MMP8 and MMP9 gene polymorphisms were associated with breast cancer risk in a Chinese Han population

Matrix metalloproteinases (MMPs) are a group of zinc-dependent endopeptidases that can breakdown almost all extracellular matrix components. MMP8 and MMP9 have been shown to be associated with breast cancer (BC) risk in European and American populations. However, few studies have focused on the polymorphisms of MMP8 and MMP9 in Chinese Han BC patients. We investigated nine single nucleotide polymorphisms (SNPs) in 571 BC cases and 578 controls to evaluating their association with risk of BC. The frequency of the “A” allele of rs3787268 was significantly lower in BC cases than in controls (P = 0.025). In the genetic model analysis, the minor allele “T” of rs11225394 in MMP8 was associated with increased risk of BC under the recessive model (P = 0.019), and the minor allele “A” of rs3787268 was associated with decreased risk of BC under the dominant model (P = 0.014). Additionally, the haplotype “AGTCA” constructed by rs3740938, rs2012390, rs1940475, rs11225394, and rs11225395 and the haplotype “CCG” constructed by rs3918249, rs3918254 and rs3787268 were associated with increased risk of BC (P < 0.05). Our data showed that polymorphisms of MMP8 and MMP9 may be associated with BC risk in the Chinese Han population.

MMP8 and MMP9 in Chinese Han BC patients. We intended to identify more SNPs in MMP8 and MMP9 gene that may be associated with BC risk.
The SNPs rs3740938, rs2012390 and rs11225394 were reported to be associated with the risk of osteonecrosis of the femoral head in the Chinese Han population 20,21 . Rs1940475 may take part in modifying the host response to inflammatory stimuli 22 . Rs11225395 and rs2274755 were considered as candidate SNPs in previous association studies on bladder cancer and glaucoma 23,24 . In addition, the haplotype constructed by rs3918249, rs3918254 and rs3787268 in MMP-9 were associated with primary angle-closure glaucoma in a Chinese Han population 25 . In this case-control study, we selected and genotyped the above nine SNPs in MMP8 and MMP9 to evaluate their association with risk of BC.

Results
A total of 571 BC patients and 578 healthy controls were recruited in the study. The ages of the patient and control groups are described in Table 1. The mean age of the participants was 50.91 years in the case group and 50.21 years in the control group. There was no significant difference in the distribution of age between BC patients and healthy controls (P > 0.05). The basic information of the MMP8 and MMP9 polymorphisms (rs3740938, rs2012390, rs1940475, rs11225394, rs11225395, rs3918249, rs2274755, rs3918254 and rs3787268) is shown in Table 2, including gene, band, position, role, alleles and minor allele frequency (MAF). One SNP (rs2274755) was excluded due to significant deviation from Hardy-Weinberg equilibrium (P < 0.05). By comparing the difference of allele frequencies between two groups, we found that the frequency of the "A" allele of rs3787268 was significantly lower in BC cases than in controls (32.2% versus 37.7%), which suggested that the "A" allele of rs3787268 may be associated with a decreased risk of BC [odds ratios (ORs) = 0.822, and 95% confidence intervals (CIs): 0.692-0.975, P = 0.025].
Next, we investigated the association between each SNP and risk of BC based on different genetic models (Table 3). When the sum of AIC and BIC were at the minimum, the model was considered as the best model. Under the best model-recessive model, the minor allele "T" of rs11225394 in MMP8 was associated with an increased risk of BC (OR = 3.94, 95% CI: 1.10-14.07, P = 0.019). Under the best model-dominant model, the minor allele "A" of rs3787268 was associated with decreased risk of BC (OR = 0.74, 95% CI: 0.59-0.94, P = 0.014).
The association of MMP8 and MMP9 haplotypes with the risk of BC was analysed. Figures 1 and 2 show the linkage disequilibrium (LD) block in MMP8 and MMP9. The association between different haplotypes and BC risk is shown in Table 4. The haplotype "AGTCA" constructed by rs3740938, rs2012390, rs1940475, rs11225394, and rs11225395 was associated with an increased risk of BC after the adjustment (OR = 1.23; 95% CI = 1.00-1.51; P = 0.048). The haplotype "CCG" constructed by rs3918249, rs3918254 and rs3787268 was also associated with an increased risk of BC after the adjustment (OR = 1.37; 95% CI = 1.05-1.77; P = 0.019).
Finally, we summarized the OR and p-values of candidate SNP from previous studies. The results are listed in Table 5.

Discussion
In this study, we investigated the associations between nine candidate SNPs in MMP8 and MMP9 genes and BC risk in a Chinese Han population. We found that rs11225394 in MMP8 and rs3787268 in MMP9 are significantly associated with BC risk. We also found two haplotypes were associated with increased risk of BC: haplotype "AGTCA" constructed by rs3740938, rs2012390, rs1940475, rs11225394, and rs11225395 in MMP8, and haplotype "CCG" constructed by rs3918249, rs3918254 and rs3787268 in MMP9. Our results suggest that the polymorphisms of MMP8 and MMP9 may play an important role in the risk of BC in the Chinese Han population.
MMPs are a family of endopeptidases involved in the degradation of extracellular matrix and basement membrane barriers. As such, they contribute when tumour cells separate from adjacent normal tissues 26,27 . Several studies have revealed that MMPs are associated with the invasion and metastasis of tumour cells. MMP8 and MMP9 are members of the MMPs family. Plasma MMP8 levels were lower in BC patients than in healthy control individuals, and MMP8 was found to protect against lymph node metastasis in BC patients 28 . Differential expression of MMP9 was found in BC cells with different degrees of cellular differentiation 29 . We identified two SNPs in MMP8 and MMP9 that were associated with risk of BC, which further corroborates the association between MMPs and BC risk.
In the present study, we investigated nine SNPs in the MMP8 and MMP9 genes. Among these SNPs, rs3740938, rs2012390, rs11225394, rs2274755 and rs3918254 were identified that have been associated with risk of osteonecrosis of the femoral head in the Chinese Han population 20,21,30 . We report for the first time that rs11225394 was associated with BC risk. This association needs to be confirmed in a further study with a larger sample size. Rs1940475 was found to have association with community-acquired pneumonia-associated sepsis and inflammatory response in Caucasians 22,31 . Rs11225395 was associated with bladder cancer risk in a Polish population 23 . Rs3918249 was demonstrated to be associated with risk for primary angle-closure glaucoma in Chinese Han population 25 and risk for asthma in Mexican paediatric patients 32 . Rs3787268 was also correlated with BC prognosis in previous studies; the "GA" and "AA" genotypes of rs3787268 were significantly associated with 1.52-fold risk of BC in Native American women 33 . However, in our study, we found that the minor allele "A" of rs3787268 was associated with decreased risk of BC in Chinese women, which is inconsistent with previous result. This inconsistency may be due to the limited sample size and different populations.
Our study has some intrinsic limitations. First, the sample size was relatively small, so the results identified here just provide some pilot data for continued in-depth studies on BC. Second, the occurrence of BC was also influenced by BMI, menarche, marital status, parity, and so on. We failed to collect these specific data, which is a weakness of the study. Additionally, although we identified significant SNPs and haplotypes with risk of BC, this is still not enough to explain the molecular mechanism between MMP and the onset and development of BC.
In conclusion, the present study showed that polymorphisms of MMP8 and MMP9 are associated with risk of BC in a Chinese Han population. Further studies will focus on two directions. One is to demonstrate the association between MMP and BC risk in larger sample sizes and different populations; the other is to investigate the mechanisms or pathways by which MMP influence BC risk. SNP selection and genotyping. Nine SNPs (rs3740938, rs2012390, rs1940475, rs11225394, rs11225395, rs3918249, rs2274755, rs3918254 and rs3787268) in MMP8 and MMP9 were selected from previous association studies. DNA extraction and concentrations were done as previously described 34 . SNP genotyping was performed by the Sequenom MassARRAY RS1000, and the data analyses were completed by Sequenom Type 4.0. The polymerase chain reaction (PCR) primers used for the study are listed in Table 6.
Statistical Analyses. All of the statistical analyses were performed with Microsoft Excel (Microsoft Corporation, Redmond, WA, USA) and the SPSS 21.0 statistical package (SPSS, Chicago, IL, USA). SNP allele frequencies in the control subjects were tested for departure from Hardy-Weinberg Equilibrium (HWE) before analysis. Differences in the distributions of allele frequencies between the cases and controls were evaluated using   Table 6. Primers used in this study. UEP SEQ, unextended mini-sequencing primer.
chi-square tests. Four models (co-dominant, dominant, recessive, and log-additive) were used to assess the association between each genotype and the BC risk 35 . Akaike's Information criterion (AIC) and Bayesian Information criterion (BIC) were used to select the best model for each SNP. ORs and 95%CIs were calculated with and without adjustment for age. All p values presented in this study were two sided, and p = 0.05 was considered the cutoff for statistical significance. Haploview software version 4.2 was used to analyse the association between haplotypes and BC. Linkage disequilibrium (LD) analysis was performed using genotype data from all the subjects. Statistical significance was established when p < 0.05.
Approval. All of the participants provided written informed consent. The Human Research Committee for Approval of Research Involving Human Subjects, the First Affiliated Hospital of Xi'an Jiaotong University, approved the use of human blood samples in this study. All methods were performed in accordance with the relevant guidelines and regulations.

Data Availability
The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.