GAS5 rs2067079 and miR-137 rs1625579 functional SNPs and risk of chronic hepatitis B virus infection among Egyptian patients

Hepatitis B virus (HBV) infection is a significant health issue worldwide.. We attempted to fulfill the molecular mechanisms of epigenetic and genetic factors associated with chronic HBV (CHBV). Expression levels of the lncRNA growth arrest-specific 5 (GAS5) and miR-137 and their corresponding SNPs, rs2067079 (C/T) and rs1625579 (G/T) were analyzed in 117 CHBV patients and 120 controls to investigate the probable association between these biomarkers and CHBV pathogenesis in the Egyptian population. Serum expression levels of GAS5 and miR-137 were significantly down-regulated in cases vs controls. Regarding GAS5 (rs2067079), the mutant TT genotype showed an increased risk of CHBV (p < 0.001), while the dominant CC was a protective factor (p = 0.004). Regarding miR-137 rs1625579, the mutant genotype TT was reported as a risk factor for CHBV (p < 0.001) and the normal GG genotype was a protective factor, p < 0.001. The serum GAS5 was significantly higher in the mutant TT genotype of GAS5 SNP as compared to the other genotypes (p = 0.007). Concerning miR-137 rs1625579, the mutant TT genotype was significantly associated with a lower serum expression level of miR-137 (p = 0.018). We revealed the dysregulated expression levels of GAS5 and miR-137 linked to their functioning SNPs were associated with CHBV risk and might act as potential therapeutic targets.


Results
Demographic and clinical characterization of the study groups. This case-control study included 117 patients with chronic HBV infection and 120 controls that had previous HBV infection and completely recovered from infection. Among the patients, the male/female ratio was 3.88 and it was 3.81 for control. No significant difference was observed between cases and controls regarding age and sex (p = 0.206 and 0.951, respectively) ( Table 1). Table 1 also demonstrates the clinical and laboratory data of the patients recruited in the study. In CHBV patients, the serum expression level of GAS5 and miR-137 was significantly down-regulated in cases vs controls, p < 0.001 for each (Table 1 and Fig. 1).

Association of GAS5 rs2067079 (C/T) and miR-137 rs1625579 (G/T) with the risk of CHBV.
The strength of the association between the SNPs and risk of CHBV was assessed by the ORs with the corresponding 95% CI for the following genetic models: (1) the codominant model (C/C vs C/T vs T/T for rs2067079 and G/G vs G/T vs T/T for rs1625579), (2) the dominant genetic model: (CC vs T/T + C/T) (C is the major allele and T is the minor allele for rs2067079) and (G/G vs T/T + G/T) (G is the major allele and T is the minor allele for rs1625579), (3) the recessive genetic model (T/T vs. C/T + C/C and T/T vs. G/T + G/G for rs2067079 and rs1625579, respectively), (4) the over-dominant model (CT vs CC + TT and GT vs GG + TT for rs2067079 and rs1625579, respectively) and (5) C allele vs the T allele and G allele vs T allele analysis for rs2067079 and rs1625579, respectively) ( Table 2).
For both SNPs, the minor allele frequency (MAF) in the controls was 25.8% for rs2067079 and 33.8% for rs1625579, which was slightly higher than the global MAF (T = 19% for rs2067079 and G = 20% for rs1625579) as reported in Ensembl GRCh37 release 16 .
The genotypes distribution of rs2067079 and rs1625579 in the control did not significantly deviate from what was expected by the Hardy-Weinberg equilibrium (HWE) (p = 0.056 and 0.08 respectively).

Association of GAS5 rs2067079 (C/T) with the risk of CHBV. The genotype and allele frequencies
for rs2067079 in CHBV patients and control subjects are shown in Table 2. Regarding the rs2067079, the recessive T allele predominated among cases (42.3% vs 25.8% in controls) and the dominant C allele predominated among controls (74.2% vs 57.7% in cases). In the co-dominant model (CC vs CT vs TT), the distribution of the rs2067079 genotypes was significantly different between CHBV patients and control subjects (CT genotype vs CC genotype, p < 0.001; and TT genotype vs CC genotype, p = 0.005). The CT and TT genotypes represented an increased risk of CHBV for about 8 and 5 folds respectively. According to the recessive model (TT vs CC + CT), the TT genotype showed a 6. Association of miR-137 rs1625579 (G/T) with the risk of CHBV. As reported in Table 2 Association of serum GAS5 and miR-137 levels with rs2067079 and rs1625579 in CHBV patients. We evaluated serum expression levels of GAS5 and miR-137 in CHBV patients having the different SNP genotypes to interpret the role of rs2067079 and rs1625579 in this disease (Fig. 3). By studying the effect of rs2067079, we found that the serum GAS5 expression level was significantly higher in the mutant TT genotype carriers vs those with the CT or CC genotypes (p = 0.007) (Fig. 3a). However, this SNP did not affect the expression level of miR-137 (p = 0.056) (Fig. 3b).
Regarding the rs1625579, we did not find a significant difference in the serum expression of GAS5 among CHBV patients with different rs1625579 genotypes (p = 0.489) (Fig. 3c). In contrast, serum miR-137 expression level was significantly down-regulated with the mutant TT genotype versus GT and GG genotypes among CHBV patients (p = 0.018) (Fig. 3d).
Effect of GAS5 rs2067079 (C/T) and miR-137 rs1625579 (G/T) on demographic, clinical, and laboratory variables of chronic viral hepatitis B patients. By studying the effect of GAS5 rs2067079 (C/T) on demographic, clinical, and laboratory variables of CVHB B patients (Table 3), the recessive TT genotype was significantly associated with low body mass index (BMI), p = 0.001, and with high GAS5 serum expression level, p = 0.007. The CT genotype was predominantly associated with males p = 0.033.
On the other hand, as regards the miR-137 rs1625579 (G/T) effect on patients' characteristics (Table 4), the mutant TT genotype was significantly associated with young age (p = 0.033), high levels of total and direct Table 1. Demographic, clinical and laboratory of study subjects. Data are expressed as mean ± SEM (standard error of mean); expression values of GAS5 and miR-137 in control group were set as 1. BMI body mass index, GAS-5 lncRNA growth arrest-specific 5, miR-137 microRNA 137, AFP alpha fetoprotein, FBS fasting blood sugar, IQR interquartile range, Hb hemoglobin, TLC total leucocytic count, PLT platelet count, INR the international normalized ratio, kPa kilopascals, ALT alanine aminotransferase, AST aspartate aminotransferase, BDL below detection level (below 20 copies/ml), low load viral load 20-10 4 copies/ml, high load viral load > 10 4 copies/ml. *Significant. www.nature.com/scientificreports/ bilirubin (p = 0.011 and < 0.001, respectively), and lower serum expression level of miR-137 (p = 0.018). The GT genotype was associated with high fasting blood sugar (FBS) (p = 0.011).
No association was found between GAS-5 serum expression level and presence of HBeAg (p = 0.736) or HBV viral load in the serum of CHBV cases (p = 0.263). Similarly, miR-137 had no association with HBeAg (p = 0.709), but it was significantly high with very low HBV viral load (less than 20 copies/ml) among CHBV cases (p = 0.035) ( Table 6).  Table 2. Association of GAS5 rs2067079 (C/T) and miR-137 rs1625579 (G/T) with the risk of chronic hepatitis B viral infection. For rs2067079 (C/T), C is the dominant allele, for rs1625579 (G/T), G is the dominant allele. GAS-5 lncRNA growth arrest-specific 5, miR-137 microRNA 137, OR odds ratio, CI confidence interval. *Significant. www.nature.com/scientificreports/ Logistic regression analysis. In our study, in the multivariate analysis, the rs2067079 mutant TT genotype (p = 0.033) and the rs1625579 mutant TT genotype (p < 0.0001) were revealed to be significant positive independent predictors of CHBV risk (Table 7).

Discussion
Chronic infection with HBV is associated with severe clinical consequences including HCC development 17 .
We have explored the influence of the expression levels of GAS5 and its rs2067079 SNP and miR-137 and its rs1625579 SNP on CHBV susceptibility, and we have demonstrated them as potential genetic biomarkers. In our study, GAS5 expression was significantly down-regulated in CHBV patients compared to controls. Our results were consistent with Feng et al., who reported down-regulation of GAS5 in patients with CHBV when compared to healthy controls 18 . Furthermore, in a study that identified the lncRNAs expression patterns in CHBV, GAS5 was down-regulated as compared to the healthy controls. However, our results were inconsistent with previous studies performed by Tu et al. and Chang et al., who determined that there was no significant correlation between GAS5 expression and HBV infection in patients with HCC 19,20 .
Accumulating evidence reveals that dysregulated GAS5 has been documented in many human diseases including malignancy 21 , childhood pneumonia 22 , autoimmune disorders 23 , heart failure, diabetes mellitus 24 , and neuropsychiatric disorders 24 . Importantly, GAS5 is widely considered as a tumor suppressor. It has been found to suppress liver fibrosis and inhibit the migration and invasiveness of HCC cells through miR-21 sequestration affecting cell survival 25 .
It is worth noting that both IFN-α and IFN-λ levels remained low during HCV infection after GAS5 overexpression, revealing the ability of GAS5 to inhibit innate immune responses after viral infection 18 .
In this study we detected, for the first time up to our knowledge, a significant down-regulation in serum miR-137 expression level in CHBV patients compared to controls. Also, it was significantly high with very low HBV viral load (less than 20 copies/ml) in contrast to those with high viral load (> 10 4 copies/ml) among CHBV cases (p = 0.035). It was found that HBV X protein (HBx), a protein produced by HBV, can modulate the expression and activity of many genes and epigenetic molecules like lncRNAs and miRNAs resulting in the dysfunction of several pathways 3 . Also, cellular miRNAs can regulate HBV infection by targeting transcription factors or by direct binding to HBV transcripts. Interestingly, miR-137 overexpression decreased the cancer-initiating cell features, impaired invasion abilities, and metastasis-associated properties of HCC cells 26 . Depending on the observations of many studies, miRNAs expression is stable, unlike antigens, because they are not directly dependent on the body's immune response. Thus, miRNAs are considered excellent biomarkers associated with HBV-related HCC 28 . It was reported that miR-137 was down-regulated in HCC cells, suggesting a potential therapeutic role of miR-137 in HCC treatment 29 .
Our results demonstrated that both GAS5 and miR-137 serum expression levels had a significant positive correlation with each other. These results were in agreement with Bian et al., who reported that there was a positive association between GAS5 and miR-137 in melanoma tissues and GAS5 was positively regulated the expression of miR-137. Simultaneously, it was determined that GAS5 could inhibit cell proliferation, migration, and invasion of melanoma cells through miR-137 30 . However, a previous study indicated that the expression of miR-137 was negatively associated with GAS5 up-regulation in the mice exposed to occlusion of the middle cerebral artery and to oxygen-glucose deprivation that stimulated primary brain neurons 31 .
Regarding GAS5 rs2067079 (C/T) in our study, the mutant CT and TT genotypes represented an increased risk of CHBV while the normal CC genotype was defined as a protective factor against CHBV risk. Up to our knowledge, no one has linked GAS5 rs2067079 with CHBV risk or pathogenesis although differentially expressed levels of GAS5 have been documented in many studies. It was reported that GAS5 rs2067079 (TT vs CC) was associated with chemo-radiotherapy induced severe myelosuppression and severe neutropenia in patients with nasopharyngeal carcinoma 32 . It was also found that this SNP was associated with the risk of development of bladder cancer 33 , multiple sclerosis 23 , and systemic lupus erythematosus (SLE) 34 .
Besides, our findings showed that the serum GAS5 expression level was significantly higher in the risky rs2067079 TT genotype compared to the CT or CC genotypes in CHBV patients. The supporting ChIP-seq data from different human cell types revealed that the site of the rs2067079 site was considered as a strong promoter region. Since the expression levels could be regulated by genetic variants in regulatory elements, this SNP might affect GAS5 transcriptional activity. It was revealed that rs2067079 has a strong feature of expression in many tissues. Furthermore, rs2067079 had a strong effect on GAS5 secondary structure which is critical for its performance 32 . Inconsistent with our results, Li et al., 2017 determined that the rs2067079 did not affect GAS5 expression levels among Chinese Han patients with SLE 34 . This disagreement may be due to the differences in ethnicity and disease mechanisms. Table 3. Effect of GAS5 rs2067079 (C/T) on demographic, clinical and laboratory variables of chronic viral hepatitis B patients. BMI body mass index, GAS-5 lncRNA growth arrest-specific 5, miR-137 microRNA 137, AFP alpha fetoprotein, FBS fasting blood sugar, IQR interquartile range, SD standard deviation, Hb hemoglobin, TLC total leucocytic count, PLT platelet count, INR the international normalized ratio, kPa kilopascals, ALT alanine aminotransferase, AST aspartate aminotransferase. *Significant.    www.nature.com/scientificreports/ Furthermore, regarding miR-137 rs1625579, the mutant TT genotype and the T allele were predominantly found in CHBV cases as compared to controls. Also, there was a strong association between rs1625579 and CHBV risk. The mutant genotype TT was detected as a significant risk factor while the normal GG genotype was reported as a protective factor for CHBV. These results are suggesting a potential role of this SNP in CHBV pathogenesis.
Up to our knowledge, we demonstrated for the first time the influence of miR-137 polymorphism on CHBV susceptibility. MiR-137 rs1625579 was investigated mainly in neurodegenerative diseases such as schizophrenia. In a genome-wide association study, the strong association of the miR-137 locus with schizophrenia was evidenced 35 . Furthermore, the rs1625579 polymorphism, which is an intron SNP in the miR-137 gene, was associated with the onset of schizophrenia in Asian and European populations 36 .
Regarding the rs1625579, we reported a significant down-regulation in serum miR-137 expression level with the mutant TT genotype against GT and GG genotypes among CHBV patients. It was demonstrated that genetic variation of the miR-137 may affect the transcription process or binding ability between the miRNA and its target genes and contribute to its abnormal expression, eventually triggering the onset of many disorders 37 . It was indicated that variations in the non-coding region of miR-137 locus lead to defective mRNA structure, splicing, and stability 38 . Decreased miR-137 expression was associated with rs1625579 in different neuropsychiatric disorders 39 . Conversely, in another genotype-tissue expression study, the miR-137 locus was proved to correlate with increased miR-137 expression, especially with rs1702294, particularly, in the hippocampus of carriers of the schizophrenia-associated genotype 40 .
MiRNAs, lncRNAs, and their target genes are emerging as biomarkers of many diseases 18 . Our study is the first to investigate miR-137 expression level in CHBV patients which could be considered as a probable diagnostic marker for these patients. The role of GAS5 and miR-137 and their functioning SNPs as potential genetic predictors for CHBV risk was reported in this study. Thus interestingly, we demonstrated the associated effect of genetic and epigenetic regulatory mechanisms controlling CHBV pathogenesis. Also, the development of new therapeutic strategies based on non-coding RNAs may be a good alternative to the current conventional therapy for HBV as the conventional treatment options only does not cure it but only suppress the replication of the virus. Therefore, a lifelong treatment is essential 41 . Additionally, Peg-IFN-α has a low response rate and difficult to tolerate 42 . The current conventional therapy can reduce but not eliminate the risk for HCC 42 .
Considering our observations in CHBV, more studies with larger sample size are essential to document our findings. Although population homogeneity reduces genetic variability; our reports should be considered in large independent population studies. Further studies are needed to evaluate the exact role of the rs2067079 and rs1625579 SNPs on the tissue level of GAS5 and miR-137, respectively.
In conclusion, we identified that the expression levels of GAS5 and miR-137 were down-regulated significantly among CHBV patients. Our findings indicated that rs2067079 within GAS5 and rs1625579 within miR-137 were contributed to the pathogenesis of CHBV and could serve as potential genetic biomarkers for CHBV susceptibility. We found that the CC genotype of rs2067079 (C/T) was a protective factor while the CT and TT Table 6. Relation between HBeAg and HBV viral load and lncRNA growth arrest-specific 5 (GAS5) and miR-137 serum expression levels among chronic HBV cases. BDL below detection level (below 20 copies/ml), low load viral load 20-10 4 copies/ml, high load viral load > 10 4 copies/ml. *Significant p value. www.nature.com/scientificreports/ genotypes showed an increased risk of CHBV. Regarding rs1625579 (G/T), it was demonstrated that the normal GG genotype was a protective factor while the mutant genotype TT was reported as a risk factor for CHBV. These potential genetic biomarkers could provide a better understanding of the pathogenesis of CHBV and this allows for the development of possible therapeutic approaches.

Subjects and methods
Subjects. The present case-control research was conducted on 117 patients with chronic HBV infection as well as 120 control individuals who had previous HBV infection, attending the outpatient clinic, at the Department of Internal Medicine, Fayoum University hospital, Fayoum, Egypt. The diagnosis of CHBV was based on the presence of hepatitis B virus DNA (either quantitative, or qualitative,) or detection of HBsAg (hepatitis B surface antigen) or HBeAg (hepatitis B e-antigen), in addition to the absence of anti-HBc IgM (immunoglobulin M antibodies to HBV core antigen) 43 . Any clinical and/or fibroscan (transient elastography) findings compatible with chronic liver disease were also considered.
Detailed personal and medical history was taken from each subject. Also, routine laboratory investigations such as tests for assessment of liver function, renal function tests as well as complete blood count (CBC) were determined.
The selected control individuals had positive anti-HBs IgG and anti-HBc IgG without detection of HBsAg in addition to normal liver function tests.
The exclusion criteria for all participants comprised: (1) evidence of other types of hepatic disorders (2) antiviral treatments or immunosuppressant drugs (3) presence of positive anti-HCV or anti-HIV antibodies (4) smoking more than one cigarette pack per day.
The Ethics Committee of the Fayoum university hospital has approved this research under the Declaration of Helsinki. Each participant has given written informed consent.
Handling blood samples. Venous blood samples which were withdrawn from each subject in the current work were deposited into plain vacutainers in which a gel separator was included. The samples were left to clot for 15 min and centrifuged at 4000×g for 10 min allowing the serum to be separated. Enzyme-linked immunosorbent assay (ELISA) was done on serum samples to measure HBeAg (Sunlong Biotech, China), anti-HBc, and HBsAg (Bioelisa, Biokit, Barcelona, Spain).in agreement with the manufacturer's instruction. HBV DNA was measured using COBAS ampliprep/COBAS TAQMAN HBV TEST V2.0 kit (Roche Diagnostics) following the steps recommended by the manufacturer.
The remaining serum samples were directly stored at −80 ˚C till the time of RNA extraction. EDTA-containing tubes were used to collect additional whole blood samples to be used in DNA extraction and genotyping of rs2067079 and rs1625579.

Extraction of total RNA and reverse transcription. Total RNAs (including microRNAs and lncRNAs)
were extracted from the abovementioned serum samples by using a miRNeasy extraction kit (Qiagen, Hilden, Germany) after the addition of QIAzol lysis reagent following the manufacturer's protocol. NanoDrop (ND)-1000 Spectrophotometer (NanoDrop Technologies, Inc. Wilmington, USA) was used to determine the purity of the extracted RNA.
Reverse-transcription of the extracted RNAs was performed using RT2 first strand kit (Qiagen, Maryland, USA) with a whole volume of 20 μl/ reaction for GAS5 testing. MiR-137 was analyzed by miScript II RT Kit (Qiagen Maryland, USA) using a 20 μl RT reaction mix following manufacturers' instructions.
GAS5 and miR-137 quantitation by RT-qPCR. Assessment of GAS5 was evaluated by the use of RT2 SYBR Green PCR kit (Qiagen, Maryland, USA). However, the miScript SYBR Green PCR kit (Qiagen, Valenica, CA, USA) was used in the assessment of miR-137 according to the kit instructions with the aid of Rotor gene Q System (Qiagen) on 20 μl reaction mixture.
The RefSeq Accession no. of GAS5 was NR_002578.2 and the catalog number of miR-137 was MS00003486. Assessment of GAS5 was done per the following cycling steps: 95 °C for 10 min, afterward, 40 cycles at 95 °C for 15 s and 60 °C for 60 s. On the other hand, assessment of miR-137 was performed according to the subsequent cycling steps: 95 °C for 30 min, subsequently, 40 cycles at 94 °C for 15 s, 55 °C for 30 s, and 70 °C for 30 s.
GAS5 expression values were normalized by the use of GAPDH as an internal control 44,45 . Though, miR-137 expression values were normalized using SNORD 68 as an internal reference gene. The GAPDH primer sequences were forward, 5′-CCC TTC ATT GAC CTC AAC TA-3′, and reverse 5′-TGG AAG ATG GTG ATG GGA TT-3′. The catalog number of SNORD 68 was MS00033712. Calculation of the fold change (FC) of GAS5 and miR-137 was done with the aid of the equation 2 −ΔΔCt46 . The FC of control individuals was assumed as 1.
Genotyping of rs2067079 and rs1625579. According to the instructions of the kits, genomic DNA was extracted from whole blood through a Qia-amplification DNA extraction kit (Qiagen, USA). To estimate the purity and quantity of DNA samples, NanoDrop (ND)-1000 spectrophotometer (NanoDrop Technologies, Inc. Wilmington, USA) was used. Predesigned TaqMan SNP genotyping assays (Applied Biosystems, Thermo Fisher Scientific, USA) were used to genotype both rs2067079 C/T [Assay ID: C_16166809_20] and rs1625579 G/T [Assay ID: C_8946584_20] following manufacturer's instructions. Rotor gene Q System (Qiagen) was used in RT-PCR analyses. Cycling steps were along these lines: denaturation for 10 min at 95 °C, 45 cycles for 15 s at 92 °C, and 60 °C for annealing and extension at 90 s. www.nature.com/scientificreports/ Statistical methods. The collected data were organized and statistically analyzed using SPSS software statistical computer package version 16 (SPSS, Inc., USA). For quantitative data, the mean, median, standard deviation, and range were calculated. The Kolmogorov-Smirnov test was performed as a test of normality. The Mann-Whitney U test or Kruskal-Wallis test was used to compare any 2 groups or 3 groups, respectively. Qualitative data were presented as numbers and percentages, and chi-square (× 2) was used as a test of significance. Spearman's correlation was used to test the association of quantitative variables. ROC curve was done to evaluate the diagnostic performance of the biomarkers and the best cut-off point of the biomarkers was determined, Odds ratio (OR) with 95% confidence was calculated for different forms of polymorphism to identify its association with disease. Multiple logistic regression analysis was done to detect the most significantly associated factors with CHBV disease. The homozygote genotype or the major allele detected among the control subjects was selected as the reference group. Significance was defined as p < 0.05.