Genetic Polymorphisms of Matrix Metalloproteinases: Susceptibility and Prognostic Implications for Prostate Cancer

Genetic Polymorphisms of Matrix Metalloproteinases: Susceptibility and Prognostic Implications for Prostate Cancer Sabrina Thalita dos Reis, José Pontes, Jr., Fabiola Elizabeth Villanova, Priscila Maria de Andrade Borra, Alberto Azoubel Antunes, Marcos Francisco Dall’oglio, Miguel Srougi and Katia Ramos Moreira Leite* From the Laboratory of Medical Investigation (LIM55), Urology Department, Faculdade de Medicina da Universidade de São Paulo, São Paulo, Brazil

Abbreviations and Acronyms MMP ⫽ matrix metalloproteinase PCa ⫽ prostate cancer PCR ⫽ polymerase chain reaction PSA ⫽ prostate specific antigen SNP ⫽ single nucleotide polymorphism Submitted for publication August 25, 2008. Study received institutional board of ethics approval. * Correspondence: Faculdade de Medicina da Universidade de São Paulo, Av. Dr. Arnaldo 455, 2 Andar, Sala 2141, Cerqueira César, São Paulo, São Paulo, Brasil CEP: 01246-903 (e-mail: [email protected]).

Purpose: Prostate cancer is the most common tumor in males in Brazil. Single nucleotide polymorphisms have been demonstrated to exist in the promoter regions of matrix metalloproteinase genes and they are associated with the development and progression of some cancers. We investigated the correlation between MMP1, 2, 7 and 9 polymorphisms with susceptibility to prostate cancer, and classic prognostic parameters of prostate cancer. Materials and Methods: Genomic DNA was extracted using conventional protocols. The DNA sequence containing the polymorphic site was amplified by realtime polymerase chain reaction using TaqMan® fluorescent probes. Results: For the MMP1 gene the polymorphic allele was more common in the control group than in the prostate cancer group (p ⬍0.001). For the MMP9 gene the incidence of the polymorphic homozygote genotype was higher in the prostate cancer group (p ⬍0.001). For higher stage tumors (pT3) a polymorphic allele in the MMP2 gene was more common (p ⫽ 0.026). When considering Gleason score, the polymorphic homozygote genotype of MMP9 was more common in Gleason 6 or less tumors (p ⫽ 0.003), while a polymorphic allele in the MMP2 gene was more common in Gleason 7 or greater tumors (p ⫽ 0.042). Conclusions: MMP1 and MMP2 may protect against prostate cancer development and MMP9 may be related to higher risk. In contrast, MMP9 polymorphism was associated with a lower Gleason score and MMP2 polymorphism was associated with nonorgan confined disease. Key Words: prostate; prostatic neoplasms; polymorphisms, genetic; matrix metalloproteinases; gene expression

2320

www.jurology.com

PROSTATE cancer is a common disease with a multifactorial and complex etiology. It is the most common male malignancy and the second leading cause of death in many countries, including Brazil. The worldwide incidence of PCa varies widely among different ethnic groups and geographic areas.1 The widespread use of PSA testing has increased the detection of this cancer at earlier stages, although

this diagnostic method has proved to be insufficient to identify the disease.2 Staging pathological and Gleason scores are the most important prognostic factors but they have been shown to imperfectly discriminate patients at risk for progression.3 Therefore, research has been directed toward identifying molecular markers that can predict PCa predisposition and progression.

0022-5347/09/1815-2320/0 THE JOURNAL OF UROLOGY® Copyright © 2009 by AMERICAN UROLOGICAL ASSOCIATION

Vol. 181, 2320-2325, May 2009 Printed in U.S.A. DOI:10.1016/j.juro.2009.01.012

MATRIX METALLOPROTEINASE GENETIC POLYMORPHISMS AND PROSTATE CANCER

MMPs are an important family of metal dependent enzymes that are responsible for the degradation of extracellular matrix components, including the basement membrane, interstitial collagen, fibronectin and various proteoglycans that are involved in various physiological processes, such as embryogenesis and tissue remodeling.4 They also have a key role in the invasion of tumor cells and metastasis, which require proteolysis of basal membranes and extracellular matrix. For a long time MMPs were considered to be almost exclusively important for these 2 steps of carcinogenesis. However, recent studies suggest that MMPs are involved in several other processes associated with cancer development. Indeed, they regulate tumor growth and apoptosis, promote angiogenesis and induce the loss of cell adhesion, facilitating invasion and metastasis. Finally, some of them are also required for immune responses to cancer.5 Although MMPs are over expressed in many tumor tissues, somatic cell mutations and gene transpositions of MMPs are rarely seen. Therefore, genetic variants that influence the level of MMP gene expression or protein function could be expected to influence the role of these enzymes in tumor invasion and metastasis.6 A functional SNP has been reported in the MMP1 gene promoter, consisting of a guanosine (G) insertion at position –1607. This SNP generates a new 5=-GGA-3= core recognition sequence for members of the Ets family of transcription factors and the 2G/2G genotype was reported to be linked to an increased risk of colorectal cancer.7 A common functional polymorphism that abolishes an Sp1 binding site has been described for the promoter region of MMP2 (⫺1306 C/T)8 and the TT genotype has been associated with a decreased risk of lung9 and gastric cardia10 cancers. In the MMP7 gene promoter region 1 SNP (⫺181A/G) has been shown to modify gene transcription activity, and it is associated with increased susceptibility to esophageal squamous cell carcinoma, gastric cardiac adenocarcinoma and nonsmall cell lung carcinoma.11 In the MMP9 gene a functional SNP (R279Q), which presumably enhances substrate binding12 and potentially alters the protein structure of MMP9, may have some functional relevance that affects individual susceptibility to cancer.13 We explored whether MMP1, MMP2, MMP7 and MMP9 gene polymorphisms are involved in the risk and prognosis of prostate cancer.

2321

MATERIALS AND METHODS Patients A case-control study was done using surgical specimens from 100 patients with a mean ⫾ SD age of 65 ⫾ 6.9 years who underwent radical prostatectomy at our hospital from 1993 to 2007. Gleason grade and pathological TNM 2002 stage were used as prognostic factors. For analysis tumors were defined as organ confined (pT2) or nonorgan confined (pT3). Gleason score was classified as low grade—Gleason score 6 or less and high grade—Gleason score 7 or greater. The prostate was formalin fixed, paraffin embedded and always examined in toto by the same uropathologist (KRML). A total of 100 serum samples from men with a mean age of 54 ⫾ 7.4 years at the same hospital served as controls. Their PSA levels were within the normal limit (less than 2.5 ng/ml) and digital rectal examination of the prostate was unremarkable. To establish the incidence of these polymorphisms in the Brazilian population we studied another 147 healthy men from the community with no urological hospital records. All men in each group provided informed consent to participate in the study and allow their biological samples to be genetically analyzed. Approval for the study was provided by the institutional board of ethics (769/06).

Analysis of SNPs in MMP Genes Genomic DNA was extracted from paraffin blocks and serum using conventional protocols14 and the GFX kit (GE Healthcare, Chalfont St. Giles, United Kingdom), respectively. SNPs were genotyped using the TaqMan SNP Genotyping Assay Kit and an ABI™ 7500 fast system. SNP specific PCR primers and fluorogenic probes were designed using Primer Express® (table 1). Fluorogenic probes were labeled with a reporter dye (FAM or VIC) and they were specific for 1 of the 2 possible bases identified at the 4 SNPs in the MMP genes. The target sequence was amplified in a 10 ␮l reaction volume containing 5 ␮l TaqMan Universal PCR Master Mix, 0.25 ␮l SNP genotyping assay (primers and probes), 1 ␮l genomic DNA and 3.75 ␮l deoxyribonuclease-free water. PCR cycling conditions were 2 minutes at 50C and 10 minutes at 95C, followed by 40 cycles of 15 seconds at 95C and 60 seconds at 60C. After PCR amplification end point plate reading was performed using the 7500 Fast Real-Time PCR System (Applied Biosystems®). Sequence Detection System software (Applied Biosystems) uses the fluorescence measurements made during the plate read to plot fluorescence (Rn) values based on the signals from each well. The plotted fluorescence signals indicate which alleles are in each sample. Since different specimens were used for analysis (DNA extracted from serum and paraffin blocks), in 3 instances we performed the PCR reaction using serum and tissue

Table 1. Primers used to genotype SNPs in MMP genes Gene

SNP

SNP Database

Primers 5=–3=

MMP1 MMP2 MMP7 MMP9

⫺1607 1G/2G ⫺1306 C/T ⫺181 A/G R279Q

Rs 1799750 Rs 243865 Rs 11568818 Rs 17576

TGACTTTTAAAACATAGTCTATGTTCA, TCTTGGATTGATTTGAGATAAGTCATAGC CTGACCCCCAGTCCTATCTGCC, TGTTGGGAACGCCTGACTTCAG TGGTACCATAATGTCCTGAATG, TCGTTATTGGCAGGAAGCACACAATGAATT GAGAGATGGGATGAACTG, GTGGTGGAAATGTGGTGT

2322

MATRIX METALLOPROTEINASE GENETIC POLYMORPHISMS AND PROSTATE CANCER

from the same patient for comparison. Results were similar for the 2 specimens, and so we assumed that paraffin sections were suitable for study.

Statistical Analysis Associations between genotype and allelic frequencies, and the prognostic factors of PCa were examined with the Fisher exact test. The OR was calculated as an approximation of relative risk and the respective 95% CI. The chi-square test was also used to compare prognostic variables.

RESULTS Table 2 lists the demographic characteristics of the studied population. Table 3 lists genotype distributions in patients with PCa and controls. They did not significantly deviate from the values expected for Hardy-Weinberg equilibrium. We found statistically significant differences in genotype distribution between patients and controls for the MMP1, MMP2 and MMP9 genes. For the MMP1 gene the incidence of the 1G/1G, 1G/2G and 2G/2G genotypes was 55%, 34% and 11% in healthy controls, and 27%, 52% and 21% in patients with PCa, respectively (p ⬍0.001). For the MMP2 gene the incidence of the TT, CT and CC genotypes was 21%, 20% and 59% in healthy controls, and 12%, 38% and 50% in patients with PCa, respectively (p ⫽ 0.012). For the MMP9 gene the incidence of the AA, AG and GG genotypes was 5%, 93% and 2% in healthy controls, and 1%, 43% and 56% in patients with PCa, respectively (p ⬍0.001). For the MMP7 gene the incidence of the AA, AG and GG genotypes was 25%, 39% and 36% in healthy controls, and 33%, 41% and 26% in patients with PCa, respectively (p ⫽ 0.251). There were also statistical differences in allelic frequency between patients with PCa and controls for MMP1 and MMP9. The 1G and 2G MMP1 alleles were detected in 72% and 28% of healthy controls, and in 53% and 47% of patients with PCa, respectively (p ⬍0.001). For MMP9 the A and G alleles

Table 3. MMP gene genotypes in patients with PCa and healthy controls No. PCa (%)

No. Controls (%)

OR (95% CI)

21 (21) 27 (27) 52 (52)

11 (11) 55 (55) 34 (34)

1.00 3.89 (1.64–9.21) 1.25 (0.53–2.91)

50 (50) 12 (12) 38 (38)

59 (59) 21 (21) 20 (20)

1.00 1.49 (0.66–3.31) 0.45 (0.23–0.86)

33 (33) 26 (26) 41 (41)

25 (25) 36 (36) 39 (39)

1.00 1.82 (0.88–3.77) 1.25 (0.63–2.47)

1 (1) 56 (56)† 43 (43)

5 (5) 2 (2) 93 (93)

1.00 140.0 (10.7–1827) 2.31 (0.26–20395)

p Value ⬍0.001

MMP1: 1G/1G* 2G/2G 1G/2G MMP2: CC* TT CT MMP7: AA* GG AG MMP9: AA* GG AG

0.012

0.251

⬍0.001

* Genotype homozygote WT. † PCa risk increased 140-fold.

were detected in 53% and 47% of healthy controls, and in 22% and 78% of patients with PCa, respectively (p ⬍0.001). For the MMP2 and MMP7 genes the differences did not attain statistical significance. For MMP2 the T and C alleles were present in 31% and 69% of healthy controls, and in 31% and 69% of patients with PCa, respectively (p ⫽ 0.920). For MMP7 the A and G alleles were detected in 44% and 56% of healthy controls, and in 54% and 46% of patients with PCa, respectively (p ⫽ 0.089, table 4). Regarding prognosis, additional analysis was performed according to pathological stage and Gleason score. There were differences between genotype and allele frequencies regarding the Gleason score in the MMP9 and MMP2 genes. The incidence of the polymorphic homozygote genotype and polymorphic allele was higher in well differentiated tumors with a Gleason score of 6 or less in the MMP9 gene (p ⫽ 0.003 and 0.010, respectively), Table 4. MMP gene alleles in patients with PCa and healthy controls

Table 2. Study patient characteristics PCa Mean ⫾ SD age No. race (%): White Not white Median ng/ml PSA (range) No. Gleason score (%): 6 or Less 7 or Greater No. pathological stage (%): pT2 pT3 * Significantly different vs PCa.

Controls

65 ⫾ 6.9

54 ⫾ 7.4*

86 (86) 14 (14) 7.5 (1.6–120.0)

81 (81) 19 (19) 0.6 (0.1–1.1)*

81 19

(81) (19)

— —

74 26

(74) (26)

— —

p Value ⬍0.001 0.213

⬍0.001

No. PCa (%)

No. Controls (%)

OR (95% CI)

94 (47) 106 (53)

56 (28) 144 (72)

1.00 0.43 (0.28–0.66)

138 (69) 62 (31)

138 (69) 62 (31)

1.00 1.00 (0.65–1.52)

107 (54) 93 (46)

89 (44) 111 (56)

1.00 0.70 (0.47–1.03)

45 (22) 155 (78)

103 (52) 97 (48)

1.00 3.65 (2.37–5.63)

MMP1: 1G* 2G MMP2: C* T MMP7: A* G MMP9: A* G * Allele WT.

p Value ⬍0.001

0.920

0.089

⬍0.001

MATRIX METALLOPROTEINASE GENETIC POLYMORPHISMS AND PROSTATE CANCER

while a polymorphic allele was more common in tumors with a Gleason score of 7 or greater in the MMP2 gene (p ⫽ 0.042, table 5). When analyzing pathological stage, and genotype and allele distributions, we found statistical differences in the MMP2 gene. The polymorphic allele was more common in nonorgan confined tumors (pT3) (p ⫽ 0.026, table 6).

DISCUSSION

Table 6. MMP gene alleles and genotypes vs pathological stage No. pT2 (%)

Genotype: CC* TT CT Allele: C T

Table 5. MMP gene alleles and genotypes vs Gleason score

Genotype: AA* GG AG Allele: A G

6 or Less

7 or Greater

OR (95% CI)

p Value

MMP1 Genotype: 1G/1G* 2G/2G 1G/2G Allele: 1G 2G Genotype: CC* TT CT Allele: C T Genotype: AA* GG AG Allele: A G Genotype: AA* GG AG Allele: A G

0.538 18 (22) 20 (25) 43 (53)

3 (16) 7 (37) 9 (47)

1.00 2.10 (0.47–9.36) 1.26 (0.30–5.18)

15 (40) 23 (60) MMP2

1.00 1.46 (0.71–3.00) 0.090

43 (53) 7 (9) 31 (38)

7 (37) 5 (26) 7 (37)

1.00 4.39 (1.08–17.7) 1.39 (0.44–4.35) 0.042

117 (72) 45 (28)

21 (55) 17 (45) MMP7

1.00 2.10 (1.02–4.35) 0.903

26 (32) 21 (26) 34 (42)

7 (37) 5 (26) 7 (37)

1.00 0.88 (0.24–3.19) 0.77 (0.23–2.45) 0.809

86 (53) 76 (47)

21 (55) 17 (45) MMP9

0 51 (63) 30 (37)

1 (5) 5 (26) 13 (69)

Could not estimate

30 (19) 132 (81)

15 (40) 23 (60)

1.00 0.35 (0.16–0.75)

1.00 0.92 (0.45–1.86) 0.003

0.005

* Genotype homozygote WT.

OR (95% CI)

p Value

Genotype: AA* GG AG Allele: A G

0.278 14 (19) 23 (31) 37 (50)

7 (27) 4 (15) 15 (58)

1.00 0.35 (0.08–1.40) 0.80 (0.27–2.40)

65 (44) 83 (56)

29 (56) 23 (44) MMP2

1.00 0.62 (0.33–1.17)

42 (57) 7 (9) 25 (34)

8 (31) 5 (19) 13 (50)

1.00 3.75 (0.94–14.8) 2.73 (0.99–7.49)

29 (56) 23 (44) MMP7

1.00 2.22 (1.15–4.28)

23 (31) 23 (31) 28 (38)

10 (39) 3 (11) 13 (50)

1.00 0.30 (0.88–3.77) 1.07 (0.63–2.47)

74 (50) 74 (50)

33 (63) 19 (37) MMP9

1.00 1.73 (0.30–1.10)

1 (2) 44 (59) 29 (39)

0 12 (46) 14 (54)

Could not estimate

31 (21) 117 (79)

14 (27) 38 (73)

1.00 0.72 (0.35–1.49)

0.189

0.065

0.026 109 (74) 39 (26)

0.147

0.130

0.383

0.488

* Genotype homozygote WT. 0.302

79 (49) 83 (51)

No. pT3 (%) MMP1

Genotype: 1G/1G* 2G/2G 1G/2G Allele: 1G 2G

In this case-control study we investigated the associations of SNPs in the MMP1, 2, 7 and 9 genes with PCA susceptibility, staging and Gleason score. The 2G polymorphic allele in the MMP1 gene was significantly more prevalent in controls, suggesting that these polymorphisms may have a role in protecting against the development of PCa in our population. Most likely this polymorphism confers lower activity of the promoter in the MMP1 gene, which may then be associated with a decreased risk of PCa. The G polymorphic allele in the MMP9 gene was more prevalent in patients with PCa, suggesting that the

No. Gleason Score (%)

2323

polymorphism may be related to a predisposition to PCa. Others have noted an association of the MMP9 R279Q genotype with lung carcinoma.15 MMPs are zinc metalloproteases that degrade collagens of the extracellular matrix that are important for tissue remodeling and repair during development and inflammation. MMPs are involved in cell cycle checkpoint controls, genomic instability and cell adhesion. They also contribute to tumor initiation and development by altering the cellular microenvironment that facilitates tumor formation, including angiogenesis.16 Excessive or inappropriate expression of MMPs may contribute to the pathogenesis of cancer in a wide variety of diseases by facilitating tissue degradation. Currently more than 20 MMPs have been identified, which can be categorized by substrate specificity. There are functional polymorphisms associated with the characteristics of expression of these genes that may serve as genetic biomarkers for the diagnosis and prognosis of neoplasms.17 In PCa specifically only Sfar et al have investigated the impact of the ⫺1562C/T polymorphism in the MMP9 gene.18 They suggested that the polymor-

2324

MATRIX METALLOPROTEINASE GENETIC POLYMORPHISMS AND PROSTATE CANCER

phism is an independent risk factor for PCa development and progression in a population from Tunisia. Interestingly we were able to correlate polymorphisms in the MMP genes with tumor stage and grade. The MMP2 polymorphic allele was more common in pT3 and Gleason 7 or greater PCa cases, and the polymorphic genotype and allele of the MMP9 gene were related to low grade tumors with a Gleason score of 6 or less. In cell cultures the genotype CT of MMP2 was related to lower levels of MMP2 mRNA19 and surprisingly lower circulating levels of MMP2 were related to more aggressive prostate cancer. We can make a case that the polymorphism that we have related to the more aggressive phenotype in PCa may be responsible for lower expression of the gene and lower MMP2 levels. To our knowledge the relationship between prognostic factors of PCa and MMP polymorphisms has never been reported previously and these data should be confirmed in a larger study. In other tumors there are descriptions of MMP polymorphisms and prognosis. When studying head and neck cancer, Lai et al noted an association between the 16071G/2G polymorphism in the MMP1 gene and invasiveness.20 Hu et al reported that the R279Q polymorphism in the MMP9 gene is related to the development of lung cancer metastasis.15 In cases of colorectal cancer the 1306C/T polymorphism in the MMP2 gene was related to tumor progression and invasion.19 Since we were aware that our control group also consisted of men with urological hospital records,

which could have interfered with our results, we studied healthy men from the community with no urological records to compare MMP genotypes. When studying 147 men, we found differences in distribution only for the MMP1 gene and the homozygote polymorphic genotype was more common in men with hospital records. However, this fact is even more interesting since we observed that the polymorphism is related to protection against PCa development. Men with hospital records had previously undergone examination and PSA measurement, and the possibility of PCa was considered to be zero. However, the community population did not undergo screening, and so it may be affected in some proportion by the neoplasm.

CONCLUSIONS To our knowledge this is the first study to evaluate polymorphisms of MMP1, MMP2, MMP7 and MMP9 in prostate cancer. Our results show a relationship between MMP1 gene polymorphisms and protection against the development of PCa, and a relationship between the MMP9 polymorphic genotype and the risk of PCa. Moreover, we found that the MMP2 and MMP9 polymorphisms are related to a prognostic parameter of this disease. This is a preclinical, exploratory study to determine a new tumor marker and it should be submitted to further experiments to confirm the value of MMPs polymorphisms as markers of PCa predisposition and prognosis.

REFERENCES 1. Jemal A, Siegel R, Ward E, Hao Y, Xu J, Murray T and Thun MJ: Cancer Statistics, 2008. CA cancer J Clin 2008; 58: 71. 2. Partin AW, Kattan MW, Subong EN, Walsh PC, Wojno KJ, Oesterling JE et al: Combination of prostate-specific antigen, clinical stage, and Gleason score to predict pathological stage of localized prostate cancer. A multi-institutional update. JAMA 1997; 277: 1445. 3. Gleason DF: Histologic grading of prostate cancer: a perspective. Hum Pathol 1992; 23: 273. 4. Nelson AR, Fingleton B, Rothenberg ML and Matrisian LM: Matrix metalloproteinases: biologic activity and clinical implications. J Clin Oncol 2000; 18: 1135. 5. Visse R and Nagase H: Matrix metalloproteinases and tissue inhibitors of metalloproteinases: structure, function, and biochemistry. Circ Res 2003; 92: 827. 6. Curran S and Murray GI: Matrix metalloproteinases in tumor invasion and metastasis. J Pathol 1999; 189: 300.

7. Zinzindohoue F, Lecomte T, Ferraz JM, Houllier AM, Cugnene PH, Berger A et al: Prognostic significance of MMP-1 and MMP-3 functional promoter polymorphisms in colorectal cancer. Clin Cancer Res 2005; 15: 594. 8. Price SJ, Greaves DR and Watkins H: Identifcation of novel, functional genetic variants in the human matrix metalloproteinase-2 gene: role of SP-1 in allele-specific transcriptional regulation. J Biol Chem 2001; 276: 7549. 9. Yu C, Pan K, Xing D, Liang G, Tan W, Zhang L and Lin D: Correlation between a single nucleotide polymorphism in the matrix metalloproteinase-2 promoter and risk of lung cancer. Cancer Res 2002; 62: 6430. 10. Miao X, Yu C, Tan W, Xiong P, Liang G, Lu W et al: A functional polymorphism in the matrix metalloproteinase-2 gene promoter (⫺1306 C/T) is associated with risk of development but not metastasis of gastric cardia adenocarcinoma. Cancer Res 2003; 63: 3987.

11. Zhang J, Jin X, Fang S, Wang R, Li Y, Wang N et al: The functional polymorphism in the matrix metalloproteinase-7 promoter increases susceptibility to esophageal squamous cell carcinoma, gastric cardiac adenocarcinoma and non-small cell lung carcinoma. Carcinogenesis 2005; 26: 1748. 12. Allan JA, Docherty AJ, Barker PJ, Huskisson NS, Reynolds JJ and Murphy G: Binding of gelatinases A and B to type-I collagen and other matrix components. Biochem J 1995; 309: 299. 13. O’Farrell TJ and Pourmotabbed T: Identification of structural elements important for matrix metalloproteinase type V collagenolytic activity as revealed by chimeric enzymes. Role of fibronectinlike domain and active site of gelatinase. J Biol Chem 2000; 275: 27964. 14. Wright DK and Manos MM: Sample Preparation From Paraffin-Embedded Tissues. New York: Academic Press 1990; pp 153–158. 15. Hu Z, Huo X, Lu D, Qian J, Zhou J, Chen Y et al: Functional polymorphisms of matrix metallopro-

MATRIX METALLOPROTEINASE GENETIC POLYMORPHISMS AND PROSTATE CANCER

teinase-9 are associated with risk of occurrence and metastasis of lung cancer. Clin Cancer Res 2005; 11: 5433. 16. Ohbayashi H: Matrix metalloproteinases in lung diseases. Curr Protein Pept Sci 2002; 3: 409. 17. Zhu Y, Spitz MR, Lei L, Mills GB and Wu X: A Single nucleotide polymorphism in the matrix

metalloproteinase-1 promoter enhances lung cancer susceptibility. Cancer Res 2001; 61: 7825. 18. Sfar S, Saad H, Mosbah F, Gabbouj S and Chouchane L: TSP1 and MMP9 genetic variants in sporadic prostate cancer. Cancer Genet Cytogenet 2007; 172: 38. 19. Xu E, Lai M, Lv B, Xing X, Huang Q and Xia X: A Single nucleotide polymorphism in the matrix

2325

metalloproteinase-2 promoter is associated with colorectal cancer. Biochem Biophys Commun 2004; 324: 999. 20. Lai HC, Chu CM, Lin YW, Chang CC, Nieh S, Yu MH et al: Matrix metalloproteinase 1 gene polymorphism as a prognostic predictor of invasive cervical cancer. Gynecol Oncol 2005; 96: 314.