC, healthy control.

Figure 4

PCA for each of the four cohorts (A–D). (A–C) were the analysis results in the human liver samples collected, the GEO dataset GSE63067, and the GEO dataset GSE89632, respectively. (D) belonged to the analysis results of mouse data in our previous study. A scattered point of PCA represented a sample and included the calculated values of all 476 drug metabolism genes for each sample. The ellipse indicated the 95% confidence interval within a group for the PCA. The HC groups are in red, and the NAFLD groups are blue in the human samples, while the ND groups are in red, and the HFD groups are blue in the mouse data. PCA, principal component analyses; NAFLD, nonalcoholic fatty liver disease; HC, healthy control; ND, normal diet; HFD, high-fat diet.

Figure 5

Volcano plot of differentially expressed drug metabolism genes for four cohorts. (A–D) The four cohorts were the human liver samples collected, the GEO dataset GSE63067, the GEO dataset GSE89632, and our previous mouse data. The screening criteria for DEGs were as follows: P-value <0.05 and |fold change|>1. Red represented up-regulated genes, green represented down-regulated genes, and black represented no significant DEGs. DEGs, differentially expressed genes.

Figure 6

Venn diagram of common DEGs associated with drug metabolism among the four cohorts. (A) Venn diagram of up-regulated genes; (B) Venn diagram of down-regulated genes. The four cohorts were the human liver samples collected, the GEO dataset GSE63067, the GEO dataset GSE89632, and our previous mouse data.

Figure 7

Validation of RNA-seq data on nine common DEGs in the human liver tissues using qRT-PCR technology. P<0.05 was checked by Pearson’s correlation coefficient. RNA-seq, RNA sequencing; qRT-PCR, quantitative real-time polymerase chain reaction; DEGs, differentially expressed genes.

Figure 8

Variation of nine common DEGs between the NAFLD group and the HC group in human liver tissues by qRT-PCR. *P<0.05 was checked by the student t-test. qRT-PCR, quantitative real-time polymerase chain reaction; DEGs, differentially expressed genes; NAFLD, nonalcoholic fatty liver disease; HC, healthy control.