Influence of vascular endothelial growth factor single ... - Nature

Genes and Immunity (2002) 3, 229–232  2002 Nature Publishing Group All rights reserved 1466-4879/02 $25.00 www.nature.com/gene

BRIEF COMMUNICATION

Influence of vascular endothelial growth factor single nucleotide polymorphisms on tumour development in cutaneous malignant melanoma WM Howell1,2, AC Bateman3, SJ Turner1, A Collins1 and JM Theaker3 1

Department of Human Genetics, University of Southampton, Southampton, UK; 2Histocompatibility and Immunogenetics Laboratory, Southampton General Hospital, Southampton, UK; 3Department of Cellular Pathology, Southampton General Hospital, Southampton, UK

Vascular endothelial growth factor (VEGF) is a potent regulator of vasculogenesis and tumour angiogenesis. We have investigated whether the VEGF −2578, −1154, +405 and +936 SNPs and associated haplotypes confer susceptibility to and/or influence prognosis in cutaneous malignant melanoma (CMM) skin cancer. A total of 152 CMM patients and 266 controls were genotyped for VEGF promoter SNPs by ARMS-PCR. Strong linkage disequilibrium between the −2578, −1154 and +405 SNPs was detected (association, ␳ = 0.488–0.965), but not between these SNPs and SNP +936 (association, ␳ = 0.004–0.130). No SNPs or three SNP haplotypes (−2578, −1154, +405) were significantly associated with CMM, although a number of non-significant trends were observed. However, the VEGF −1154 AA genotype and −2578, −1154, +405 CAC haplotype were both significantly associated with less advanced (Stage 1) disease (P = 0.03). In addition, the VEGF −1154 AA genotype was associated with thinner primary vertical growth phase tumours (P = 0.002), while VEGF −1154 GG was associated with thicker primary tumours (P = 0.02). These preliminary results indicate that VEGF genotype may influence tumour growth in CMM, possibly via the effects of differential VEGF expression on tumour angiogenesis. Genes and Immunity (2002) 3, 229–232. doi:10.1038/sj.gene.6363851 Keywords: VEGF; SNPs; angiogenesis; malignant melanoma; cancer

Vascular endothelial growth factor (VEGF) is a potent endothelial cell-specific regulator of vasculogenesis and angiogenesis.1,2 Increased VEGF expression, resulting in VEGF-induced angiogenesis is associated with tumour growth and metastasis,3 since solid tumours cannot grow larger than 2–3 mm3 and cannot metastasise without the formation of new blood vessels. In such tumours, VEGF, whose expression is regulated by a number of hormones, growth factors and cytokines, such as interleukin (IL)-10, can be produced by tumour cells and tumour infiltrating T cells and macrophages.4,5 In cutaneous malignant melanoma (CMM), a number of studies have reported increased immunoreactivity for VEGF in neoplastic as opposed to benign melanocytic lesions,6 with elevated expression in metastatic melanoma.7 Tumour vascularity in metastatic melanoma has also been shown to be of prognostic value.7 In addition, in a murine model, it has been claimed that constitutive, genetically determined levels of VEGF expression may be more important than Correspondence: Dr WM Howell, Histocompatibility and Immunogenetics Laboratory, Tenovus Building, Southampton General Hospital, Tremona Road, Southampton SO16 6YD, UK. Email: wmh1얀soton.ac.uk This work was supported by a grant from the Association for International Cancer Research, St Andrews, Scotland (No. 99–121) Received 23 August 2001; revised 27 December 2001; accepted 3 January 2002

hypoxia-induced upregulation of VEGF in the angiogenesis of human melanoma cell line xenografts.8 Recently, a number of single nucleotide polymorphisms (SNPs) have been described in the VEGF gene (chrom 6p12), which have been reported to be associated with differential VEGF expression in vitro.9–12 Two of these SNPs (positions −2578 and −1154) are located in the VEGF promoter region,9,12 one SNP (+405) in exon 1 of the VEGF gene10 and a further SNP (+936) located in exon 8, corresponding to the 3′ untranslated (UTR) region of the gene.11 Other SNPs have also been described, although a relationship with VEGF expression has not been demonstrated.10 Based on the above, in this study we have determined whether the above VEGF-associated SNPs are associated with susceptibility to or prognosis in CMM. To achieve this, genotyping and haplotype analysis for VEGF −2578, −1154, +405 and +936 SNPs in a series of 152 CMM patients was performed, utilizing DNA extracted from diagnostic histopathological formalin-fixed, wax-embedded tissues.13 The following clinical and pathological markers were examined as indicators of prognosis: initial growth phase (horizontal vs vertical), mitotic count within vertical growth phase tumours, Breslow depth of invasive CMM and stage of disease.14 Among these factors, Breslow thickness at presentation is the most important single prognostic indicator in CMM.15–17

VEGF polymorphisms and malignant melanoma WM Howell et al

230

Patient survival was also determined according to VEGF genotype. In addition, VEGF genotypes and assigned haplotypes were compared between the CMM patients and a series of 266 unrelated controls (bone marrow and cadaveric solid organ donors), to determine whether VEGF genotype influences susceptibility to CMM. Linkage disequilibrium and haplotype analyses were performed according to previously published methods.18,19

Results and discussion All control samples were genotyped for VEGF −2578, and between 238 and 263 samples were genotyped for the remaining three SNPs, depending on DNA availability. Genotype and allele frequencies were in close agreement with those previously published for healthy Caucasian individuals.9–11 Similarly, all CMM samples were genotyped for VEGF −1154 and between 137 and 144 samples were genotyped for the remaining three SNPs, subject to DNA availability. There were no significant differences in VEGF genotype frequencies between CMM patients and controls at P = 0.05 (Table 1). In addition, no significant associations with tumour growth phase were demonstrated. However, results indicate that the VEGF −1154 AA genotype was decreased in frequency among CMM patients with Stage II–IV disease, compared with Stage I disease, at a marginal level of significance (3/55 (5.5%) vs 14/84 (16.7%); Fisher’s exact P = 0.03; OR = 0.3 (0.1– 1.1)) (VEGF −1154 AA has been shown to be associated with significantly lower VEGF expression than GA or GG genotypes12). VEGF genotypes were also compared with the frequency of mitotic figures in vertical growth phase tumours. No genotypes were significantly associated with mitotic count. Finally, VEGF genotypes were compared among patients with vertical growth phase tumours stratified according to Breslow thickness of their primary tumours (Table 2). At the clinically significant cut-off point of 0.75 mm, the VEGF −1154 AA genotype was significantly decreased in frequency among tumours with a thickness of ⭓0.75 mm (4/80 (5.0%) vs 7/22 (31.8%); Fisher’s exact P = 0.002; OR = 0.1 (0.03–0.4)). When stratifying tumours according to Breslow thickness of greater or less than 1.5 mm, the VEGF −1154 GG genotype was significantly increased in frequency among patients with thicker tumours (32/48 (66.7%) vs 24/54 (44.4%); P = 0.02; OR = 2.5 (1.1–6.0)). Insufficient CMM patients had tumours with a Breslow thickness exceeding 3.5 mm for a viable analysis at this cut-off point. In addition, mean vertical growth phase tumour Breslow thickness was significantly greater among CMM patients with VEGF −1154 GG vs AA genotypes (3.1 ± 3.9 mm vs 0.8 ± 0.5 mm; Mann-Whitney U test, P = 0.002). Mean tumour thickness in CMM patients of VEGF −1154 GA genotype was of an intermediate value (2.1 ± 2.2 mm). Since VEGF −1154 GG has been shown to be associated with high VEGF expression compared with VEGF −1154 AA,12 these results suggest that genotypes associated with increased VEGF expression are also associated with increased tumour Breslow thickness. Only individuals typed for all four SNPs were included in the haplotype analysis (114 CMM patients and 224 controls). In this combined series, haplotype analysis revealed that the VEGF −2578, −1154 and +405 SNPs are in tight linkage disequilibrium (pairwise linkage disequilibrium coefficient ␳ = 0.488–0.965), while the VEGF +936 Genes and Immunity

Table 1 VEGF SNP frequencies in CMM cases and controls Genotype

CMM Cases n(%)

Controls n(%)

P

54 (40.3) 49 (36.6) 31 (23.1) n = 134

82 (30.8) 115 (43.2) 69 (25.9) n = 266

0.08

83 (54.6) 52 (34.2) 52 (34.2) 17 (11.2) n = 152

120 (45.6) 109 (41.4) 109 (41.4) 34 (12.9) n = 263

0.07

VEGF +405 GG VEGF +405 GC VEGF +405 CC

63 (46.0) 56 (40.9) 18 (13.1) n = 137

121 (50.2) 90 (37.3) 30 (12.4) n = 241

VEGF +936 TT VEGF +936 TC VEGF +936 CC

4 (2.8) 28 (19.4) 112 (77.8) n = 144

2 (0.8) 54 (22.7) 182 (76.5) n = 238

VEGF −2578 CC VEGF −2578 CA VEGF −2578 AA VEGF VEGF VEGF VEGF

−1154 GG −1154 GA −q1154 GA −1154 AA

In Tables 1 and 2 P values and odds ratios are given where P ⭐ 0.05. P values for non-significant trends (P = 0.05–0.10) are also given. All other P values were non-significant. Genotype distributions were compared by 2 × 2 tables using ␹2 analysis or Fisher’s exact test if one or more variables in 2 × 2 tables were ⬍5. PCR genotyping reactions (10 ␮l) comprised 1× AS reaction buffer (ABgene, Epsom, UK), 200 ␮M each dNTP, 12% (w/v) sucrose, 200 ␮M cresol red, 1 ␮M each specific/common primer (5′TAGGCC AGACCCTGGCAC3′ or 5′TAGGCCAGACCCTGGCAA3′ with 5′T GCCCCAGGGAACAAAGT3′ (−2578); 5′GCCCGAGCCGCGTGTG GAG3′ or 5′GCCCGAGCCGCGTGTGGAA3′ with 5′CCCCGCTAC CAGCCGACTT3′ (−1154); 5′CGTGCGAGCAGCGAAAGG3′ or 5′C GTGCGAGCAGCGAAAGC3′ with 5′TGTCCGTCAGCGCGACT G3′ (+405); 5′GGTCGGGTGACCCAGCAC3′ or 5′GGTCGGGTGAC CCAGCAT3′ with 5′GGGTGGGTGTGTCTACAGGA3′ (+936)), 0.2 ␮M each internal control primer22 (5′TGCCAAGTGGAGCACC CAA3′ and 5′GCATCTTGCTCTGTGCAGAT3′), 0.25 units ThermoprimePLUS DNA polymerase (ABgene) and 25–100 ng DNA. MgCl2 concentrations were optimised for each SNP (2.0 mM for +405 and +936; 2.75 mM for −2578 and −1154). PCR conditions comprised 1 min at 96°, followed by 10 cycles of 96° for 15 s, Tao for each SNP (61°C for −1154; 65 °C for +405 and +936; 65.5°C for −2578) for 50 s, 72°C for 40 s; then 20 cycles of 96°C for 10 s, 60°C for 50 s, 72°C for 40 s. (9600 Thermal Cycler (ABI, Foster City, CA, USA) or an MWG Primus 96Plus Thermal Cycler (MWG Biotech UK Ltd, Milton Keynes, UK).) PCR products were loaded directly onto 2% agarose gels (containing 0.5 mg/ml ethidium bromide), electrophoresed and visualised by photography under UV transillumination.

SNP is only weakly associated with the other SNPs (␳ = 0.004–0.130). Given this, combined with the rarity of the VEGF +936 TT genotype (Table 1), further haplotype analysis was restricted to the −2578, −1154 and +405 SNPs. From Table 3 it can be seen that of a possible eight haplotypes, one haplotype (VEGF −2578, −1154, +405 AGC) was not represented among CMM cases or controls, while a further three haplotypes (AAC, CAG and CAC) had a frequency of less than 0.1 in both patients and controls. No significant differences in haplotype frequencies between CMM patients and controls were detected. These three SNP haplotypes were also compared within the patient series stratified according to clinical and histopathological prognostic indictors, by logistic regression and Breslow thickness by stepwise regression.


231

Table 2 VEGF SNPs in vertical growth phase CMM patients according to Breslow depth of the primary tumour Genotype

Breslow depth: Breslow depth ⭐0.74 mm n(%) ⭓0.75 mm n(%)

P

OR (95% CI)

Breslow depth ⭐1.4 mm n(%)

Breslow depth ⭓1.5 mm n(%)

VEGF −2578 CC VEGF −2578 CA VEGF −2578 AA

6 (42.9) 7 (50.0) 1 (7.1) n = 14

32 (43.9) 24 (32.9) 17 (23.3) n = 73

14 (35.9) 17 (43.6) 8 (20.5) n = 39

24 (50.0) 14 (29.2) 10 (20.8) n = 48

VEGF −1154 GG VEGF −1154 GA VEGF −1154

9 (41.0) 6 (27.3) 7 (31.8) n = 22

47 (58.8) 29 (36.3) 4 (5.0) n = 80

24 (44.4) 21 (38.9) 9 (16.7) n = 54

32 (66.7) 14 (29.2) 2 (4.2) n = 48

VEGF +405 GG VEGF +405 GC VEGF +405 CC

10 (45.5) 10 (45.5) 2 (9.0) n = 22

32 (45.1) 29 (40.8) 10 (14.1) n = 71

26 (53.1) 18 (36.7) 5 (10.2) n = 49

16 (36.4) 21 (47.7) 7 (15.9) n = 44

VEGF +936 TT VEGF +936 TC VEGF +936 CC

1 (5.0) 3 (15.0) 16 (80.0) n = 20

2 (2.7) 15 (20.0) 58 (77.3) n = 75

2 (4.3) 7 (14.9) 38 (80.9) n = 47

1 (2.1) 11 (22.9) 36 (75.0) n = 48

*0.002

0.1 (0.03–0.4)

P

OR (95% CI)

0.02

2.5 (1.1–5.6)

*0.06

*Fishers exact P. Table 3 VEGF −2578, −1154, +405 haplotypes in CMM cases and controls Haplotype −2578, −1154, +405 AAG AAC AGG AGC CAG CAC CGG CGC

Frequency in CMM cases (n = 114)

Frequency in controls (n = 224)

0.207 0.000 0.211 0.000 0.016 0.045 0.221 0.300

0.265 0.008 0.212 0.000 0.023 0.041 0.196 0.255

P = non-significant for all comparisons.

From this analysis stage of disease was shown to be associated with the rare VEGF −2578, −1154, +405 CAC haplotype (␹2 = 4.659, 1 df, P = 0.03), with this haplotype associated with less advanced disease. Finally, analysis of patient survival was performed in 108 subjects in which the primary tumour was fully excised by surgery. This revealed a non-significant trend towards improved survival among individuals of VEGF −1154 AA genotype compared with VEGF −1154 GA and GG genotypes (Log rank P = 0.2915) (Figure 1). This is the first study to investigate whether SNPs reported to be associated with differential VEGF expression influence cancer susceptibility and tumour development, with CMM as exemplar. This is also the first study to investigate linkage disequilibrium and probable haplotypes for the four SNPs in question. Results obtained show evidence for strong linkage disequilibrium between the VEGF −2578, −1154 and +405 SNPs, but not between these SNPs and SNP +936. The reason for this lack of strong linkage disequilibrium between SNP +405 (exon 1) and SNP +936 (exon 8) is not clear. Findings also suggest that it is unlikely that any of the four VEGF SNPs or associated haplotypes investigated play a major role in susceptibility to CMM,

Figure 1 Patient survival in CMM series according to VEGF −1154 genotype.

although this cannot be entirely discounted by the present small study. The most significant finding in the present study is that VEGF −1154 genotype appears to be associated with CMM tumour Breslow thickness, with the VEGF −1154 GG ‘high expression’ genotype found at a significantly higher frequency in CMM patients with a vertical growth phase tumour thickness equal to or greater than 1.5 mm compared with patients with thinner tumours. In addition, the VEGF −1154 AA ‘low expression’ genotype was found at a significantly higher frequency in CMM patients with vertical growth phase tumours of less than 0.75 mm compared with patients with thicker tumours. These findings are supported by further data analysis demonstrating that mean Breslow tumour thickness is significantly greater among CMM patients of VEGF −1154 GG ‘high expression’ genotype compared with patients of VEGF −1154 AA ‘low expression’ genotype. Of the remaining prognostic indicators examined, the VEGF −1154 AA ‘low expression’ genotype was found at a significantly increased frequency among CMM patients with Stage I as opposed to Stage II–IV disease. The findGenes and Immunity


232

ing is supported by haplotype analysis showing that the VEGF −2578, −1154, +405 CAC haplotype was associated with less advanced stage of disease. All of the above associations were significant before correction of P values for the number of genotypes compared (three genotypes for each of four SNPs). The VEGF −1154 AA association with thinner primary tumours remains significant after correction for the number of genotypes. In addition, there was a non-significant trend towards improved survival among patients of VEGF −1154 AA ‘low expression’ genotype, consistent with a direct influence of polymorphisms affecting VEGF expression on tumour growth. Taken together, these results suggest that the VEGF −1154 SNP – or polymorphisms in linkage disequilibrium with this SNP – may influence tumour invasion in CMM, possibly via increased expression of VEGF, acting to promote tumour angiogenesis. We have previously reported that IL-10 promoter SNPs are associated with susceptibility to CMM and with tumour growth phase, Stage of disease and Breslow thickness in this patient series and have hypothesised that these associations may be due to IL-10 downregulation of a number of angiogenic factors, including VEGF.13 Other explanations of IL-10 action, including downregulation of HLA class I expression and increased NK cell lysis remain plausible.20,21 An investigation of combined VEGF and IL-10 genotypes may therefore be informative, which could be addressed by a larger study than the present. Results from this preliminary study therefore suggest that VEGF gene polymorphisms (which may be associated with differential VEGF expression) may influence tumour development in CMM. This preliminary finding requires confirmation in an independent study of a large CMM patient series, along with a definitive investigation of VEGF genotype influences on patient survival in this malignancy.

6

7

8

9

10

11

12

13

14

15 16

17

References 1 Leung DW, Cachianes G, Kuang WJ, Goeddel DV, Ferrara N. Vascular endothelial growth factor is a secreted angiogenic mitogen. Science 1989; 246: 1306–1309. 2 Tischer E, Mitchell R, Hartman T et al. The human gene for vascular endothelial growth factor. Multiple protein forms are encoded through alternative exon splicing. J Biol Chem 1991; 266: 11947–11954. 3 Claffey KP, Robinson GS. Regulation of VEGF/VPF expression in tumour cells: consequences for tumour growth and metastasis. Cancer Metastasis Rev 1996; 15: 165–176. 4 Berse B, Brown LF, Van de Water L, Dvorak HF, Senger DR. Vascular permeability factor (vascular endothelial growth factor) gene is expressed differently in normal tissues, macrophages and tumors. Mol Biol Cell 1992; 3: 211–220. 5 Freeman MR, Schneck FX, Gagnon ML. Peripheral blood lymphocytes infiltrating human cancers express vascular endo-

Genes and Immunity

18 19 20

21

22

thelial growth factor: a potential role for T cells in angiogenesis. Cancer Res 1995; 55: 4140–4145. Bayer-Garner IB, Hough AJ, Smoller MD. Vascular endothelial growth factor expression in malignant melanoma: prognostic versus diagnostic usefulness. Mod Pathol 1999; 12: 770–774. Vlaykova T, Laurila P, Muhonen T et al. Prognostic value of tumour vascularity in metastatic melanoma and association of blood vessel density with vascular endothelial growth factor expression. Melanoma Res 1999; 9: 59–68. Danielsen T, Rofstad EK. The constitutive level of vascular endothelial growth factor (VEGF) is more important than hypoxia-induced VEGF up-regulation in the angiogenesis of human melanoma xenografts. Br J Cancer 2000; 82: 1528–1534. Brogan IJ, Khan N, Isaac K, Hutchinson JA, Pravica V, Hutchinson IV. Novel polymorphisms in the promoter and 5′ UTR regions of the human vascular endothelial growth factor gene. Hum Immunol 1999; 60: 1245–1249. Watson CJ, Webb NJA, Bottomley MJ, Brenchley PEC. Identification of polymorphisms within the vascular endothelial growth factor (VEGF) gene: correlation with variation in VEGF protein production. Cytokine 2000; 12: 1232–1235. Renner W, Kotschan S, Hoffmann C, Obermayer-Pietsch B, Pilger E. A common 936 C/T mutation in the gene for vascular endothelial growth factor is associated with vascular endothelial growth factor plasma levels. J Vasc Res 2000; 37: 443–448. Shahbazi M, Fryer AA, Pravica V, Brogan IJ, Ramsay HM, Hutchinson IV. Polymorphisms in vascular endothelial growth factor gene are associated with increased risk of acute rejection in renal transplant recipients. J Am Soc Nephrol 2002; 13: 260–264. Howell WM, Turner SJ, Bateman AC, Theaker JM. IL-10 promoter polymorphisms influence tumour development in cutaneous malignant melanoma. Genes Immun 2001; 2: 25–31. Clark WH Jr, Elder DE, Guerry D IV et al. Model predicting survival in stage I melanoma based on tumour progression. J Natl Cancer Inst 1989: 81: 1893–1904. Mackie RM, Smyth JF, Soutar DS et al. Malignant melanoma in Scotland 1979–1983. Lancet 1985; 2: 859. Mackie R, Hunter JA, Aitchison TC et al. Cutaneous malignant melanoma, Scotland, 1979–1983. The Scottish Melanoma Group. Lancet 1992; 339: 971–975. Vollmer RT, Seigler HF. Using a continuous transformation of the Breslow thickness for prognosis in cutaneous melanoma. Am J Clin Pathol 2001; 115: 205–212. Collins A, Morton NE. Mapping a disease locus by allelic association. Proc Natl Acad Sci USA 1998; 95: 1741–1745. Terwilliger JD, Ott J. Handbook of Human Genetic Linkage. The Johns Hopkins University Press; Baltimore, 1994. Yue FY, Dummer R, Geertsen R et al. Interleukin-10 is a growth factor for human melanoma cells and down-regulates HLA class-I, HLA class-II and ICAM-1 molecules. Int J Cancer 1997; 71: 630–637. Zheng LM, Ojcius DM, Garaud F et al. Interleukin-10 inhibits tumor metastasis through an NK cell-dependent mechanism. J Exp Med 1996; 184: 579–584. Olerup O, Zetterquist H. HLA-DR typing by PCR amplification and sequence-specific primers (PCR-SSP) in 2 hours: an alternative to serological DR typing in clinical practice including donor-recipient matching in cadaveric transplantation. Tissue Antigens 1992; 39: 225–235.