Abstract
Recently, we reported on the associations of seven single-nucleotide polymorphisms (SNPs) in the promoter region of MMP1 gene with susceptibility to cutaneous melanoma (CM). Considering the reported correlation between MMP1 expression and melanoma progression, we hypothesized that these promoter SNPs might affect CM progression and prognosis. In this study, we examined the associations of the seven SNPs with overall survival as well as six clinicopathologic factors in 754 patients with CM. After adjustment for 11 covariates, we observed significant association of the SNP −422A > T (rs475007) with ulceration status (P = 0.012), primary tumor thickness (P = 0.040), and anatomic site (P = 0.030). We also observed significant association of the SNP −755T > G (rs498186) with ulceration status (P = 0.038) and anatomic site (P = 0.003). Two SNPs −839G > A and −519A > G were marginally associated with primary tumor thickness, ulceration status, and anatomic site. Furthermore, the frequency of haplotype 2G-G-G-A-A-G-T was higher in patients with ulceration (odds ratio [OR] = 2.18, 95% confidence interval [CI] 1.08–4.40, P = 0.030) than that in those without ulceration. However, we did not find significant associations of these SNPs with overall survival and other clinical factors. Since primary tumor thickness and ulceration status are two important indicators of tumor progression and have significant associations with melanoma prognosis, our results suggested that these promoter SNPs in MMP1 might have potential effects on melanoma progression and prognosis by influencing related clinical factors.
Keywords: Genotypes, Melanoma, Survival, Tumor characters, Molecular epidemiology
Introduction
Melanoma is the most aggressive form of skin cancer, and its incidence is increasing [1]. Melanoma progresses in two phases: the radial and vertical growth phases. The cure rate for melanoma in the radial growth phase is nearly 100% after surgery [2]. However, melanoma in the vertical growth phase is more aggressive and able to metastasize. Because few treatments are effective against advanced melanoma, the prognosis for it is poor. The one year survival rates were 33%–62% in patients with distant metastatic melanoma [3]. Therefore, identification of predictive markers that distinguish high risk melanoma cases is of great clinical importance.
Melanoma progression is a multiple-step process that involves many genes. Degradation of the extracellular matrix (ECM) and basement membrane barriers is a key step in this process. Researchers have shown that the matrix metalloproteinases (MMPs), a family of calcium- and zinc-dependent endopeptidases, are critical for ECM degradation [4]. To date, studies have identified at least 24 MMPs in humans [5]. They are classified into four groups based on their substrate specificity and cellular location: collagenases, gelatinases, stromelysins, and membrane-associated MMPs. MMP1, one of the most widely expressed interstitial collagenases, plays a significant role in ECM degradation. The MMP1 gene is located on 11q22 and expressed in various cells [6]. Previous studies have shown that increased expression of MMP1 modulates melanoma development, progression, and metastasis [7–10]. Specifically, patients with melanoma having increased expression of MMP1 have exhibited worse outcomes than patients having lower expression of it [11].
It has been reported that the promoter region of the MMP1 gene contains binding sites for various transcription factors, such as AP-1, AP-2, and Ets/PEA-3, and response elements to glucocorticoids, retinoic acid, and cyclic adenosine monophosphate [12, 13]. Polymorphisms in this region may regulate MMP1 expression by influencing the binding of these transcription factors. The polymorphism of insertion/deletion of guanine at position −1607 has been found to have a role in the regulation of MMP1 expression by creating an Ets binding site [14]. Groups have widely studied the association of this variation with risk of development or progression of various cancers [15–18]. Specifically, −1607 1G/2G and six other SNPs (−839G > A, −755T > G, −519A > G, −422A > T, −340A > G, and −320T > C) in the MMP1 promoter region exhibited haplotype effects on MMP1 promoter activity [19]. We recently reported the significant associations of several of these SNPs in the MMP1 promoter region with cutaneous melanoma (CM) susceptibility [20]. However, until now, little is known about the correlation between these SNPs and melanoma progression and prognosis [21]. In this study, we investigated the association of these seven promoter SNPs with CM progression and prognosis in 754 CM patients with available genotyping and clinical data.
Materials and methods
Study subjects
Recruitment of the study subjects was described previously [20]. Briefly, 754 patients with newly diagnosed, histologically confirmed, and untreated CM were recruited consecutively at The University of Texas MD Anderson Cancer Center from April 1994 to April 2008. A onetime blood sample (30 ml) was drawn from each study participant. Informed consent to participate in the study was obtained from the patients, and the study was approved by the MD Anderson Institutional Review Board.
Patients’ clinicopathological information, including disease stage at diagnosis (based on the 2001 American Joint Committee on Cancer [AJCC] staging system), disease progression, and survival duration, was obtained. The main clinical factors used in this study included primary tumor thickness (Breslow), ulceration status, Clark level, and anatomic site, as well as sentinel lymph node status and stage [22]. In the analysis, tumor thicknesses were divided into three categories: thin (≤1 mm), intermediate (1.01–4.00 mm), and thick (>4.00 mm) [3]. The five Clark levels of invasion were grouped into three categories: I, II–III, and IV–V. Considering melanomas on the extremities have better prognosis than those on the head, neck, or trunk (axial) [23], we treated the primary CM anatomic site as a dichotomous variable: extremity (e.g., arm, hand, foot, leg) and axial (e.g., face, forehead, ear, cheek, nose, neck, eye, scalp, trunk, buttock, groin).
SNP selection and genotyping
The SNP selection method we used was described previously [20]. Briefly, seven reported common (minor allele frequency ≥ 5%) SNPs in the MMP1 promoter region (−1607 1G/2G [rs1799750], −839 G > A [rs473509], −755 T > G [rs498186], −519 A > G [rs1144393], −422 A > T [rs475007], −340 A > G [rs514921], and −320 T > C [rs494379]) were selected for genotyping. These SNPs were not in linkage disequilibrium (LD) as reported previously [15, 20].
The TaqMan genotyping method (Applied Biosystems, Foster City, CA, USA) was used as described previously [20] with genomic DNA extracted from whole blood. Each 96-well plate had one positive control, one negative control and 5 duplicate samples.
Statistical analysis
The chi-square test and unconditional logistic regression analysis were used to evaluate the correlations between SNPs and the six clinicopathological factors for CM. The raw and adjusted odds ratios (ORs) and 95% confidence intervals (CIs) were calculated without/with adjustment for the 11 covariates including age, sex, family history of any cancer in first-degree relatives, and the sunlight exposure-related factors including moles, dysplastic nevi, hair color, eye color, skin color, tanning ability, freckling in the sun as a child and history of sunburn. In the survival analysis, deaths from any causes were treated as events, and living subjects were censored at the time of last contact. The overall survival was calculated from the date of diagnosis of CM till death or censorship. The association between each SNP and overall survival duration was estimated using the Kaplan-Meier method and log-rank test. The hazard ratio (HR) and 95% CI were assessed in a Cox regression analysis without/with adjustment for sex, age, stage, primary tumor thickness, ulceration status, Clark level, and anatomic site. Because sentinel lymph node biopsy was not performed in 248 patients (32.9%) with thin melanoma or in situ melanoma, this clinical variant was not included in the Cox model. Haplotype analysis was carried out using SAS (Statistical Analysis System)/Genetics Module.
All statistical analyses were performed using the SAS software (version 9.2; SAS Institute Inc., Cary, NC). Two-sided P values less than 0.05 were considered statistically significant.
Results
The clinical and histological characteristics of the study population are shown in Table 1. All of the subjects were non-Hispanic white individuals. The final study population consisted of 490 male and 264 female patients. The age at diagnosis ranged from 21 to 84 years old (median = 54 years). At diagnosis, 79 patients had melanoma in situ, 567 had stage I–II CM (415 at stage I, 43 at stage I/II, and 109 at stage II), and 108 had stage III–IV CM (106 at stage III and 2 at stage IV).
Table 1.
Primary clinical and histological characteristics of the 754 study patients.
| Variable | # Patients (%) |
|---|---|
| Age (years) | |
| Median | 54 |
| Range | 21–84 |
| Primary tumor thickness | |
| Thin (≤1 mm) | 317 (42.0) |
| Intermediate (1.01–4.00 mm) | 271 (35.9) |
| Thick (>4 mm) | 37 (4.9) |
| In situ only | 79 (10.5) |
| Unknown | 50 (6.6) |
| Presence of ulceration | |
| No | 503 (66.7) |
| Yes | 80 (10.6) |
| In situ only | 79 (10.5) |
| Unknown | 92 (12.2) |
| Sentinel lymph node status | |
| Negative | 433 (57.4) |
| Positive | 73 (9.7) |
| No SLN biopsy performeda | 248 (32.9) |
| Sex | |
| Female | 264 (35.0) |
| Male | 490 (65.0) |
| Clark level | |
| I | 78 (10.3) |
| II–III | 339 (45.0) |
| IV–V | 285 (37.8) |
| Unknown | 52 (6.9) |
| Primary CM anatomic site | |
| Axial | 447 (59.3) |
| Extremity | 300 (39.8) |
| Unknown | 7 (0.9) |
| Stage at diagnosis | |
| In situ only | 79 (10.5) |
| I–II | 567 (75.2) |
| III–IV | 108 (14.3) |
Sentinel lymph node (SLN) biopsy was not performed for most patients with in situ or thin melanoma.
As shown in Table 2, even though none of the seven SNPs was significantly associated with primary tumor thickness in the univariate analysis, after adjustment for the 11 covariates, SNP −422A > T showed significant association with primary tumor thickness under the dominant genetic model (P = 0.04); and another SNP −839 G > A showed marginally association with primary tumor thickness under the codominant model (P = 0.062) (Table 2). In addition, more patients with thin tumors than with thick tumors were variant allele carriers of −839 G > A (GA or AA genotype) and −422A > T (AT or TT genotype).
Table 2.
Associations between SNPs in the MMP1 promoter region and primary tumor thickness.
| SNP | # Patients (%)
|
P-valuea | P-valueb | ||
|---|---|---|---|---|---|
| Thin (≤1 mm) | Intermediate (1.01–4.00 mm) | Thick (>4 mm) | |||
| −1607 1G/2G (rs1799750) | |||||
| 1G | 81 (25.8) | 79 (29.5) | 10 (27.8) | 0.398 | 0.761 |
| 1G/2G | 160 (51.0) | 134 (50.0) | 14 (38.9) | ||
| 2G/2G | 73 (23.2) | 55 (20.5) | 12 (33.3) | ||
| 1G/2G + 2G/2G | 233 (74.2) | 189 (70.5) | 26 (72.2) | 0.611 | 0.482 |
| −839 G > A (rs473509) | |||||
| GG | 103 (32.5) | 94 (35.1) | 18 (50.0) | 0.127 | 0.062 |
| GA | 170 (53.6) | 126 (47.0) | 13 (36.1) | ||
| AA | 44 (13.9) | 48 (17.9) | 5 (1.9) | ||
| GA + AA | 214 (67.5) | 174 (64.9) | 18 (50.0) | 0.110 | 0.068 |
| −755 T > G (rs498186) | |||||
| TT | 106 (33.8) | 82 (30.8) | 12 (33.3) | 0.423 | 0.721 |
| TG | 146 (46.5) | 133 (50.0) | 13 (36.1) | ||
| GG | 62 (19.8) | 51 (19.2) | 11 (30.6) | ||
| TG + GG | 208 (66.3) | 184 (69.2) | 24 (66.7) | 0.749 | 0.532 |
| −519 A > G (rs1144393) | |||||
| AA | 122 (38.9) | 107 (40.1) | 19 (52.8) | 0.598 | 0.457 |
| AG | 154 (49.0) | 126 (47.2) | 14 (38.9) | ||
| GG | 38 (12.1) | 34 (12.7) | 3 (8.3) | ||
| AG + GG | 192 (61.1) | 160 (59.9) | 17 (47.2) | 0.272 | 0.211 |
| −422 A > T (rs475007) | |||||
| AA | 82 (26.0) | 76 (28.2) | 16 (43.2) | 0.157 | 0.078 |
| AT | 167 (52.9) | 132 (49.1) | 12 (32.4) | ||
| TT | 67 (21.2) | 61 (22.7) | 9 (24.3) | ||
| AT + TT | 234 (74.1) | 193 (71.8) | 21 (56.7) | 0.085 | 0.040 |
| −340 A > G (rs514921) | |||||
| AA | 165 (52.2) | 129 (48.1) | 20 (54.1) | 0.876 | 0.995 |
| AG | 121 (38.3) | 111 (41.4) | 14 (37.8) | ||
| GG | 30 (9.5) | 28 (10.5) | 3 (8.1) | ||
| AG + GG | 151 (47.8) | 139 (51.9) | 17 (45.9) | 0.561 | 0.957 |
| −320T > C (rs494379) | |||||
| TT | 185 (58.5) | 161 (60.3) | 22 (59.5) | 0.732 | 0.438 |
| TC | 108 (34.2) | 94 (35.2) | 13 (35.1) | ||
| CC | 23 (7.3) | 12 (4.5) | 2 (5.4) | ||
| TC + CC | 131 (41.5) | 106 (39.7) | 15 (40.5) | 0.912 | 0.879 |
SNP, single nucleotide polymorphism.
Two-sided chi-square test for genotype distribution.
Adjusted for age, sex, skin color, hair color, eye color, freckling in the sun as a child, sunburns, tanning ability, number of moles, dysplastic nevi, and first-degree relatives with any cancer.
As shown in Table 3, SNP −422A > T showed significant association with ulceration status (P = 0.023 for codominant) in the univariate analysis. After adjustment for the 11 covariates, two SNPs (−755T > G and −422A > T) had significant association with ulceration status under the dominant genetic model. Compared with wild-type homozygotes (TT genotype), variant carriers (TG or GG genotypes) of −755T > G had an OR of 1.85 (95% CI: 1.04–3.33, P = 0.038) for ulceration, whereas variant carriers (AT or TT genotypes) of −422A > T had an OR of 0.52 (95% CI: 0.31–0.86, P = 0.012) compared with the homozygotes (AA genotype). SNP −839 G > A exhibited a marginally significant association with ulceration status (P = 0.055) under the dominant model.
Table 3.
Associations between SNPs in the MMP1 promoter region and ulceration status.
| SNP | # Patients (%)
|
P-valuea | OR (95% CI)
|
P-valueb | ||
|---|---|---|---|---|---|---|
| Ulceration absent | Ulceration present | Crude | Adjustedb | |||
| −1607 1G/2G (rs1799750) | ||||||
| 1G | 138 (37.6) | 20 (25.3) | 0.417 | 1.00 (Ref) | 1.00 (Ref) | 0.425 |
| 1G/2G | 256 (51.2) | 37 (46.8) | 1.00 (0.55–1.79) | 1.21 (0.66–2.21) | ||
| 2G/2G | 106 (21.2) | 22 (27.9) | 1.43 (0.74–2.76) | 1.58 (0.79–3.14) | ||
| 1G/2G + 2G/2G | 362 (72.4) | 59 (74.7) | 0.672 | 1.13 (0.65–1.94) | 1.32 (0.75–2.33) | 0.342 |
| −839 G > A (rs473509) | ||||||
| GG | 163 (32.6) | 35 (43.8) | 0.148 | 1.00 (Ref) | 1.00 (Ref) | 0.159 |
| GA | 256 (51.2) | 34 (42.5) | 0.62 (0.37–1.03) | 0.61 (0.36–1.04) | ||
| AA | 81 (16.2) | 11 (13.8) | 0.63 (0.31–1.31) | 0.62 (0.29–1.30) | ||
| GA + AA | 337 (67.4) | 45 (56.3) | 0.051 | 0.62 (0.39–1.01) | 0.61 (0.37–1.01) | 0.055 |
| −755 T > G (rs498186) | ||||||
| TT | 165 (33.2) | 18 (23.5) | 0.108 | 1.00 (Ref) | 1.00 (Ref) | 0.059 |
| TG | 234 (47.1) | 47 (58.0) | 1.84 (1.03–3.28) | 2.04 (1.12–3.73) | ||
| GG | 98 (19.7) | 15 (18.5) | 1.40 (0.68–2.91) | 1.40 (0.65–3.04) | ||
| TG + GG | 332 (66.8) | 62 (76.5) | 0.056 | 1.71 (0.98–2.99) | 1.85 (1.04–3.33) | 0.038 |
| −519 A > G (rs1144393) | ||||||
| AA | 192 (38.6) | 38 (47.5) | 0.268 | 1.00 (Ref) | 1.00 (Ref) | 0.361 |
| AG | 241 (48.5) | 35 (43.8) | 0.74 (0.45–1.21) | 0.76 (0.45–1.28) | ||
| GG | 64 (12.9) | 7 (8.8) | 0.55 (0.24–1.30) | 0.57 (0.24–1.37) | ||
| AG + GG | 305 (61.4) | 42 (52.6) | 0.134 | 0.70 (0.43–1.12) | 0.72 (0.44–1.18) | 0.195 |
| −422 A > T (rs475007) | ||||||
| AA | 128 (25.5) | 31 (39.5) | 0.023 | 1.00 (Ref) | 1.00 (Ref) | 0.022 |
| AT | 263 (52.5) | 30 (37.0) | 0.47 (0.27–0.81) | 0.45 (0.26–0.79) | ||
| TT | 110 (22.0) | 19 (23.5) | 0.71 (0.38–1.33) | 0.67 (0.35–1.28) | ||
| AT + TT | 373 (74.5) | 49 (60.5) | 0.014 | 0.54 (0.33–0.89) | 0.52 (0.31–0.86) | 0.012 |
| −340 A > G (rs514921) | ||||||
| AA | 261 (52.0) | 37 (46.3) | 0.322 | 1.00 (Ref) | 1.00 (Ref) | 0.278 |
| AG | 197 (39.2) | 32 (40.0) | 1.15 (0.69–1.91) | 1.20 (0.71–2.04) | ||
| GG | 44 (8.8) | 11 (13.8) | 1.76 (0.84–3.72) | 1.56 (0.70–3.48) | ||
| AG + GG | 241 (48.0) | 43 (53.8) | 0.341 | 1.26 (0.78–2.02) | 1.27 (0.77–2.09) | 0.344 |
| −320T > C (rs494379) | ||||||
| TT | 299 (59.8) | 42 (53.2) | 0.531 | 1.00 (Ref) | 1.00 (Ref) | 0.635 |
| TC | 172 (34.4) | 32 (40.5) | 1.33 (0.81–2.18) | 1.28 (0.76–2.16) | ||
| CC | 29 (5.8) | 5 (6.3) | 1.23 (0.45–3.35) | 1.23 (0.44–3.45) | ||
| TC + CC | 201 (40.2) | 37 (46.8) | 0.266 | 1.31 (0.81–2.11) | 1.27 (0.77–2.10) | 0.343 |
SNP, single nucleotide polymorphism; OR, odds ratio; CI, confidence interval.
Two-sided chi-square test for genotype distribution.
Adjusted for age, sex, skin color, hair color, eye color, freckling in the sun as a child, sunburns, tanning ability, number of moles, dysplastic nevi, and first-degree relatives with any cancer.
Table 4 showed the correlation results between MMP1 promoter SNPs and anatomic site. In the univariate analysis, significant association was found between SNP −755T > G and primary anatomic site (Crude OR = 1.52, 95% CI: 1.11–2.07, P = 0.009 under the dominant model). After adjustment for the 11 covariates, two SNPs were significantly associated with anatomic site (P = 0.003 for −755T > G under the dominant genetic model and P = 0.030 for −422A > T under the codominant genetic model). More patients with axial than with extremity CM were variant allele carriers of −755T > G (TG or GG genotype) (70.6% vs. 61.4%, adjusted OR = 1.67, 95% CI: 1.120–2.34). Also, the SNP −519A > G was marginally associated with anatomic site (adjusted OR = 0.71, 95% CI: 0.51–0.99, P = 0.044 in the dominant model).
Table 4.
Associations between SNPs in the MMP1 promoter region and primary CM anatomic site.
| SNP | # Patients (%)
|
P-valuea | OR (95% CI)
|
P-valueb | ||
|---|---|---|---|---|---|---|
| Extremity | Axial | Crude | Adjustedb | |||
| −1607 1G/2G (rs1799750) | ||||||
| 1G | 88 (29.8) | 121 (27.2) | 0.335 | 1.00 (Ref) | 1.00 (Ref) | 0.177 |
| 1G/2G | 150 (50.9) | 218 (49.0) | 1.06 (0.75–1.49) | 1.20 (0.83–1.75) | ||
| 2G/2G | 57 (19.3) | 106 (23.8) | 1.35 (0.89–2.07) | 1.55 (0.98–2.45) | ||
| 1G/2G + 2G/2G | 207 (70.2) | 324 (72.8) | 0.435 | 1.14 (0.82–1.58) | 1.30 (0.91–1.84) | 0.146 |
| −839 G > A (rs473509) | ||||||
| GG | 93 (31.4) | 160 (35.8) | 0.465 | 1.00 (Ref) | 1.00 (Ref) | 0.412 |
| GA | 153 (51.7) | 215 (48.1) | 0.82 (0.59–1.14) | 0.80 (0.56–1.15) | ||
| AA | 50 (16.9) | 72 (16.1) | 0.84 (0.54–1.30) | 0.78 (0.48–1.25) | ||
| GA + AA | 203 (68.6) | 287 (64.2) | 0.218 | 0.82 (0.60–1.12) | 0.80 (0.57–1.12) | 0.185 |
| −755 T > G (rs498186) | ||||||
| TT | 114 (38.6) | 130 (29.4) | 0.031 | 1.00 (Ref) | 1.00 (Ref) | 0.011 |
| TG | 130 (44.1) | 226 (51.0) | 1.52 (1.10–2.12) | 1.71 (1.19–2.44) | ||
| GG | 51 (17.3) | 87 (19.6) | 1.50 (0.98–2.29) | 1.58 (1.00–2.52) | ||
| TG + GG | 181 (61.4) | 313 (70.6) | 0.009 | 1.52 (1.11–2.07) | 1.67 (1.20–2.34) | 0.003 |
| −519 A > G (rs1144393) | ||||||
| AA | 102 (34.8) | 187 (42.0) | 0.144 | 0.113 | ||
| AG | 150 (51.2) | 204 (45.8) | 0.74 (0.54–1.02) | 0.73 (0.52–1.04) | ||
| GG | 41 (14.0) | 54 (12.1) | 0.72 (0.45–1.15) | 0.64 (0.38–1.07) | ||
| AG + GG | 191 (65.2) | 258 (57.9) | 0.050 | 0.74 (0.54–1.00) | 0.71 (0.51–0.99) | 0.044 |
| −422 A > T (rs475007) | ||||||
| AA | 89 (29.9) | 125 (28.0) | 0.057 | 1.00 (Ref) | 1.00 (Ref) | 0.030 |
| AT | 133 (44.6) | 235 (52.7) | 1.26 (0.89–1.78) | 1.27 (0.87–1.85) | ||
| TT | 76 (25.5) | 86 (19.3) | 0.81 (0.53–1.22) | 0.74 (0.48–1.15) | ||
| AT + TT | 309 (70.1) | 321 (72.0) | 0.587 | 1.09 (0.79–1.51) | 1.07 (0.75–1.51) | 0.723 |
| −340 A > G (rs514921) | ||||||
| AA | 159 (53.5) | 224 (50.2) | 0.565 | 1.00 (Ref) | 1.00 (Ref) | 0.717 |
| AG | 113 (38.1) | 176 (39.5) | 1.11 (0.81–1.51) | 1.03 (0.74–1.44) | ||
| GG | 25 (8.4) | 46 (10.3) | 1.31 (0.77–2.21) | 1.27 (0.72–2.26) | ||
| AG + GG | 138 (46.5) | 222 (48.8) | 0.376 | 1.14 (0.85–1.53) | 1.07 (0.78–1.47) | 0.666 |
| −320T > C (rs494379) | ||||||
| TT | 173 (58.3) | 269 (60.5) | 0.702 | 1.00 (Ref) | 1.00 (Ref) | 0.880 |
| TC | 106 (35.7) | 146 (32.8) | 0.89 (0.65–1.21) | 0.92 (0.66–1.30) | ||
| CC | 18 (6.1) | 30 (6.7) | 1.07 (0.58–1.98) | 1.03 (0.54–1.99) | ||
| TC + CC | 124 (41.8) | 176 (39.5) | 0.550 | 0.92 (0.68–1.23) | 0.94 (0.68–1.30) | 0.703 |
SNP, single nucleotide polymorphism; OR, odds ratio; CI, confidence interval; Ref, reference.
Two-sided chi-square test for genotype distribution.
Adjusted for age, sex, skin color, hair color, eye color, freckling in the sun as a child, sunburns, tanning ability, number of moles, dysplastic nevi, and first-degree relatives with any cancer.
We also performed haplotype analysis to evaluate the joint effect of the seven SNPs on CM overall survival and six clinical factors. There were seven haplotypes with frequencies greater than 0.05, and we combined the rest haplotypes into one group in this analysis (Table 5). Compared with the most frequent haplotype G-A-T-G-T-A-T (the SNP order was −1607 2G/1G, −839G > A, −755T > G, −519A > G, −422A > T, −340A > G, and −320T > C), the frequency of haplotype 2G-G-G-A-A-G-T was significantly higher in CM patients with ulceration (adjusted OR = 2.18, 95% CI: 1.08–4.40, P = 0.030) than that in those without ulceration. No other significant associations were observed between haplotypes and other clinical phenotypes (data were not shown).
Table 5.
Haplotype distribution of the seven SNPs in the study patients with and without ulceration.
| Haplotypea | # Patients (%)
|
P-valueb | OR (95% CI)
|
P-valuec | ||
|---|---|---|---|---|---|---|
| Ulceration absent | Ulceration present | Raw | Adjustedc | |||
| G-A-T-G-T-A-T | 205 (21.2) | 24 (16.3) | --- | 1.00 (Ref) | 1.00 (Ref) | --- |
| 2G-G-G-A-A-A-T | 103 (10.6) | 18 (12.2) | 0.231 | 1.49 (0.78–2.88) | 1.56 (0.78–3.01) | 0.205 |
| G-A-T-G-A-A-T | 84 (8.7) | 9 (6.1) | 0.830 | 0.92 (0.41–2.05) | 1.03 (0.45–2.34) | 0.954 |
| 2G-G-G-A-A-G-T | 76 (7.9) | 18 (12.2) | 0.038 | 2.02 (1.04–3.94) | 2.18 (1.08–4.40) | 0.030 |
| 2G-G-T-A-A-A-C | 73 (7.5) | 12 (8.1) | 0.370 | 1.40 (0.67–2.95) | 1.48 (0.69–3.18) | 0.313 |
| 2G-G-G-A-A-A-C | 52 (5.4) | 9 (6.1) | 0.353 | 1.48 (0.65–3.37) | 1.58 (0.68–3.67) | 0.291 |
| 2G-G-G-A-T-G-T | 53 (5.5) | 8 (5.4) | 0.560 | 1.29 (0.55–3.03) | 1.44 (0.60–3.46) | 0.413 |
| Other d | 322 (33.3) | 50 (33.8) | 0.285 | 1.33 (0.79–2.23) | 1.33 (0.77–2.28) | 0.305 |
OR, odds ratio; CI, confidence interval; Ref, reference.
The SNP order was −1607 2G/1G, −839G > A, −755T > G, −519A > G, −422A > T, −340A > G, and −320T > C.
Two-sided chi-square test for haplotype distribution.
Adjusted for age, sex, skin color, hair color, eye color, freckling in the sun as a child, sunburns, tanning ability, number of moles, dysplastic nevi, and first-degree relatives with any cancer.
Rare haplotypes with frequencies less than 0.05 were combined in one group.
We also examined the association of these seven SNPs with overall survival. The median follow-up time among all subjects was 32.9 months (range: 0.8 – 364.1 months). Forty-six patients died during follow-up because of melanoma or other reasons. However, no SNPs showed significant effect on melanoma survival in the univariate and multivariate Cox regression analyses. We did not find any SNPs in the MMP1 promoter region were significantly associated with Clark level, sentinel lymph node status, and melanoma stages either (data not shown).
Discussion
In this study, for the first time we investigated the associations of seven promoter SNPs in MMP1 with six melanoma clinicopathologic factors and overall survival. Our results showed that four SNPs (−839G > A, −755T > G, −519A > G and −422A > T) were associated with three clinical factors: primary tumor thickness, ulceration status, and anatomic site, even though no significant association was observed between these SNPs and melanoma overall survival with only 46 deaths out of 754 CM patients. Therefore, our results partially supported the hypothesis that MMP1 promoter SNPs might have potential effects on melanoma progression and prognosis by influencing these clinical factors.
Researchers have shown that −1607 1G/2G can regulate MMP1 expression [14]. The association of this SNP with cancer progression and prognosis were not consistent among previous studies. For example, −1607 1G/2G was significantly associated with disease-free or overall survivals in patients with epithelial ovarian cancer [24], colorectal cancer [13]. However, the significant association was not found in breast cancer [18], esophageal adenocarcinoma [17] and lung cancer [25]. In the present study, we did not observe significant association of this variant with melanoma overall survival and the six clinicopathologic factors. This was consistent with one previous study[21], which observed that this variant had no significant effect on melanoma disease free survival, as well as primary tumor thickness and stage. These results may indicate a multilevel regulation of MMP1 expression during tumor progression. In addition to the promoter SNPs, it is known that other factors are involved in the regulation of MMP1 transcription, such as external stimuli and promoter methylation [26–29].
Except for −1607 1G/2G, the effects of other promoter SNPs in MMP1 on cancer progression are little known. In our study, we observed −422A > T was associated with thinner melanoma and absence of ulceration, whereas −755 T > G was associated with presence of ulceration. Although there are little functional studies for the two SNPs, in silico prediction using TFSEARCH (http://www.cbrc.jp/research/db/TFSEARCH.html) shows variant allele T in the position of −422 would abolish the binding site of human transcript factor USF, whereas the variant allele G in −755 could create a new binding site for p300 [30]. Further functional analysis may be warranted for the validation of these predictions.
A previous study demonstrated that the seven promoter SNPs exerted a haplotype effect on the MMP1 promoter activity in cancer cells [19]. In the present study, we observed that the haplotype 2G-G-G-A-A-G-T was more frequent in CM patients with ulceration than patients without ulceration. This result was partially supported by the results of the aforementioned study[19], in which this haplotype (reported as GG-G-G-A-T-C-T) and the haplotype 2G-G-GA-T-A-T (identical to the reported GG-G-G-A-T-T-T) had significantly higher promoter activity than that of other haplotypes in the A2058 melanoma cell line. In our recent study, two MMP1 haplotypes (1G-A-G-A-T-G-T and 2G-G-G-A-T-A-T) were also significantly associated with increased risk of CM [20]. Thus, both of our results are evidence that these seven promoter SNPs of MMP1 may jointly affect CM development and progression.
The correlations between abnormal expression of MMPs and melanoma progression have been evidenced in a number of studies [31]. However, previous studies on functional polymorphisms in MMP2, MMP3 and MMP9 did not provide strong evidence for the roles of these genes in melanoma progression [32, 33]. In this study, we also did not observe direct association between promoter SNPs in MMP1 and melanoma progression and clinical outcome, which indicated that these SNPs might just have moderate effects on melanoma progression. Future investigations on more other MMPs genes and their interactions may be helpful to better understand the genetic mechanism underlying melanoma progression and reveal new progression predictive biomarkers. For example, recently one systematic mutation analysis of MMPs family found the mutation of MMP8 could facilitate melanoma progression [34].
In conclusion, we observed significant associations of SNPs in the MMP1 promoter with three clinicopathologic factors, and one common haplotype composed of the seven promoter SNPs was over-represent among patients with ulceration. These findings provide evidence that potential functional SNPs in the MMP1 promoter may individually or jointly influence CM progression. Certainly, there are limitations in this study. First, because we have less patients with advanced stages, the number of outcome events was small (n = 46), which limited the power of our survival analysis. Second, we do not have the tumor tissues for MMP1 gene expression analysis to correlate with the SNP data. However, the analysis of the available SNPs in the MMP1 promoter region and its mRNA expression from the 90 HapMap CEU cell lines showed a significant correlation for SNP −755T > G (rs498186) [35]. Although the expression data were obtained from human peripheral blood cells not tissue samples, the result might provide some functional evidence for our findings. Definitely, further validation of these results in a larger patient population, the functional significance of each of the promoter SNPs, and their correlation with gene expression levels in tissue samples are warranted to confirm the roles of these SNPs in melanoma progression. Recently available genome-wide association studies may also provide great opportunities for us to further examine the effect of tagging SNPs in the entire MMP1 and other MMPs genes on CM risk and progression.
Acknowledgments
This study was supported by National Institutes of Health grants CA100264 (to Q. W.), P50 CA093459 (to E. A. G.), and CA016672 (to MD Anderson). We thank Margaret Lung and Cesar A. Maldonado for assistance in recruiting the subjects; Kejing Xu and Jianzhong He for laboratory assistance; Julie M. Gardner of the MD Anderson Melanoma Informatics, Tissue Resource, and Pathology Core (MelCore) for assistance in collecting the clinical and pathological data; Donald R. Norwood for the scientific editing.
Footnotes
Conflict of interest statement
None declared.
References
- 1.Ries L, Melbert D, Krapcho M, Stinchcomb D, Howlader N, Horner M, et al. SEER Cancer Statistics Review, 1975–2005. National Cancer Institute; Bethesda, MD: 2008. http://seer.cancer.gov/csr/1975_2005/, based on November 2007 SEER data submission, posted to the SEER web site, 2008. [Google Scholar]
- 2.Elder DE, Van Belle P, Elenitsas R, Halpern A, Guerry D. Neoplastic progression and prognosis in melanoma. Semin Cutan Med Surg. 1996;15:336–348. doi: 10.1016/s1085-5629(96)80047-2. [DOI] [PubMed] [Google Scholar]
- 3.Balch CM, Gershenwald JE, Soong SJ, Thompson JF, Atkins MB, Byrd DR, et al. Final version of 2009 AJCC melanoma staging and classification. J Clin Oncol. 2009;27:6199–6206. doi: 10.1200/JCO.2009.23.4799. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.McCawley LJ, Matrisian LM. Matrix metalloproteinases: multifunctional contributors to tumor progression. Mol Med Today. 2000;6:149–156. doi: 10.1016/s1357-4310(00)01686-5. [DOI] [PubMed] [Google Scholar]
- 5.Roy R, Yang J, Moses MA. Matrix metalloproteinases as novel biomarkers and potential therapeutic targets in human cancer. J Clin Oncol. 2009;27:5287–5297. doi: 10.1200/JCO.2009.23.5556. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Arakaki PA, Marques MR, Santos MC. MMP-1 polymorphism and its relationship to pathological processes. J Biosci. 2009;34:313–320. doi: 10.1007/s12038-009-0035-1. [DOI] [PubMed] [Google Scholar]
- 7.Huntington JT, Shields JM, Der CJ, Wyatt CA, Benbow U, Slingluff CL, Jr, et al. Overexpression of collagenase 1 (MMP-1) is mediated by the ERK pathway in invasive melanoma cells: role of BRAF mutation and fibroblast growth factor signaling. J Biol Chem. 2004;279:33168–33176. doi: 10.1074/jbc.M405102200. [DOI] [PubMed] [Google Scholar]
- 8.Hofmann UB, Houben R, Brocker EB, Becker JC. Role of matrix metalloproteinases in melanoma cell invasion. Biochimie. 2005;87:307–314. doi: 10.1016/j.biochi.2005.01.013. [DOI] [PubMed] [Google Scholar]
- 9.Blackburn JS, Rhodes CH, Coon CI, Brinckerhoff CE. RNA interference inhibition of matrix metalloproteinase-1 prevents melanoma metastasis by reducing tumor collagenase activity and angiogenesis. Cancer Res. 2007;67:10849–10858. doi: 10.1158/0008-5472.CAN-07-1791. [DOI] [PubMed] [Google Scholar]
- 10.Hofmann UB, Westphal JR, Van Muijen GN, Ruiter DJ. Matrix metalloproteinases in human melanoma. J Invest Dermatol. 2000;115:337–344. doi: 10.1046/j.1523-1747.2000.00068.x. [DOI] [PubMed] [Google Scholar]
- 11.Nikkola J, Vihinen P, Vlaykova T, Hahka-Kemppinen M, Kahari VM, Pyrhonen S. High expression levels of collagenase-1 and stromelysin-1 correlate with shorter disease-free survival in human metastatic melanoma. Int J Cancer. 2002;97:432–438. doi: 10.1002/ijc.1636. [DOI] [PubMed] [Google Scholar]
- 12.Rutter JL, Benbow U, Coon CI, Brinckerhoff CE. Cell-type specific regulation of human interstitial collagenase-1 gene expression by interleukin-1 beta (IL-1 beta) in human fibroblasts and BC-8701 breast cancer cells. J Cell Biochem. 1997;66:322–336. [PubMed] [Google Scholar]
- 13.Hettiaratchi A, Hawkins NJ, McKenzie G, Ward RL, Hunt JE, Wakefield D, et al. The collagenase-1 (MMP-1) gene promoter polymorphism - 1607/2G is associated with favourable prognosis in patients with colorectal cancer. Br J Cancer. 2007;96:783–792. doi: 10.1038/sj.bjc.6603630. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Rutter JL, Mitchell TI, Buttice G, Meyers J, Gusella JF, Ozelius LJ, et al. A single nucleotide polymorphism in the matrix metalloproteinase-1 promoter creates an Ets binding site and augments transcription. Cancer Res. 1998;58:5321–5325. [PubMed] [Google Scholar]
- 15.Peng B, Cao L, Wang W, Xian L, Jiang D, Zhao J, et al. Polymorphisms in the promoter regions of matrix metalloproteinases 1 and 3 and cancer risk: a meta-analysis of 50 case-control studies. Mutagenesis. 2009;25:41–48. doi: 10.1093/mutage/gep041. [DOI] [PubMed] [Google Scholar]
- 16.Lei H, Hemminki K, Altieri A, Johansson R, Enquist K, Hallmans G, et al. Promoter polymorphisms in matrix metalloproteinases and their inhibitors: few associations with breast cancer susceptibility and progression. Breast Cancer Res Treat. 2007;103:61–69. doi: 10.1007/s10549-006-9345-2. [DOI] [PubMed] [Google Scholar]
- 17.Bradbury PA, Zhai R, Hopkins J, Kulke MH, Heist RS, Singh S, et al. Matrix metalloproteinase 1, 3 and 12 polymorphisms and esophageal adenocarcinoma risk and prognosis. Carcinogenesis. 2009;30:793–798. doi: 10.1093/carcin/bgp065. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Knechtel G, Hofmann G, Gerger A, Renner W, Langsenlehner T, Szkandera J, et al. Analysis of common germline polymorphisms as prognostic factors in patients with lymph node-positive breast cancer. J Cancer Res Clin Oncol. 2010;136:1813–1819. doi: 10.1007/s00432-010-0839-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Pearce EG, Laxton RC, Pereira AC, Ye S. Haplotype effects on matrix metalloproteinase-1 gene promoter activity in cancer cells. Mol Cancer Res. 2007;5:221–227. doi: 10.1158/1541-7786.MCR-06-0139. [DOI] [PubMed] [Google Scholar]
- 20.Wang LE, Huang YJ, Yin M, Gershenwald JE, Prieto VG, Lee JE, et al. Promoter polymorphisms in matrix metallopeptidase 1 and risk of cutaneous melanoma. Eur J Cancer. 2010 doi: 10.1016/j.ejca.2010.06.129. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Ye S, Dhillon S, Turner SJ, Bateman AC, Theaker JM, Pickering RM, et al. Invasiveness of cutaneous malignant melanoma is influenced by matrix metalloproteinase 1 gene polymorphism. Cancer Res. 2001;61:1296–1298. [PubMed] [Google Scholar]
- 22.Balch CM, Soong SJ, Gershenwald JE, Thompson JF, Reintgen DS, Cascinelli N, et al. Prognostic factors analysis of 17,600 melanoma patients: validation of the American Joint Committee on Cancer melanoma staging system. J Clin Oncol. 2001;19:3622–3634. doi: 10.1200/JCO.2001.19.16.3622. [DOI] [PubMed] [Google Scholar]
- 23.Dickson PV, Gershenwald JE. Staging and prognosis of cutaneous melanoma. Surg Oncol Clin N Am. 2010;20:1–17. doi: 10.1016/j.soc.2010.09.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Six L, Grimm C, Leodolter S, Tempfer C, Zeillinger R, Sliutz G, et al. A polymorphism in the matrix metalloproteinase-1 gene promoter is associated with the prognosis of patients with ovarian cancer. Gynecol Oncol. 2006;100:506–510. doi: 10.1016/j.ygyno.2005.08.049. [DOI] [PubMed] [Google Scholar]
- 25.Scherf DB, Dally H, Muller P, Werle-Schneider G, Jager B, Edler L, et al. Single nucleotide polymorphisms in matrix metalloproteinase genes and lung cancer chemotherapy response and prognosis. Eur Respir J. 2009;35:381–390. doi: 10.1183/09031936.00125608. [DOI] [PubMed] [Google Scholar]
- 26.Iida J, McCarthy JB. Expression of collagenase-1 (MMP-1) promotes melanoma growth through the generation of active transforming growth factor-beta. Melanoma Res. 2007;17:205–213. doi: 10.1097/CMR.0b013e3282a660ad. [DOI] [PubMed] [Google Scholar]
- 27.Repeke CE, Trombone AP, Ferreira SB, Jr, Cardoso CR, Silveira EM, Martins W, Jr, et al. Strong and persistent microbial and inflammatory stimuli overcome the genetic predisposition to higher matrix metalloproteinase-1 (MMP-1) expression: a mechanistic explanation for the lack of association of MMP1-1607 single-nucleotide polymorphism genotypes with MMP-1 expression in chronic periodontitis lesions. J Clin Periodontol. 2009;36:726–738. doi: 10.1111/j.1600-051X.2009.01447.x. [DOI] [PubMed] [Google Scholar]
- 28.Wang H, Ogawa M, Wood JR, Bartolomei MS, Sammel MD, Kusanovic JP, et al. Genetic and epigenetic mechanisms combine to control MMP1 expression and its association with preterm premature rupture of membranes. Hum Mol Genet. 2008;17:1087–1096. doi: 10.1093/hmg/ddm381. [DOI] [PubMed] [Google Scholar]
- 29.Benbow U, Tower GB, Wyatt CA, Buttice G, Brinckerhoff CE. High levels of MMP-1 expression in the absence of the 2G single nucleotide polymorphism is mediated by p38 and ERK1/2 mitogen-activated protein kinases in VMM5 melanoma cells. J Cell Biochem. 2002;86:307–319. doi: 10.1002/jcb.10225. [DOI] [PubMed] [Google Scholar]
- 30.Heinemeyer T, Wingender E, Reuter I, Hermjakob H, Kel AE, Kel OV, et al. Databases on transcriptional regulation: TRANSFAC, TRRD and COMPEL. Nucleic Acids Res. 1998;26:362–367. doi: 10.1093/nar/26.1.362. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Deryugina EI, Quigley JP. Matrix metalloproteinases and tumor metastasis. Cancer Metastasis Rev. 2006;25:9–34. doi: 10.1007/s10555-006-7886-9. [DOI] [PubMed] [Google Scholar]
- 32.Cotignola J, Roy P, Patel A, Ishill N, Shah S, Houghton A, et al. Functional polymorphisms in the promoter regions of MMP2 and MMP3 are not associated with melanoma progression. J Negat Results Biomed. 2007;6:9. doi: 10.1186/1477-5751-6-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Cotignola J, Reva B, Mitra N, Ishill N, Chuai S, Patel A, et al. Matrix Metalloproteinase-9 (MMP-9) polymorphisms in patients with cutaneous malignant melanoma. BMC Med Genet. 2007;8:10. doi: 10.1186/1471-2350-8-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Palavalli LH, Prickett TD, Wunderlich JR, Wei X, Burrell AS, Porter-Gill P, et al. Analysis of the matrix metalloproteinase family reveals that MMP8 is often mutated in melanoma. Nat Genet. 2009;41:518–520. doi: 10.1038/ng.340. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Holm K, Melum E, Franke A, Karlsen TH. SNPexp - A web tool for calculating and visualizing correlation between HapMap genotypes and gene expression levels. BMC Bioinformatics. 2010;11:600. doi: 10.1186/1471-2105-11-600. [DOI] [PMC free article] [PubMed] [Google Scholar]