HTML
-
Baseline age and sex data were matched between the CHB patients and HCs. HBV was more active in females than in males, as HBV DNA and HBeAg levels were higher in females (P < 0.001). However, males exhibited worse liver dysfunction, as indicated by higher serum concentrations of alanine aminotransferase (ALT), aspartate transaminase (AST), gamma-glutamyltransferase (GGT), and albumin (ALB) (P < 0.05). The AST-to-ALT ratio (AAR), S index (1000 × GGT (IU/L)/[platelet(109/L × ALB2(g/L)]), and GGT-to-platelet ratio (GPRI) have been suggested as inexpensive and readily available methods to assess the degree of liver fibrosis in CHB patients (Bonacini et al. 1997; Zhou et al. 2010; Zeng et al. 2015). The GPRI score was positively correlated with the liver stiffness measurement (LSM) value (r = 0.513, P < 0.001) and closely related to the histopathological score (r = 0.372, P < 0.001). GPRI has good diagnostic accuracy and is a robust serum prediction model for liver fibrosis in CHB patients (Zeng et al. 2015). In the present study, male CHB patients had a higher S index and GPRI scores than did female CHB patients (P < 0.001), suggesting that liver fibrosis was more extensive in male CHB patients. Interestingly, the clinical parameters mentioned above were not significantly different between male and female HCs (P > 0.05) (Table 1).
Index Male HCs Female HCs Male CHBs Female CHBs Case number 18 11 717 207 Age (year) 34.27 ± 11.14 41.15 ± 11.62 35.95 ± 10.24 35.50 ± 11.64 ALT (U/L) 20.47 ± 5.49 21.20 ± 12.31 76.72 ± 103.74 52.38 ± 69.17** AST (U/L) 20.88 ± 5.57 21.20 ± 7.98 53.91 ± 59.47 46.24 ± 43.02* GGT (U/L) 20.59 ± 4.50 23.40 ± 14.71 59.51 ± 73.49 30.58 ± 40.05** ALB (g/L) 43.63 ± 4.06 44.00 ± 6.54 45.42 ± 3.95 44.32 ± 4.49** Uric acid (μmol/L) 328.88 ± 52.21 277.80 ± 40.77* 304.62 ± 76.94 237.76 ± 52.08** Bile acid (μmol/L) 8.41 ± 2.12 6.80 ± 1.23 13.83 ± 23.16 7.29 ± 9.80 PT (s) 12.76 ± 0.96 12.44 ± 0.52 12.96 ± 2.00 12.66 ± 1.33 CD3 + (%) 72.85 ± 5.88 69.14 ± 8.71 70.11 ± 9.20 69.87 ± 9.05 CD4 + (%) 36.12 ± 6.28 32.52 ± 8.80 35.40 ± 8.69 38.24 ± 8.30 CD8 + (%) 36.32 ± 8.22 38.00 ± 11.45 27.41 ± 7.69 26.22 ± 5.98 Lymphocyte (× 109/L) 1.79 ± 0.41 1.72 ± 0.37 1.839 ± 0.663 1.79 ± 0.57 IgA(g/L) 2.16 ± 0.90 2.42 ± 0.95 3.45 ± 1.59 3.40 ± 0.95 IgM (g/L) 1.26 ± 0.47 1.96 ± 0.91* 1.68 ± 0.80* 2.12 ± 1.09 IgG (g/L) 11.99 ± 2.61 14.24 ± 2.53 17.49 ± 4.42 18.01 ± 4.21 AAR 1.06 ± 0.28 1.19 ± 0.57 0.91 ± 0.41 1.08 ± 0.41** S index 0.03 ± 0.01 0.04 ± 0.02 0.14 ± 0.22 0.06 ± 0.10** GPRI 0.11 ± 0.04 0.14 ± 0.09 0.46 ± 0.68 0.20 ± 0.32** HBsAg (IU/mL) 0.00 ± 0.01 0.01 ± 0.01 237.75 ± 100.67 232.64 ± 57.74 HBeAg (PEIU/mL) – – 275.35 ± 444.71 423.44 ± 516.1** HBeAb (1S/CO) – – 12.54 ± 18.00 19.09 ± 20.43** HBcAb (1S/CO) – – 20.99 ± 300.97 9.17 ± 2.83* HbcAb-IgM(1S/CO) – – 0.54 ± 2.27 0.31 ± 0.46* HBV-DNA (copies/mL) – – 2.13 × 107 ± 1.89 × 108 2.3 × 107 ± 6.35 × 107** ALT alanine aminotransferase, AST aspartate aminotransferase, GGT Glutamyl transpeptidase, ALB albumin, PT prothrombin time, IgA Immunoglobulin A, IgM Immunoglobulin M, IgG Immunoglobulin G, AAR AST-to-ALT ratio, GRPI GGT-to-platelet ratio.
Data was shown as Mean ± SD.
*P < 0.05: male HCs vs female HCs or male CHBs vs female CHBs.
**P < 0.001: male HCs vs female HCs or male CHBs vs female CHBs.Table 1. Clinical parameters of CHB patients and HCs.
-
Genome-wide DNA methylation detection revealed 5529 and 4984 DM loci between male and female CHB patients and male and female HCs, respectively (|Δβ value| > 0.17 and P < 0.05). The top 25% of DM loci among these four groups were analyzed with PLS-DA, demonstrating that males and females, both CHB patients and HCs, were clearly separated (Fig. 1A). In total, 229 DM CpG loci (160 hypomethylated loci and 69 hypermethylated loci) in 127 genes were observed between male CHB patients and male HCs, and 254 DM loci (91 hypomethylated loci and 163 hypermethylated loci) in 142 genes were observed between female CHB patients and female HCs (Supplementary Table S1, S2). These results suggested that male CHB patients tended to exhibit more DNA hypomethylation, while female CHB patients were more susceptible to DNA hypermethylation (Fig. 1A). DM loci between male CHB patients and male HCs were predominantly located in the gene body (39%), followed by the intergenic region (IG, 34%) and promoter (including TSS 1500, TSS 200, 5′UTR, and 1st Exon; 23%). In contrast, DM loci between female CHB patients and female HCs were mainly distributed at the promoter (33%), IG (30%), and gene body (34%). About half of those DM loci were located in nonCpG islands (male CHB patients: 52%, female CHB patients: 47%), followed by CpG islands (male CHB patients: 18%, female CHB patients: 23%) (Fig. 1B, 1C). The genomic distribution of DM loci between sex was not statistically different (P > 0.05).
Figure 1. Genomic DNA methylation profiles of male and female CHB patients and HCs. A PLS-DA analysis of the top 25% DM loci among the 4 groups. B Distribution of DM loci between males with CHB and male HCs in relation to chromosome, CpG island, and gene region. C Distribution of DM loci between females with CHB and female HCs in relation to chromosome, CpG island, and gene region. Δβ value refers to the DNA methylation level of male/female CHB patients minus that of male/female HCs.
-
Venn diagram analysis identified 111 male CHB patientspecific and 126 female CHB patient-specific DM genes (Fig. 2A). The male CHB patient-specific DM genes were mainly enriched in the biological processes of brain development, such as cerebellar granular layer development, cell differentiation in hindbrain, and cerebellar cortex formation. On the other hand, female CHB patientspecific DM genes were primarily enriched in immunerelated processes, including intestinal immune network for IgA production, asthma, allograft rejection, and graft-versus-host disease (P < 0.05). Furthermore, the biological process terms were shown to closely interact (Fig. 2B, 2C).
Figure 2. Venn diagram analysis and DM gene involvement in biological processes between sex in CHB patients. A Venn diagram analysis; B Interaction network of biological processes related to male CHB patientspecific DM genes; C Interaction network of biological processes related to female CHB patient-specific DM genes. Different colors of nodes represent different biological processes. Larger node size signifies smaller P value.
-
miR509-1, miR509-2, and miR509-3 were located in sex chromosomes (CHR X: 146340534, TSS 200) and were hypermethylated in both male and female CHB patients than in the corresponding HCs. Among male CHB patientspecific and female CHB patient-specific DM genes, VCX, TTTY1, and TSPY1 were also differentially methylated between male CHB patients and female CHB patients (|Δβ value| > 0.17 and P < 0.05) but not between male HCs and female HCs (P > 0.05). The CHR X: 7810800 locus in the 5′UTR of VCX was hypomethylated in male CHB patients as compared to that in female CHB patients (0.356 ± 0.346 vs 0.820 ± 0.070, P < 0.001), whereas its methylation level was not significantly different between male HCs and female HCs (0.860 ± 0.025 vs 0.897 ± 0.001, P > 0.05), according to the Human Methylation 450 K Assay results. The |Δβ value| of the CHR X: 7810800 locus was the highest (-0.504), and the 5′UTR is critical in gene expression. Therefore, its DNA methylation level was further validated, and its mRNA expression and SNP genotypes were detected in a larger sample set.
Pyrosequencing of larger sample sets showed that the CHR X: 7810800 locus was significantly hypermethylated in female CHB patients as compared to that in male CHB patients (0.298 ± 0.159 vs 0.233 ± 0.219, P < 0.05), whereas it was hypomethylated in female HCs as compared to that in male HCs (0.161 ± 0.161 vs 0.334 ± 0.196, P > 0.05) (Fig. 3A). VCX expression was upregulated in female CHB patients as compared to that in male CHB patients (11.437 ± 15.933 vs 1.778 ± 2.429, P < 0.05) but did not differ significantly between female HCs and male HCs (2.873 ± 4.150 vs 1.000 ± 1.026, P > 0.05) (Fig. 3B). In this study, the CHR X: 7810800 locus displayed three genotypes, namely CC, CT, and TT. The frequency of CC + CT was considerably higher in female CHB patients than in male CHB patients (females vs males: 76% vs 40%, P < 0.05), but was not correspondingly higher in female HCs than in male HCs (females vs males: 42% vs 65%, P > 0.05) (Fig. 3C).
-
Correlation analysis was performed to explore the relationship among SNP genotypes and DNA methylation at the CHR X: 7810800 locus and the VCX transcriptional level. CC + CT genotypes displayed a higher DNA methylation level than the TT genotype (0.392 ± 0.133 vs 0.110 ± 0.158, P < 0.001) (Fig. 4A), while VCX mRNA expression was not statistically different between genotypes (1.000 ± 2.148 vs 0.680 ± 1.522, P > 0.05). In CHB patients with CC + CT genotypes, the VCX mRNA level was negatively correlated with its DNA methylation level at the CHR X: 7810800 locus (r = -0.271, P < 0.01) (Fig. 4B). To further evaluate the involvement of DNA methylation in gene expression, we treated the K562 cell line with 5-Aza, a commonly used demethylating agent, at concentrations of 0.001, 0.1, and 10 μmol/L, respectively. Ten μmol/L 5-Aza led to notably higher VCX mRNA expression in K562 cells than 0, 0.001, and 0.1 μmol/L 5-Aza did, indicating that general DNA demethylation may increase VCX mRNA expression (Fig. 4C).
Figure 4. Relationship among SNP genotypes, DNA methylation level at the CHR X: 7810800 locus, VCX mRNA level, and liver function and fibrosis in CHB patients. A Relationship between SNP genotypes and DNA methylation level; B Correlation between VCX mRNA and its DNA methylation levels in the CC + CT genotype; C VCX mRNA expression following 5-Aza treatment; D Relationship between VCX mRNA and AST levels; E Relationship between VCX mRNA and GGT levels; F Relationship between VCX mRNA levels and GPRI values; G Relationship between VCX mRNA levels and HBV DNA load. *P < 0.05; **P < 0.001.
GPRI has been reported as a non-invasive serum marker for diagnosis of severe liver fibrosis in CHB patients, as it was positively correlated with histopathological scores (r = 0.372, P < 0.001); the optimal cutoff value of GPRI is 0.2343 with 58.24% specificity and 79.35% sensitivity (Zeng et al. 2015). Therefore, we used the cutoff values of 40 U/L, 50 U/L, and 0.2343 for AST, GGT, and GPRI, respectively, to determine the liver function and extent of fibrosis in CHB patients. As shown in Figs. 4D–F, VCX mRNA expression was lower in CHB patients with higher AST and GGT levels as well as GPRI score. Additionally, the CHR X: 7810800 methylation level was negatively correlated with GGT (r = -0.144, P < 0.05) and uric acid levels (r = -0.178, P < 0.05) but was positively correlated with the proportion of CD4 + T cells (r = 0.154, P < 0.05) and lymphocyte counts (r = 0.151, P < 0.05). VCX mRNA expression was negatively correlated with ALT (r= -0.246, P < 0.05), ALP (r = -0.332, P < 0.001), bile acid levels (r =-0.212, P < 0.05), and prothrombin time (PT) (r = -0.195, P < 0.05). On the other hand, VCX mRNA expression was positively correlated with the proportion of CD3 + T cells (r = 0.142, P < 0.05) and IgA level (r = 0.236, P < 0.05). With respect to CC + CT genotypes at the CHR X: 7810800 locus, VCX expression was increased in CHB patients with a higher HBV DNA load (over 1 × 106 copies/mL) (Fig. 4G). These results indicated that CHR X: 7810800 hypomethylation and VCX mRNA downregulation in male CHB patients cause predisposition to worse liver dysfunction, less active immunity, and more significant fibrosis than in female CHB patients.
-
To explore whether VCX protein was secreted into the plasma and to evaluate sex-based differences in expression, we detected the VCX plasma concentration in 15 male CHB patients, 15 female CHB patients, 6 male HCs, and 6 female HCs using Western blot. As shown in Fig. 5, plasma VCX concentration in female CHB patients was higher than that in male CHB patients (P < 0.05), whereas it was not statistically different between male and female HCs (P > 0.05), which was consistent with the VCX mRNA expression in PBMCs (Fig. 5). Therefore, we inferred that higher plasma VCX expression contributed to better resistance to CHB in female patients.