Association of MBOAT7-TMC4 rs641738 with the risk of hepatocellular carcinoma and persistent infection of hepatitis B virus: a case-control study and meta-analysis

Background: Recent studies reported that a hot genetic variant, rs641738 within the membrane bound O-acyltransferase domain containing 7(MBOAT7) and transmembrane channel-like 4 (TMC4) , was associated with several liver diseases. However, results are still conicting. We conducted this study to explore the role of MBOAT7-TMC4 rs641738 in the risk of hepatocellular carcinoma (HCC) and persistent hepatitis B virus (HBV) infection. Methods: We rst performed a case-control study by including 779 HCC cases and 1412 cancer-free controls. Controls are consisted of 678 HBV persistent carriers and 734 spontaneously recovered subjects. Rs641738 was genotyped by MassARRAY platform. Results were analyzed by multivariate logistic regression analysis under ve genetic models. Second, we conducted a systematic review and meta-analysis to further explore the role of this variant in HCC risk. Results: Results suggested no association between MBOAT7-TMC4 rs641738 and HCC risk in most genetic models (All P > 0.05), although a marginally signicant association was observed in TT vs. CC ( P = 0.037) and recessive model ( P = 0.044). Further meta-analysis including 2135 HCC cases and 4388 controls supported that this variant was not related to HCC risk, even in the TT vs. CC and recessive models. Besides, we identied that this variant also had no inuence on persistent HBV infection. Conclusion: Our work highlights that MBOAT7-TMC4 rs641738 is not associated with the risk of HCC or persistent HBV infection. This study provides some clues to identify the “truth” of potential disease-related genetic factors in the post-genome era.


Background
Hepatocellular carcinoma (HCC) is a common cancer with rapid progress and high mortality in the world, especially in China [1]. As the fourth common cancer in China, it was estimated to have 466100 new cases in 2015 [2]. Environmental factors such as infection of hepatitis B virus (HBV) are con rmed to be a crucial pathogenesis of HCC [3]. Increasing evidence has revealed that genetic factors also play an important role in the development of HCC [4].
HCC is a complicated disease with high genetic heterogeneity. Candidate gene researches and genome-wide association studies (GWASs) have found hundreds of genes and loci associated with HCC risk. How to identify the "truth" of these genetic factors is becoming an urgent issue in the post-genome era. In 2015, a GWAS study reported a genetic variant near the membrane bound O-acyltransferase domain containing 7(MBOAT7) gene, rs641738, C>T, increasing the risk of alcohol-related cirrhosis [5]. Later it is also reported to be located in the transmembrane channel-like 4 (TMC4) gene. In 2016, Thabet demonstrated that MBOAT7-TMC4 rs641738 could be a risk factor of hepatic in ammation and liver brosis [6]. Subsequent studies further investigated the association between this variant and HCC risk, but led to a con icting conclusion [7][8][9].
Here, we rst conducted a case-control study including 779 HCC cases and 1412 cancer-free controls, aiming to explore the effects of MBOAT7-TMC4 rs641738 on HCC risk in a Chinese population. Then we performed a systematic review and meta-analysis to further explore the role of this variant and validate our results.

Study Subjects
This case-control study included 779 HCC cases and 1412 cancer-free controls. Cases were pathologically con rmed and enrolled between January 2014 and June 2016 at Tongji Hopital of Huazhong University of Science and Technology (HUST), central China. Controls are consisted of 678 hepatitis B virus (HBV) persistent carriers and 734 spontaneously recovered subjects, who were recruited from a health screening in the same hospital during the same period as the cases were included. HBV persistent carriers were people who were positive for both of HBsAg and HBcAb, but negative for anti-HCV. Spontaneously recovered subjects were people who were negative for both of HBsAg and anti-HCV, but positive for both of HBcAb and HBsAb. All subjects were unrelated Han Chinese from Wuhan and the surrounding regions. Cases and controls were frequency-matched for sex and age. At recruitment, a 2-ml peripheral blood sample and a written informed consent were collected from each subject, and demographic information (i.e., sex, age, smoking and drinking status) were also obtained by questionnaire. De nitions of smoking and drinking status have been detailed previously [10,11]. This study was approved by the institutional ethics committee of Tongji Hospital, Tongji Medical College of HUST.

Serological Testing
Enzyme-linked-immunosorbent assay (ELISA) was applied to detect serum HBsAg, HBsAb, HBcAb and anti-HCV (IMX; Abbott Diagnostics, USA). There were three positive controls, two negative controls and one blank control in each reaction plate. About 5% of the samples were randomly chosen for repetition, and results were 100% concordant.

Genotyping
Genomic DNA was extracted from leucocyte pellets of 2-ml peripheral blood by the coagulated blood DNA mini-extraction kit (DP6101, BioTeke Corporation, China). MBOAT7-TMC4 rs641738 was genotyped by the Sequenom MassARRAY iPLEX Platform (SEQUENOM, CA, USA). All assays were conducted using 384-well plate with positive and negative controls for each plate, without the information of the disease status of samples. We randomly selected 5% of the samples as duplicate sets, and got a 100% concordance rate. The average call rate of this variant was 99.7%.

Statistical Analysis
A goodness-of-t χ 2 test was applied to assess the Hardy-Weinberg equilibrium (HWE) in controls. Differences between cases and controls are examined by independent t test, ANONVA analysis or Pearson's χ 2 test according to the category of variables. The risk of HCC or HBV persistent infection was estimated by odds ratio (OR) and 95% con dence interval (CI), using a multivariate logistic regression analysis after adjusting for age, sex, smoking and drinking status. A two-tailed P value < 0.05 was considered statistically signi cant. All the analyses above were performed by IBM SPSS Statistics version 20.0 (Chicago, IL, USA). In addition, we calculated the statistical power of this study by Power V3.0 [12]. Based on our sample size to detect an OR of 1.30, a power of 0.95 was estimated for MBOAT7-TMC4 rs641738.

Meta-analysis
This meta-analysis was performed according to the guidelines of the Preferred Reporting Items for Systematic Review and Meta-analyses (PRISMA) [13].

Literature search, Study Selection and Data Extraction
A comprehensive literature search was conducted through PubMed, Embase and Web of Science up to August, 2019 without any restrictions. The search items included "MBOAT7",, "TMC4",, "rs641738", and "hepatocellular carcinoma". References from identi ed publications were also reviewed for obtaining potential relevant studies.
The inclusion criteria was as follows: (1) a case-control or cohort study to evaluate the association between MBOAT7-TMC4 rs641738 and HCC risk, and (2) providing ORs and 95%CIs, or allele frequency and/or genotypes of this variant. A study was excluded if it met one of the following criteria: (1) review, meta-analysis, comment or conference abstract; (2) insu cient data to estimate OR and 95%CI; (3) a deviation from HWE in controls; and (4) studies with overlapped data. If a study contained overlapped data with another, we kept the one with larger sample size.
Data was extracted from each included study as follows: rst author, publication year, country, ethnicity, gender, age, sample size, genotype distribution, genotyping method, adjustment, and HWE information. Two authors independently conducted the work above. Disagreement was resolved by discussion. The quality of each included study was evaluated by the Newcastle-Ottawa scale (NOS) [14]. Quality evaluation was not an exclusion criterion for eligible studies (Additional le 1).

Statistical Analysis for Meta-analysis
We used ORs and 95%CIs to estimate the association between MBOAT7-TMC4 rs641738 and HCC risk. Multivariate-adjusted OR and 95%CI was preferentially extracted if available, otherwise unadjusted OR and 95%CI was calculated instead. Between-study heterogeneity was examined by Q test and I 2 statistic. If P < 0.10 or I 2 > 50 %, it was considered a signi cant heterogeneity and applied the random-effects model; otherwise the xed-effects model was used. To evaluate the quality, we also performed the sensitivity analysis and publication test [15,16]. The meta-analysis was conducted via Stata 12.0 software (College Station, TX, USA).

Subject Characteristics
A total of 779 HCC cases, 678 HBV persistent carriers and 734 spontaneously recovered subjects were included in this study. Demographic characteristics are summarized in Table 1. In these three groups, the male proportion of the three groups was 70.6%, 68.7% and 68.1%, respectively; and the mean age was 53.20 ± 12.49, 52.27 ± 11.45 and 52.28 ± 12.89 years, respectively. There was no signi cant difference among the three groups on sex (P = 0.551), age (P = 0.240), smoking (P = 0.071) and drinking status (P = 0.091). Genotypes of MBOAT7-TMC4 rs641738 in all groups did not deviate from HWE (HCC: P = 0.485; HBV persistent carriers: P = 0.133; spontaneously recovered subjects: P = 0.769).
Association of MBOAT7-TMC4 rs641738 with Risk of HCC and Persistent HBV Infection We performed four comparisons under ve genetic models and showed the results in Table 2. In the comparison of HCC cases vs. all controls, we found that after adjusting for sex, age, smoking and drinking status, MBOAT7-TMC4 rs641738 did not confer any increased risk of HCC risk in dominant, additive or allelic model (All P > 0.05). Although this variant seemed a risk factor of HCC in two models, the results were marginally signi cant (TT vs. CC: P = 0.037; Recessive: P = 0.044). Similar results were also observed in the further comparison of HCC vs. HBV persistent carriers. When we set spontaneously recovered subjects as controls, we observed that MBOAT7-TMC4 rs641738 was not associated with HCC risk (All P > 0.10). To explore whether this variant in uences the risk of persistent HBV infection, we conducted a further comparison between HBV persistent carriers and spontaneously recovered subjects. Results suggested that this variant was also unrelated to persistent infection or clearance of HBV (All P > 0.30).

Meta-analysis of the Association between MBOAT7-TMC4 rs641738 and HCC Risk
In addition, we conducted a systematic review and meta-analysis to further examine our results. Initially, we got 87 records by searching PubMed, Embase and Web of Science, and excluded 79 records by title and abstract review. After assessing the full-text articles, we further excluded ve studies due to deviation from HWE (n = 1) [9], insu cient data to estimate ORs (n = 1) [17], or having overlapped data (n = 3) [18][19][20]. Finally, we included three eligible articles [6][7][8] and this study for meta-analysis (Additional le 2).
A total of 2135 HCC cases and 4388 controls were included. Characteristics are detailed in Table 3. We rst applied a meta-analysis under the allelic model ( Figure 1). Pooled results demonstrated that MBOAT7-TMC4 rs641738 was not associated with HCC risk (OR = 1.10, 95%CI = 0.99-1.23, P hetergeneity = 0.453, I 2 = 0%), which was stable suggested by the sensitivity analysis (Additional le 3). Egger's or Begg's test did not suggest any publication bias. Considering that marginally signi cant results were observed in our study, we further applied metaanalyses in the TT vs. CC and recessive models. As shown in Table 4, pooled results in these models also consistently and stably showed that there was no association between MBOAT7-TMC4 rs641738 and HCC risk.

Discussion
In this study, we conducted a case-control study including 799 HCC cases and 1412 controls to investigate the effect of MBOAT7-TMC4 rs641738 on HCC risk. Although this variant showed a marginally signi cant association in TT vs. CC (P = 0.037) and recessive model (P = 0.044), the other genetic models all demonstrated non-signi cant results. We further performed a systematic review and meta-analysis to explore the role of MBOAT7-TMC4 rs641738 in the susceptibility to HCC. A total of 2135 HCC cases and 4388 controls were included. Results consistently suggested that MBOAT7-TMC4 rs641738 was not associated with HCC risk, even in the TT vs. CC and recessive models. In addition, our study revealed that this variant was also not related to persistent HBV infection in all genetic models. To the best of our knowledge, it is the rst study to investigate the association between MBOAT7-TMC4 rs641738 and the risk of HCC and persistent HBV infection in Asians.
The rs641738 variant is mapped to the 500-bp downstream of MBOAT7 and likewise located at exon 1 of the TMC4 gene. Therefore, previous studies refer to it as MBOAT7 or MBOAT7-TMC4 variant. This variant was rst identi ed as a risk locus for alcohol-related cirrhosis [5], then reported to be associated with nonalcoholic fatty liver disease (NAFLD)/nonalcoholic steatohepatitis (NASH) [21] and liver brosis [6]. In 2016, Thabet rst investigated the association of MBOAT7-TMC4 rs641738 with HCC risk, but did not fond a signi cant result [6]. Subsequently, Donati contradicted the conclusion and showed that this variant predisposed to HCC in NAFLD subjects [9]. However, the signi cant association was not con rmed by later studies [6][7][8]. Even in the NAFLD cohort, MBOAT7-TMC4 rs641738 did not show any association with the risk of Matteoni type 4 or NASH-HCC [17]. Additionally, a recent meta-analysis revealed that this variant was not related to the susceptibility of NAFLD [22]. In view of a small sample size of the previous studies, we performed this case-control and meta-analysis, and nally con rmed that MBOAT7-TMC4 rs641738 was not associated with HCC risk.
Persistent HBV infection is a crucial risk factors contributing to HCC in China. However, studies about the effect of MBOAT7-TMC4 rs641738 on HBV infection is limited up to date. Here, we also explored the association between MBOAT7-TMC4 rs641738 and persistent HBV infection. Results exhibited that this variant was not related to spontaneous clearance of HBV, which was in concordance with a previous report from a Moroccan cohort [23]. Interestingly, Thabet reported that MBOAT7-TMC4 rs641738 in uenced hepatic in ammation and brosis in patients with persistent HBV infection [24]. A possible explanation is that this variant's effect on liver disease is not associated with HBV, or it is not the causal variant.
Nowadays, the role of MBOAT7-TMC4 rs641738 in liver diseases is still con icting. MBOAT7 encodes a lysophosphatidylinositol acyltransferase involved in phospholipid metabolism. TMC4 encodes a member protein of the calcium-dependent chloride channels, which are widely expressed in epithelia of many tissues, and exist in both plasma membrane and intracellular compartments [25]. Some studies showed that rs641738 reduced mRNA and hepatic MBOAT7 expression, but some studies refuted it [9,24,26]. Similar contradictions are also reported in brosis and NAFLD [23,24,26]. In fact, the exact biological mechanisms of MBOAT7 and TMC4 on liver diseases remain elusive. So whether and how MBOAT7-TMC4 rs641738 in uences liver diseases is doubtful. Nonetheless, our study and a recent meta-analysis support that this variant is not associated with the risk of HCC and NAFLD, respectively.
There are some limitations in our study. First, this is a hospital-based case-control study, which might be in uenced by selection bias. However, our multivariate analysis, power calculation and subsequent meta-analysis ensured accuracy of our conclusion. Second, our metaanalysis only included four studies due to limited number of relevant studies, which were not performed further strati ed analyses according to factors including age, sex, smoking or other lifestyle habits. Even so, the results of meta-analysis were reliable because of robustness, no publication bias and low heterogeneity.

Conclusion
In summary, our case-control study and meta-analysis highlight that MBOAT7-TMC4 rs641738 is not associated with HCC risk. Besides, our work also suggests no relationship of this variant with persistent HBV infection in a Chinese population. Further studies are required to validate our results. Availability of data and materials All data generated or analyzed during this study are available from the corresponding author upon reasonable request. But those clinical data are not available for authors have an ethical and legal responsibility to respect participant's rights to privacy and to protect their identity.