IL-6 gene rs12700386 Polymorphism is Associated with Increased Risk of Knee Osteoarthritis Development in Chinese Han Population: A Case-Control Study

Background: There is an association between Interleukin-6 (IL-6) polymorphism and knee osteoarthritis (OA) risk. The case-control study aims at exploring how IL-6 rs12700386 polymorphism affects the knee OA risk in Chinese Han individuals. Methods: We extracted the DNA from 763 participants, thereinto, 352 were OA patients and 411 were healthy controls. The polymerase chain reaction restriction fragment length polymorphism (PCR-RFLP) assisted in genotyping the IL-6 gene polymorphism. The relative expression exhibited by IL-6 in blood samples of knee OA patients was determined via a quantitative reverse transcription PCR (qRT-PCR). Results: We found that IL-6 rs12700386 enhanced the knee OA susceptibility. Based on a subgroup analysis, the loci magnied the knee OA risk in smokers, drinkers, and subjects ≥ 55 years old or with BMI ≥ 25 kg/m 2 . The combination of smoking and drinking and rs12700386 genotype led to an increase in the knee OA risk, indicating an underlying interaction between gene and environment. Additionally, the rs12700386 was found to be related to increased IL-6 gene levels. Conclusion: These data indicate that rs12700386 polymorphism of IL-6 gene led to an increase in the knee OA risk specic to Chinese Han individuals.


Background
Osteoarthritis (OA) is related to the damage synovial joint structure and function clinically and pathologically [1]. Worldwide it is a representative joint disease and it is estimated to affect 10% of men and 18% of women > 60 years old [2]. Clinically, OA most frequently affects the knee joint [3]. The occurrence of OA is in uenced by obesity, smoking, joint damage, heredity and in ammation [4].It is a multifactorial disease with an important genetic component [5]. In the past decade, genome-wide association studies have discovered many new genetic risk sites OA [5].
IL-6 refers to a four-helical cytokine containing 184 amino acids. A lot of cell types can secrete IL-6 after proper stimulation in the process of infection, cancer or in ammation [6]. It can effectively regulate B and T cells responses and coordinate the activities of innate as well as adaptive immune systems [7]. IL-6 is an important regulator of bone homeostasis, it can trigger osteoclast differentiation and bone resorption [8]. Goldring MB et al. found that IL-6 production underwent an up-regulation by IL-1β together with matrix metalloproteinase, and the upregulated IL-6 weakened collagen-2 production, and meanwhile caused joint damage [9]. Additionally, Sakao K et.al found the positive correlation between increased IL-6 and mRNA expression and OA radiographic progression [10]. Thus, we assumed that IL-6 may be a pivotal candidate gene for OA susceptibility.
Polymorphisms of IL-6 gene, located at 7p21-24 chromosomal position, are associated with risks for OA. Nevertheless, the impact of IL-6 rs12700386 polymorphism on OA risk has never been explored in Chinese Han population. Therefore, a case-control study was conducted in the paper which contained 352 OA patients together with 411 healthy controls, aiming at investigating if IL-6 gene rs12700386 is related to knee OA risk and IL-6 gene levels in Chinese subjects.

Methods
Subjects 352 OA patients together with 411 healthy controls were recruited from the A liated Changzhou No.2 People's Hospital of Nanjing Medical University and the Second A liated Hospital of Jiaxing University, who were ethnic Han Chinese with no genetic relation. The Kellgren Lawrence scoring (KL scoring) system was used to diagnose each patient radiologically. The functional index of Lequesne was used to assess each patient's functional or symptomatic status. Visual Analog Scale (VAS) was employed for evaluating the pain. There were three inclusion criteria. 1) Symptoms and/or signs related to knee OA; 2) Radio-graphic abnormality (with K-L grade ≥ 2); and 3) No any other arthritis form. Demographic characteristics [gender, age, smoking, alcohol and body mass index (BMI)] and clinical characteristics of disease severity [Lequesne function index, visual analog scale (VAS), ESR and CRP] were obtained from the medical records. We selected healthy controls from patients in orthopedics clinics and general surgery in one hospital while collecting samples. Exclusion criteria: history of knee or hip replacement, application of corticosteroids or bi-immunosuppressors, Pakinson's disease, or sequelae of stroke. Table 1 lists the characteristics of all the subjects. Study had obtained the informed consent from all the subjects, and obtained the approval of the Ethics Committee of the hospital. The performance of experiment followed the tenets of the Helsinki Declaration.
A TIANamp blood DNA kit (Tiangen Biotech, Beijing, China) assisted in extracting blood from all patients and extracting genomic DNA from surrounding blood leukocytes as recommended by the manufacturer. The extracted samples were stored at -20°C until further use. The IL-6 rs12700386 polymorphism was genotyped with a polymerase chain reaction restriction fragment length polymorphism (PCR-RFLP) with sequence-speci c primers: 5'-GCGACAGGCCTCTCCAG TCT-3' (forward) and 5'-GCAGTCACACCGGCTAGGTC' (reverse). We selected about 5% of samples in a random manner to receive repeated assays, nding that the accordance rate reached 100%.
Statistical analyses SPSS 22.0 (SPSS Inc., USA) was applied to all the statistical analyses. The statistical graphs were applied by virtue of GraphPad Prism 5 (GraphPad Software, La Jolla, CA). A student's t-test or χ2 test assisted in assessing the difference between patients and controls in the mean and frequency distribution regarding the clinical and demographic characteristics. The observed number of genotypes was compared with a Chi-square (χ2) analysis, with that expected in accordance with Hardy-Weinberg equilibrium (HWE). An independent sample t-test or a oneway ANOVA assisted in testing the continuous variables existing in the normal distribution, with results denoted by mean ± standard deviation. We analyzed the allele and genotype distributions between patients and the controls. Strati ed analyses according to alcohol, smoking, sex, and age were conducted. The odds ratio (OR) and 95% con dence interval (CI) were calculated in a logistic regression to assess the correlation of IL-6 gene polymorphism with OA risk. A cross-over analysis was used to assess the effects of the interactions between environmental factors (smoking or drinking) and genetic factors for the OA risk. A P < 0.05 implied a signi cant level.

Results
Characteristics of the study population Table 1 lists the information of all individuals in detail. The patients and controls were averagely aged between 61.39 to 61.03 years old, respectively. The mean BMI was 24.61 and 24.58 kg/m2 for the patients and controls respectively. No signi cant between-group differences were identi ed for age, BMI, or sex. In addition, the distributions of smokers and drinkers in the cases were basically the same as the control group. Table 1 lists the relevant indexes, including affected leg, CRP, ESR, VAS, Lequensene's index, as well as K-L grade.
Relation of IL-6 rs12700386 polymorphism to OA risk The allele and genotype distribution of IL-6 rs12700386 polymorphism is shown in Table 2. The HWE test found no obvious deviation in genotypic frequency among the controls, indicating that the subjects are representative of the local population. The IL-6 rs12700386 led to an obvious increase in the OA risk as seen in dominant, homozygous and allelic models. A logistic regression analyses showed that CC genotype of rs12700386 signi cantly intensi ed the OA risk (CC vs. GG: OR and 95%CI, 2.01(1.04,3.89), P=0.045). (Table 2). An allele genetic analysis revealed that C allele was linked to a higher risk of OA (C vs. G: 1.38(1.08,1.75), P=0.010) ( Table 2). This signi cant association was also observed in the dominant models. Strati ed analyses of age, smoking, drinking status, and BMI revealed a signi cantly higher OA risk in drinkers, smokers, as well as those ≥55 years old or with BMI ≥25 kg/m2. (Table 3). Nevertheless, the rs12700386 polymorphism did not affect the OA risk with regard to abovementioned indexes (Table 5).
Based on the RT-PCR analysis, the CC genotype exhibited an obviously higher IL-6 expression relative to the GG genotype (P < 0.001) ( Figure   1).

Cross-over analysis
The OR and 95% CI speci c to two combining exposure models (IL-6 gene variants and smoking or drinking) were calculated (Table 4). For rs12700386 polymorphism, a GG genotype more obviously affected the OA risk compared with the CC genotype. Besides, smoking had no impact on OA risk. Notwithstanding, smokers carrying the CC genotype were more easily to suffer from OA compared with non-smokers carrying the GG genotype (OR, 2.86, 95%CI, 1.25-6.56; P = 0.015). This indicated that CC genotype could strongly interact with smoking.
Similarly, IL-6 rs12700386 polymorphism or drinking could not affect the OA risk. Obviously, CC genotype of rs12700386 polymorphism and drinking exerted a combined effect on OA, demonstrating that CC genotype could strongly interact with drinking.

Discussion
In this study, we found that IL-6 rs12700386 polymorphism led to an increased OA risk, especially among drinkers, smokers, and those ≥55 years old or with BMI ≥25 kg/m 2 . The interactions between IL-6 rs12700386 polymorphism and smoking and drinking contributed to an increased risk for knee OA. In addition, genotypes of rs12700386 polymorphism were associated with increased IL-6 gene levels. Nevertheless, the impact of rs12700386 on OA patients regarding the clinical parameters remains unknown.
OA is characterized by destruction of cartilage, remodeling of subchondral bone and synovial membrane in ammation, which actively affects the disease progression [11]. As demonstrated by studies, tibiofemoral cartilage injury progression can be clearly affected by reactive or in ammatory synovium [12]. Proin ammatory cytokines like IL-6, TNF, can dramatically mediate metabolic disorders as well as enhance the catabolism regarding OA joint tissue [11]. The synovial uid and serum of OA patients also presented signi cantly increased IL-6 and sIL-6R levels [13]. On the basis of a clinical trial on OA patients, if IL-6 and CRP held a high baseline level, the cartilage loss risk would increase [14].
Increase in circulating IL-6 level and high BMI could lead to radiographic keen OA risk [15]. Furthermore, IL-6 de ciency in mice has been found to lead to reduced knee arthritis cell number as well as collagen-induced arthritis response [16]. However, how IL-6 exactly affect the OA is of Furthermore, based on strati ed analyses, knee OA risk increased in smokers, drinkers, and subjects ≥55 years old or with BMI ≥25 kg/m2. Thus, the cross-over analysis was used to estimate the combined effects of IL-6 gene polymorphism and smoking and drinking on knee OA risk. As expected, the combined effects of rs12700386 polymorphism and smoking and drinking conferred susceptibility to OA. In addition, we evaluated the associations between IL-6 gene polymorphism and clinical characteristic of OA. However, no evidence was found to prove that rs12700386 is associated with the clinical parameters of OA patients. To explore the underlying mechanisms, the qRT-PCR of IL-6 gene showed the rs12700386 contributed to increase IL-6 gene levels. Thus, our study indicated that the IL-6 gene polymorphism could affect the IL-6 gene level and the knee OA risk.
There are also some shortcomings in the study. 1) it had a moderate sample size, incapable of accurately exploring how IL-6 gene      Figure 1 Expression levels of IL-6 in 3 different genotypes.