Matrix metalloproteinase-2 promoter and tissue inhibitor of metalloproteinase-2 gene polymorphisms in non-Hodgkin's lymphoma

Authors


Abstract

Single nucleotide polymorphisms in the promoter regions of matrix metalloproteinases (MMPs) and tissue inhibitors of metalloproteinase (TIMP) genes are associated with an adverse outcome in some cancers. We examined three polymorphisms: −1306C/T and −735C/T in MMP-2 and −418G/C in the TIMP-2 gene, using chain reaction restriction fragment length polymorphism typing analysis in 575 patients with non-Hodgkin's lymphoma (NHL). We examined the possible correlations between the three polymorphisms (MMP-2 (−1306C/T and −735C/T) and TIMP-2 gene (−418G/C)) and the clinical significance and survival rate in patients with NHL. The incidence of the CT, TT+CT genotypes and T allele of −735C/T was significantly higher in stage III and IV patients compared to stage I and II patients. In cases with bone marrow infiltration, the TT genotypes of the −1306C/T gene were significantly less frequent compared to CC genotypes. The CT, TT and CT+TT genotypes and T allele in patients exhibiting the −1306C/T polymorphism were significantly less frequent in patients with a large tumor size compared to a smaller tumor. The TT genotypes of the −735C/T polymorphism were more common in patients with a large tumor compared to those with a smaller tumor. The frequency of the −1306C/−735T haplotype in patients with a smaller tumor size was significantly higher compared to patients with a large tumor. The −1306T/−735C and −1306C/−735C haplotypes were significantly less frequent in patients with B-symptoms compared to those without. Interestingly, patients with the −735CT genotype exhibited a lower rate of survival. Our results demonstrate that certain MMP-2 and TIMP-2 gene polymorphisms potentially effect the progression or assessment of prognosis for NHL. This research warrants further, larger scale studies.

Matrix metalloproteinases (MMPs) are zinc-dependent neutral enzymes responsible for the degradation of the extracellular matrix (ECM) components, including the basement membrane, a natural barrier with a key role in preventing invasion and migration of tumor cells.1 Several previous studies have suggested that in many cancer types, including malignant lymphoma, MMPs play critical roles in almost every process of cancer development.2–6 They regulate growth and apoptosis of tumors, accelerate angiogenesis and reduce cell adhesion, facilitating tumor formation, metastasis and invasion and influence on immune responses to cancer.7 MMP-2, (also known as gelatinase A or type IV collagenase), one of the members of the MMP family, is primarily hydrolysis gelatine and type IV and V collagen gelatine and type IV and V collagen, which are the major components of the base membrane2 Over-expression of MMP-2 has been demonstrated in the invasion and metastasis of various cancers.3, 8–10

Human malignant non-Hodgkin's lymphomas (NHLs) can be defined as a heterogeneous condition of neoplasms. Subtypes of NHL are not only characterized by different morphology and pathophysiology but also by clinical differences. These differences may be related to the various mechanisms that control ECM metabolism. It has also been previously demonstrated that the ECM composition may differ in different types of NHL. Some reports have shown that different NHL types produce differing quantities of ECM-degrading enzymes and exhibit differing proteolytic activities.11, 12 Recent evidence has demonstrated that matrix metalloproteinases can serve as prognostic indicators in lymphomas.5, 11, 13

The tissue inhibitors of metalloproteinases (TIMPs), a family of secreted proteins, may inhibit the degradation caused by MMPs. In pathological events, TIMPs may participate in tumor invasion and metastasis and control tumorigenesis through remodeling the ECM homeostasis.14, 15 Of the four members of the TIMP family, TIMP-2 is conspicuous due to its dual functionality in terms of regulation of MMP-2 activity and its controversial contribution to tumor progression.16, 17 TIMP-2 has also been previously described as a growth factor and was demonstrated to inhibit apoptosis in some cell lines and to stimulate growth of Raji Burkitt's lymphoma cell line.18, 19 In human T-lymphocytes, however, TIMP-2 has been reported to induce apoptosis.20

Recently, a functional polymorphism in the promoter region of TIMP-2 (−418) has been demonstrated.21 This −418G to C transition, situated at the consensus sequence for the Sp1-binding site, may downregulate transcriptional activity and result in an imbalance between the function of TIMP-2 and MMP-2. Common functional polymorphisms in the promoter region of MMP-2 have been demonstrated to influence the level of gene expression. At position −1306 of the MMP-2 gene promoter, a C to T commutation abolishes the Sp1-binding site, thus decreasing promoter activity. It has been confirmed that the CC genotype of MMP-2 (−1306C/T) may exhibit a high level of transcriptional and enzymatic activity. The functional significance of this polymorphism is linked to an individual's susceptibility to cancer.22 In addition, another C to T transition in the promoter region of MMP-2 (−735) has been identified.23 It has been reported that the transition of −735 C to T also disrupts a consensus sequence for the Sp1-binding site, manifesting in lower promoter activity.23 Several studies have revealed that the functional −1306C/T and −735C/T polymorphisms in MMP2 are associated with the risk of several cancers.3, 24–26

The activity of MMP-2 is potentially regulated by TIMP. The functional balance of MMP-2/TIMP-2 is potentially a key factor that affects tumor invasion and metastasis. Several reports have demonstrated that three polymorphisms: MMP-2 (−1306C/T and −735C/T) and TIMP-2 (−418G/C) are associated with the generation and progression of certain tumor types, but the effects of the three polymorphisms on NHL have not been elucidated. In this study we investigated the associations of these three functional gene polymorphisms in MMP-2 and TIMP-2 with various clinicopathologic characteristics and survival in order to evaluate the significance of these novel molecular markers in the progression and assessment of prognosis for patients with NHL.

Material and Methods

Study subjects

This study included 575 patients with confirmed NHL. The diagnosis of lymphoma was confirmed by histopathology and in most cases by supplementary immunohistochemistry analyses. The diagnoses of chronic lymphocytic leukemia (CLL) and acute lymphocytic leukemia (ALL) were based on cytology and cytometry. Cases were categorized according to the Revised European–American Lymphoma (REAL) Classification.27 An additional assessment was performed at a later date to classify the samples according to the World Health Organisation Classification.28 We evaluated B-cell lymphoma and T-cell lymphoma histologies. However, in 42 patients the immunohistological subtype was not obtained. All patients were ethnically Han Chinese. They were consecutively recruited to the study from January 2001 to January 2010 at the Department of Hematology, Fourth Hospital, Hebei Medical University (Shijiazhuang). All subjects were from North China (Shijiazhuang and its surrounding area). Five hundred thirty out of 575 patients were enrolled at the time of diagnosis, and 45 patients were enrolled during follow-up visits after treatment with chemotherapy. Those patients were treated for NHL at our hospital between 1991 and 2010. In the meantime, some similar resources exist in our department not included in this study. The overall participation rate was 70%. The Ethics Committee of Hebei Cancer Institute approved this study. Documented, informed consent was obtained from all enrolled subjects. There were 346 men and 229 women (M:F ratio, 1.51:1) with a median age of 51 years (range: 14–84 yr) at initial diagnosis. In 307 cases (53.4%), the primary tumor was localized in the lymph node, and in the other cases the origin was extranodal. Following diagnosis all patients underwent evaluation of clinical status, computerized tomography (CT) of the chest and the abdomen and bone marrow puncture or biopsy. Some patients underwent PET-CT (positron emitted tomography-computed tomography) procedures. The cancer stage was assessed according to Costwolds modification of the Ann Arbor staging method. Three hundred patients (48%) had systemic or B symptoms, including unexplained fever, night sweats and weight loss. Patients were treated in accordance with national guidelines; however, due to certain patients' financial problems their treatment protocols were adjusted. Patients were followed up through May 2010 or death. The 575 patients' characteristics are shown in Table 1.

Table 1. Clinical and pathologic characteristics of the 575 NHL patients at diagnosis
inline image

DNA extraction and genotyping

A peripheral blood sample (5 ml from each patient) was collected from each study participant into EDTA-blood Vacutainer tubes and stored at 4°C. Genomic DNA was extracted within 1 week from the blood samples using proteinase K (Merck, Darmstadt, Germany) digestion followed by a salting-out procedure29 and stored at −70°C until SNP analysis.

MMP-2 −1306C/T, −735C/T and TIMP-2 −418G/C genotyping

The MMP-2 −1306C/T, −735C/T and TIMP-2 −418G/C polymorphisms were genotyped using polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) as previously reported.30 The primers used for amplification of the MMP2 promoter containing the −1306C/T site were 5′-CTTCCTAGGCTGGTCCTTACTGA-3′ (forward) and 5′-CTGAGACCTGAAGAGCTAAAGAGCT-3′ (reverse), resulting in a 193-bp fragment. The C allele contained a XspI restriction site and produced two bands (188 and 5 bp), but the T allele contained two XspI restriction sites and produced three fragments of 162, 26 and 5 bp. For the −735C/T site, 5′-GGATTCTTGGCTTGGCGCAGGA-3′(forward) And 5′-GGGGGCTGGGTAAAATGAGGCTG-3′ (reverse) resulted in a 391-bp fragment. The product was digested with HinfI (New England Biolabs, Beverly, MA). The −735 T allele formed two fragments of 338 and 53 bp, whereas the −735 C allele remained uncut (391 bp). For TIMP-2 −418G/C polymorphism analysis, the sites were 5′-CGTCTCTTGTTGGCTGGTCA-3′ (forward) and 5′-CCTT CAGCTCGACTCTGGAG-3′ (reverse). The products were digested with Eco88I (New England Biolabs, Beverly, MA). The −418C allele contained one Eco881 restriction site and produced two fragments (253 and 51 bp) yet the −418 G allele contained two Eco881 restriction sites and resulted in three fragments of 230, 51 and 23 bp. The PCR experiments were performed in a 20 μl final volume consisting of 100 ng of DNA template, 200 nM of each primer, 0.4 μl of 10 mmol/l deoxyribonucleotide triphosphates, 1 U of Taq DNA polymerase and 2.4 μl of 10× PCR buffer. The PCR was performed under the following conditions29: 35 cycles at 94°C for 5 minutes, followed by 45 sec at 94°C, 45 sec at 58°C for −1306C/T, 65.5°C for −735C/T and 62°C for −418G/C and 45 sec at 72°C, with a final step at 72°C for 10 minutes to allow for the complete extension of all PCR fragments. Every PCR product, using an 8-μl aliquot, was digested at 37°C overnight in a 10-μl reaction containing 10 IU of the respective restriction enzyme. The digestion products were separated on 4% agarose gels that were stained with ethidium bromide.

Statistical analysis

All statistical analyses were performed with the Social Sciences software package (version 11.5; SPSS, Chicago, IL). Chi-square tests were used to compare the genotype distributions of MMP-2 (−1306C/T and −735C/T) and TIMP-2 (−418G/C) in the study groups. The odds ratio (OR) and 95% confidence interval (CI) were calculated for the genotype clinicopathologic classification using unconditional logistic regression. The covariates were age, sex, lymphoma origin, immunohistological subtype, Ann Arbor stage, bone marrow infiltration, systemic or B symptoms and tumor size. In this study, overall survival was selected as an end-point. Survival time was calculated from the date of diagnosis to the last follow-up date or the time of death related to NHL. The Kaplan–Meier method was used for three-year survival analysis. The test for differences in survival by MMP-2 (−1306C/T and −735C/T) and TIMP-2 (−418G/C) genotype, and log-rank analysis was performed. The influence of MMP-2 (−1306C/T and −735C/T) and TIMP-2 (−418G/C) genotypes on survival were investigated using Cox's proportional hazard model. Haplotype frequencies were estimated using EH (EH-plus) software. Chi-square analysis was used to compare the distribution of haplotypes. For all experiments, a value of p ≤ 0.05 was considered to indicate a statistically significant result.

Results

MMP2 and TIMP2 genotype distribution and characteristics in NHL patients

The frequencies of the genotypes were 69.74% (CC), 24.70% (CT) and 5.56% (TT) for MMP2 −1306 C/T; 63.66% (CC), 30.43% (CT) and 5.91% (TT) for MMP2 −735C/T; and 65.22% (GG), 30.26% (GC) and 4.52% (CC) for TIMP-2 −418 G/C; all exhibited Hardy–Weinberg equilibrium (p = 0.110, p = 0.385 and p = 0.677, respectively). The associations of clinicopathological parameters of NHL patients (n = 575) at diagnosis with the MMP2 and TIMP2 genotypes were evaluated. No significant relationships between the investigated polymorphisms of MMP-2 −1306C/T, −735C/T and TIMP-2 −418G/C and the clinicopathological variables such as age, sex, B symptoms, lymphomas arised and histopathological types at diagnosis were demonstrated (data not shown). The associations of the MMP2 and TIMP2 genotypes with Ann Arbor clinical staging are displayed in Table 2. The CT and TT+CT genotypes of −735C/T were significantly more frequent in patients with advanced clinical stages III and IV compared to stages I and II (OR, 1.60; 95%CI, 1.08–2.38; p = 0.020 and OR, 1.64; 95%CI, 1.13–2.34; p = 0.010, respectively). The −735T allele was significantly more frequent in patients with stage III and IV compared to patients with stages I and II (OR, 1.53; 95%CI, 1.11–2. 11; p = 0.009). None of the evaluated genotypes of −1306 C/T and −418 G/C was significantly associated with clinical staging. The relationships of MMP2 and TIMP2 polymorphisms with bone marrow infiltration and gross tumor volume at diagnosis are revealed in Table 3. The TT genotypes of the −1306C/T were significantly less frequent in patients with bone marrow infiltration compared to CC genotypes (OR, 0.13; 95%CI, 0.02–0.93; p = 0.013). However, a correlation between the −735C/T and −418G/C polymorphisms and bone marrow infiltration was not observed. The CT, TT and CT+TT genotypes of −1306C/T were significantly less frequent in patients with a large tumor size (>5 cm) compared to a smaller tumor (≤5 cm) (OR, 0.47; 95%CI, 0.29–0.79; p = 0.004; OR, 0.10; 95%CI, 0.01–0.72; p = 0.005 and HR, 0.39; 95%CI, 0.23–0.66; p = 0.000, respectively). The −1306T allele was also less frequent in patients with a large tumor compared to a smaller tumor (OR, 0.38; 95%CI, 0.23–0.61; p = 0.000). The TT genotypes of −735 C/T were significantly more frequent in patients with a large tumor size compared to a smaller tumor (OR, 2.44; 95%CI, 1.17–5.11; p = 0.015). However, none of the genotypes of −418G/C were significantly associated with gross tumor volume and bone marrow infiltration.

Table 2. Association of the MMP-2 −1306C/T, −735C/T and TIMP-2 −418 G/C polymorphisms with Ann Arbor stage of NHL
inline image
Table 3. Association of MMP-2 −1306C/T, −735C/T and TIMP-2 −418 G/C polymorphisms with bone marrow involvement and tumor size in NHL
inline image

Association of MMP-2 −1306C/T and −735C/T haplotypes with the disease status of NHL

The combined effect of the MMP-2 −1306C/T and −735C/T gene polymorphisms on NHL was investigated using haplotype analysis. In comparison to the haplotype of −1306T/−735T, no significant associations between haplotypes and the clinicopathological parameters of age, sex, lymphoma origin, immunohistological subtype, Ann Arbor clinical stage and bone marrow infiltration were observed. However, the frequency of the −1306C/−735T haplotype in patients with smaller tumor size was significantly higher compared to patients with a larger tumor (OR, 2.72; 95%CI, 1.016–7.276; p = 0.040). The −1306T/−735C and −1306C/−735C haplotypes were significantly less frequent in patients with B symptoms compared to those without (OR, 0.459; 95%CI, 0.24–0.86; p = 0.014 and OR, 0.55; 95%CI, 0.31–0.96; p = 0.033, respectively).

Associations between the MMP2 and TIMP2 genotype and overall survival in NHL

Table 4 demonstrates the association of MMP2 and TIMP2 genotypes with overall survival. In this study the follow-up length were 180 months with a median follow-up period of 30.0 months. Twenty-four patients were lost to follow-up. There were 318 deaths among subjects who were included in the genotype study. In this analysis, the CT genotype of −735C/T demonstrated a statistically significant association with lower three-year overall survival (HR, 1.75; 95%CI, 1.15–2.65; p = 0.009). The −735C/T genotype overall survival Cox regression curve is displayed in Figure 1 and demonstrates a correlation with three-year overall survival in cases of NHL. Different genotypes of MMP-2 −735C/T were compared using the Log-rank test in Figure 1. No correlation was observed between −1306C/T and −418G/C polymorphisms and three-year overall survival in cases of NHL in our study.

Figure 1.

Cox regression overall survival curve for patients after diagnosis with NHL according to MMP-2 −735C/T polymorphisms. The overall survival rate in the carries of the TT genotype was significantly decreased compared to carriers of CC and CT genotypes (Log-rank test, p = 0.026).

Table 4. MMP2 and TIMP2 gene polymorphisms in association with NHL overall survival
inline image

Discussion

Lymphomas are comprised of a series of heterogeneous malignant diseases that originate from lymphoid tissue; there is a great deal of variability in clinical biological behavior exhibited by lymphomas.13 The prognosis of this tumor is diverse and is dependent upon the immunohistological subtype, clinical staging, programs of treatment and prognostic markers utilized. An increasing volume of evidence indicates that both tumor progression and the aggressive biological nature observed are related to MMPs, which currently serve as prognostic indicators in certain types of solid carcinomas, including lymphoma. Several SNPs in MMP-2 and TIMP2 have been demonstrated to affect the level of gene expression or protein function. In this study we evaluated the influence of MMP2 and TIMP2 promoter region gene polymorphisms on the clinical characteristics of NHL, as well as investigating whether the SNPs affect patient survival. We found that the genotypes of MMP-2 −1306C/T, −735C/T and TIMP-2 −418G/C were not significantly associated with the sex, age, B-symptoms, lymphoma origin and immunohistological classification of NHL at the time of diagnosis. We demonstrated that carriers of the variant genotype MMP2 −1306C/T were at an obviously reduced risk of bone marrow involvement and large tumor size. In contrast, patients carrying the variant genotype MMP2 −735C/T exhibited an increased risk for developing a large tumor and advanced stage of lymphoma. Moreover, we demonstrated that carrying the variant genotype of MMP2 −735C/T was associated with a decreased overall survival rate. To the best of our knowledge, no previous report has demonstrated the effects of these MMP2 and TIMP2 polymorphisms on NHL clinicopathologic variables and overall survival.

The functional polymorphism in MMP2 −1306C/T is related to susceptibility to certain tumor types. Price et al.23 reported a C to T transition at position −1306 in the MMP2 gene promoter. This SNP abolishes a Sp1-binding site and the T allele then exhibits markedly lower promoter activity. The polymorphic genotype CT or TT, and the T allele of MMP2 −1306C/T, exhibited correlation to lower MMP-2 protein circulation levels. It has been previously demonstrated that MMP-2 plays a critical role in promoting tumor angiogenesis and progression; thus, decreased expression is potentially related to a reduced cancer risk. Our results demonstrated that the TT genotype with the −1306C/T polymorphism was significantly lower in frequency in patients with infiltration of medullary bone compared to patients exhibiting a CC genotype. The CT, TT, CT+TT genotypes and T allele of the −1306C/T polymorphism were significantly less frequent in patients, exhibiting a large tumor size compared to patients with smaller tumors. Several previous findings revealed that the protective effect of the MMP2 −1306C/T polymorphism reduces the risk of tumor. O-Charoenrat et al.3 observed that the MMP2 −1306CC genotype exhibited significant correlation with tumor thickness >10 mm, nodal metastasis and advanced pathological stages in head and neck squamous cell carcinoma (HNSCC). Additionally, they found that individuals with the CC genotype of MMP2 −1306C/T exhibited a higher risk for induction of HNSCC and a more highly invasive disease. Miao et al.31, 32 suggested that the MMP2 −1306CC genotype correlates to the risk of carcinogenesis but not of metastases of gastric cardia, lung and breast cancers.

The −735C/T polymorphism is also located in the promoter region of the MMP-2 gene, and it exhibits a substantial influence on transcriptional activity30 The C to T transition of nucleotide −735 disables a Sp1-binding site, indicating that this SNP could potentially effect MMP-2 transcription.33 Previous studies have investigated the association between the MMP-2 −735C/T polymorphism and risk for disease, but the results have been discordant. Zhou et al.33, 34 noted that genotype CC of −735C/T was associated with increased MMP-2 expression and greater risk of invasion and metastasis of esophageal and lung cancer. In contrast, Rollin et al.35 demonstrated that no difference in −735C/T MMP-2 genotype was found between non-small cell lung cases and controls; the MMP-2 gene expression was lower in genotype −735CC compared to those with CT or TT in cancer patients. Moreover, the survival time was longer in patients who exhibited the MMP-2 −735T allele compared to those with the CC genotype. The relative risk of death increased 2.6-fold in the −735CC genotype. However, our data demonstrated different results: our findings clearly indicated that the CT and TT+ CT genotypes, as well as T allele −735C/T, were significantly associated with greater tumor cell invasion (stages III and IV; large tumor size) in NHL. Additionally, overall survival time was lower in patients with the MMP-2 −735T allele. Therefore, patients with the NHL variant genotype MMP2 −735C/T potentially endure a more aggressive disease. This discrepancy is most likely because there are different roles for the MMP-2 −735C/T polymorphism in different diseases.30 Furthermore, the genetic polymorphism in the promoter of MMP-1 (1G/2G) or MMP-3 (5A/6A), which also changes the transcriptional activity of the genes, may affect the occurrence or inhibition of some cancer types.36, 37

The −418G/C is a common, functional polymorphism located in the consensus sequence for the Sp-1 binding site. It has been suggested that the −418G/C transition in the promoter region of TIMP −2, which interrupts the Sp-1 binding site, may cause the down-regulation of promoter activity and further decrease the levels of TIMP-2.30 Because TIMP-2 is deemed a mesendogenous inhibitor of MMP-2, lower constitutive expression of TIMP-2 potentially stimulates the pre-cancerous parasitifers through the MMP-reliant pathway. O-Charoenrat et al.3 demonstrated that no relationship existed between the TIMP2 polymorphisms and the clinicopathological parameters of HNSCC; however, compared with the GG common allele, the temporarily augmented risk of cancer was associated with the variant allele (−418GC or CC). In concordance with the results of this study, Vairaktaris et al.25 also reported that the TIMP-2 −418G/C or C/C genotype could raise the risk of oral squamous cell carcinoma compared with the common G/G genotype. Nevertheless, some previous reports are not in concordance with those results. Wu et al.34 reported that TIMP-2 polymorphisms were not correlated with gastric cancer carcinogenesis. However, patients with the variant genotypes of TIMP-2 −418G/C exhibited a lower risk of serosal encroachment, lymphatic infestation, lymph node metastases and venous intrusion. Zhou et al.32 demonstrated that the variant genotype of TIMP-2 −418G/C was correlated with a measurably decreased risk of breast cancer, compared with the GG genotype. They suggested that lower constitutive expression of TIMP-2 may cause a host to be less susceptible to breast cancer. The reasons why the variant alleles are less impressionable and infestation to carcinosis may not be immediately apparent, maybe through the MMP-independent pathway.32 Additionally, among members of the TIMP family, TIMP-1 and TIMP-4 also exhibit promotional effects on malignant cells.38, 39 Those findings are not in concordance with our current study. We observed no considerable association between the TIMP-2 −418G/C polymorphism and clinical characteristics and overall survival rate of NHL.

Yu et al.33 found that patients with the −1306C/−735C haplotype exhibited over six-fold increased risk for development of esophageal squamous cell carcinoma and over three-fold increased risk for exhibiting distant metastasis of the cancer. However, the results of Zhou et al. suggested that the −1306C/−735C haplotype was associated with increased risk of developing lung cancer.34 These findings are in concordance with the functional analysis of the polymorphisms. They suggested that the haplotype may be a precise and powerful genetic marker for cancer risk assessment.33 However, Kang et al. reported no significant association of MMP-2 −1306C/T and −735C/T haplotypes and the risk of adenomyosis.31 We observed a significant correlation between the MMP-2 haplotype and disease status of NHL. Compared with the haplotype of −1306T/−735T, the frequency of the −1306C/−735T haplotype in patients with a smaller tumor size was significantly higher than in those with a larger tumor. The −1306T/−735C and −1306C/−735C haplotypes were significantly less frequent in patients with B-symptoms compared to those without. The reason for the difference between other studies and ours remains unknown. Resolution of this discrepancy in further studies is warranted.

In the individual tumor patients, different genetic backgrounds influence the frequencies of genotypes. A symphonious explanation for the correlation between the SNPs and the clinical biological behavior of the patients with malignant carcinomas remains to be elucidated.40 Despite of inconsistent results, it is generally accepted that genetic polymorphisms in the MMP-2 promoter and TIMP-2 gene are associated with cancer susceptibility and disease severity. The effect may occur via a polymorphism alone, or in combination with other polymorphisms. Our sample and study design have some limitations. The patients in our study were from only one hospital and there were no control subjects; thus, the subjects may not be representative of the general population. Much more work is required to confirm these findings in a large series of patients with NHL. Additional studies on other polymorphisms of MMPs and TIMPs are warranted.

Conclusion

We have demonstrated the allele and genotype frequencies and the relationship between MMP2 and TIMP2 promoter polymorphisms and clinical behavior in NHL patients. The genetically determined balance of MMP-2 and TIMP-2 potentially exhibits the capability to influence progression or prognosis in NHL. The MMP2 and TIMP2 polymorphisms, as potential markers for NHL, will allow for early recognition of people at high risk, will assist in the design of therapeutic programs and will aid in the appraisal of treatment effects.

Acknowledgements

This investigation was supported in part by the foundation of Hebei province University's Advantage and Characteristic Discipline, China. We would like to acknowledge Professor Li Yan and colleagues in the Department of Molecular Biology, Hebei Institute of Oncology, China, for their greatly appreciated assistance.

Ancillary