- Email: [email protected]
Promoter polymorphisms of the pigment epithelium-derived factor gene are associated with diabetic retinopathy
Biochemical and Biophysical Research Communications 361 (2007) 421–426 www.elsevier.com/locate/ybbrc
Promoter polymorphisms of the pigment epithelium-derived factor gene are associated with diabetic retinopathy Hiroyuki Iizuka a, Takuya Awata a,b,*, Masataka Osaki b, Tamotsu Neda b, Susumu Kurihara b, Kiyoaki Inoue b, Kouichi Inukai b, Sho Kabasawa c, Keisuke Mori c, Shin Yoneya c, Shigehiro Katayama b b
a Division of RI Laboratory, Biomedical Research Center, Saitama Medical University, Saitama, Japan Division of Endocrinology and Diabetes, Department of Medicine, Saitama Medical University, Saitama, Japan c Department of Ophthalmology, Saitama Medical University, Saitama, Japan
Received 23 June 2007 Available online 16 July 2007
Abstract Pigment epithelium-derived factor (PEDF or SERPINF1), a neuroprotective and anti-angiogenic factor, may play an important role in the pathogenesis of diabetic retinopathy (DR). In 416 patients with type 2 diabetes, four polymorphisms in the PEDF SNPs were identified, rs12150053 and rs12948385 in the promoter region, rs9913583 in the 5 0 -untranslated region, and rs1136287 (Met72Thr) in exon 3. Based on case-control studies, rs12150053 and rs12948385, but not rs9913583 and rs1136287, were significantly associated with DR. A logistic regression analysis revealed that the TC or CC genotype of rs12150053 was a significant risk factor for DR (odds ratio 2.40, p = 0.0004). The GA or AA genotype of rs12948385 was also a significant risk factor for DR. In addition, a significant interaction between the vascular endothelial growth factor (VEGF) and PEDF SNPs in the susceptibility to DR was found. These results demonstrate that the PEDF gene, in cooperation with the VEGF gene, may contribute to the development of DR. 2007 Elsevier Inc. All rights reserved. Keywords: Pigment epithelium-derived factor gene; Diabetic retinopathy; Diabetic macular edema; Vascular endothelial growth factor; Interaction
Diabetic retinopathy (DR), which frequently occurs in patients with both type 1 and type 2 diabetes, is a central cause of visual impairment in adults. Therapeutic options for the severe form of this condition (proliferative diabetic retinopathy, PDR) and diabetic macular edema (DME), which may occur at any stage of DR, are limited [1–3]. Therefore, it is important to identify the genetic factors that determine susceptibility to DR, to gain insight into Abbreviations: DR, diabetic retinopathy; PDR, proliferative diabetic retinopathy; DME, diabetic macular edema; VEGF, vascular endothelial growth factor; PEDF, pigment epithelium-derived factor; SNP, single nucleotide polymorphism; NPDR, non-proliferative diabetic retinopathy; OR, odds ratio. * Corresponding author. Address: Division of Endocrinology and Diabetes, Department of Medicine, Saitama Medical University, Saitama, Japan. Fax: +81 49 278 1834. E-mail address: [email protected] (T. Awata). 0006-291X/$ - see front matter 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2007.07.025
its pathogenesis and genetic risk factors for this condition. Most genetic studies have focused on the logical candidate genes for DR, such as aldose reductase and advanced glycation end products receptor, and several studies reported an association of these genes with DR, although further examination of these factors is required [4–6]. Previously, we have found that the C-634G SNP (rs2010963) in the 5 0 -untranslated region of the vascular endothelial growth factor (VEGF) gene was significantly associated with both DR and DME [7,8]; we have also provided evidence supporting the functional significance of the SNP [7–9]. Subsequently, rs2010963 was also found to be associated with giant cell arteritis [10], psoriasis [11], Behcet’s disease [12], proliferative retinopathy of prematurity [13], age-related macular degeneration [14] and PDR [15]. In the present study, we evaluated pigment epitheliumderived factor (PEDF or SERPINF1) as a potential
422
H. Iizuka et al. / Biochemical and Biophysical Research Communications 361 (2007) 421–426
susceptibility gene for DR. PEDF is widely expressed in the nervous system including the retina and in most tissues of the body, and it is an important neuromodulatory factor that can provide effective neuroprotection and a primary endogenous angiogenic inhibitor in the eye [16–20]. It is suggested that PEDF, as a counterpart of VEGF, is an antiinflammatory factor that may be actively involved in the pathogenesis of DR [3,21]. A decrease of PEDF levels was actually observed in the vitreous and aqueous humor from patients with DR [22,23]. Furthermore, PEDF knockout mice showed several retinal abnormalities, such as morphological alterations and increased microvessel density [24]. Therefore, PEDF is an attractive candidate gene for DR. Materials and methods Subjects. The present study included 416 type 2 diabetes patients. All of the subjects were unrelated and Japanese. The patients were recruited from the patients admitted to the Saitama Medical University Hospital in Saitama prefecture, Japan (age range: 15–83 yr). Written informed consent was obtained from each individual, and the study was approved by the Ethical Committee of Saitama Medical University and all procedures were conducted in accordance with the principles of the Declaration of Helsinki. Diagnosis, classification of diabetes and measurement of clinical features were carried out as previously reported [25]. Ophthalmologic evaluation. The funduscopic characteristics of patients were determined and graded by trained retinal specialists (S.Y. and K.M.), based on a grading scale similar to the international clinical diabetic retinopathy/macular edema disease severity scale (http://www.aao.org/aao/ education/library/international_DR.cfm), as reported previously [25]. Based on the severity grading, fluorescein angiography was performed on those patients with severe non-proliferative (NPDR) or PDR to confirm the funduscopic observations. Genotyping of the PEDF gene polymorphisms. The human PEDF gene consists of eight exons [19]. Eighteen patients were screened for variations in the promoter region and all exons and their boundary regions of the Table 1 Clinical characteristics of type 2 diabetic patients
n (male:female) Age (yr) Age at diagnosis (yr) Duration of disease (yr) Systolic BP (mmHg) Diastolic BP (mmHg) HbA1c (%) Cholesterol (mg/dl) Triglyceride (mg/dl) Insulin therapy
DR ()
DR (+)
p value
229 (112:117) 55 ± 15 48 ± 14 7±7 134 ± 20 78 ± 15 8.8 ± 2.3 196 ± 42 153 ± 118 24%
187 (96:91) 58 ± 11 46 ± 12 13 ± 8 138 ± 22 77 ± 14 8.9 ± 2.0 197 ± 51 148 ± 90 44%
NS 0.0066 NS <0.0001 0.0231 NS NS NS NS <0.0001
Data are n, means ± SD or %. BP, blood pressure; DR, diabetic retinopathy; NS, not significant.
PEDF gene by directly sequencing DNA fragments produced by PCR. This yielded the identification of 13 polymorphisms in the PEDF genes. All of these polymorphisms were included in the NCBI dbSNP database (http://www.ncbi.nlm.nih.gov/projects/SNP/index.html). Among them, we further analyzed two SNPs in the promoter region (rs12150053 and rs12948385), one in the 5 0 -untranslated region (rs9913583), and a nonsynonymous SNP in exon 3 (rs1136287; Glu298Asp) because they might affect the gene expression or the protein function; seven SNPs in introns and two synonymous SNP (Thr130Thr and Tyr321Tyr) were excluded in the present study. The genotyping of each polymorphism was carried out using polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) analysis. The PCR primers for rs12150053, rs12948385, rs9913583 and rs1136287 were 5 0 -TTGAGACAAGCGTGA CCAATGT-3 0 (forward) and 5 0 -TTCCCGTCACCATAACC-3 0 (reverse); 5 0 -CTACAGGTGCGCGCCAAC-3 0 (forward) and 5 0 -CTAAGGCGCA CACATTTGCACA-3 0 (reverse); 5 0 -ATCACAGAATCCAAATTGAGT GCAG-3 0 (forward) and 5 0 -TCTTTTCTGCTCCCTGGAGTGCCCC-3 0 (reverse); and 5 0 -GACAGTCCCTGTGCATCTCTGCGAG-3 0 (forward) and 5 0 -TCAGCCACGTTTACGCAGAGGCCAC-3 0 (reverse), respectively. The rs12150053 C allele, the rs12948385 A allele, the rs9913583 A allele, and the rs1136287 C allele result in the gain of a BsrI site, a HincII site, a BglII site, and a BssSI site, respectively. After digestion by an appropriate restriction enzyme described above, the PCR products were separated by electrophoreses on a 2.5–3.0% agarose gel and then visualized using ethidium bromide staining. Genotyping of the VEGF gene polymorphism and interaction between VEGF and PEDF SNPs. The VEGF C-634G SNP (rs2010963) was genotyped as described previously [8]. The interaction of the VEGF rs2010963 and the PEDF rs12150053 was evaluated using a logistic regression analysis; i.e. the VEGF rs2010963 (CC = 1, CG or GG = 0; recessive model), the PEDF rs12150053 (TC or CC = 1, TT = 0; dominant model) and the product of them (CC/TC or CC/CC = 1, others = 0), as an interactive term, were included together as explanatory variables in a multivariate logistic regression model, and then the interactive effects were considered to be represented by the increased OR for the product. Statistical analysis. Continuous clinical data were compared using the unpaired Student’s t-test, and categorical clinical data were compared using Fisher’s exact test. Multivariate logistic regression analyses were performed to assess the independent role of the genotype of SNPs and other variables, including sex, age at onset, duration of diabetes, systolic blood pressure, HbA1c level, cholesterol level, and treatment of diabetes (categorical variable: insulin therapy or no insulin therapy). StatView Ver.5 (SAS Institute; Cary, NC) was used for these tests. Hardy–Weinberg equilibrium, linkage disequilibrium, case-control study, and case-control haplotype analysis for multiple loci were assessed using the SNPAlyze Ver.6 software (Dynacom, Yokohama, Japan). Statistical significance was defined as p < 0.05.
Results Clinical characteristics Among the 416 type 2 diabetic patients, 229 patients had no evidence of retinopathy and the remaining 187 had DR.
Table 2 Pairwise linkage disequilibrium between PEDF SNPs SNP
rs12150053 rs12948385 rs9913583 rs1136287
Major/minor allele
T/C G/A C/A C/T
Position from transcription start site
Position from translation start site
790 358 +72 +8028
5740 5308 4879 +3078
r2 (upper)/D 0 (lower) rs12150053
rs12948385 0.919
0.974 1 0.962
1 0.907
rs9913583
rs1136287
0.035 0.037
0.215 0.198 0.163
0.934
0.0067 0.0081
0.0027
NS
423
0.0045 0.0048
NS
Table 1 summarizes the clinical characteristics of the diabetic patients included in the study. Among 187 patients with DR, 101 patients had non-proliferative retinopathy, and 86 had PDR; DME was present in 66 patients.
0.0015
p value
H. Iizuka et al. / Biochemical and Biophysical Research Communications 361 (2007) 421–426
NS NS — 1.33 (0.84–2.12) 0.0229 0.0106
NS
1.73 (1.14–2.60) 2.50 (0.62–10.12) — 1.61 (1.13–2.31) Codominant Allele
Recessive
112:69:6 G/A rs12948385
165:61:3
Dominant
Codominant Allele
Recessive
Dominant 167:60:2
NS
1.53 (0.90–2.58) –– 54:32:0 0.0120
0.0110 0.0057
NS
0.0085
1.76 (1.17–2.67) 3.76 (0.75–18.86) — 1.68 (1.17–2.41) 113:68:6
OR (95% CI)
DR, diabetic retinopathy; PDR, proliferative diabetic retinopathy; DME, macular edema; OR, odds ratio; NS, not significant. p values are vs DR ().
2.43 (1.38–4.26) 1.16 (0.12–11.33) — 1.95 (1.21–3.12) NS NS
NS NS
55:31:0
— 1.35 (0.85–2.17)
NS
1.52 (0.90–2.57) —
NS
34:31:1
34:31:1
OR (95% CI) OR (95% CI) Genotypes (11:12:22) Genotypes (11:12:22)
p value
PDR DR (+)
Case
Model in carriers of minor allele Control
T/C
Previous studies demonstrated that the C-634G SNP (rs2010963) of the VEGF gene was associated with DR [7,8]. In the present study, a possible interaction between
rs12150053
Interaction between VEGF and PEDF in susceptibility to DR
DR () Genotypes (11:12:22)
To further assess the relationship of the PEDF polymorphisms with DR, multivariate logistic regression analysis was applied, including the patients’ clinical characteristics and the PEDF genotype. As shown in Table 5, the TC or CC genotype of rs12150053 was a significant risk factor for DR (dominant model, OR 2.40, p = 0.0004). The GA or AA genotype of rs12948385, which was in strong linkage disequilibrium with rs12150053, was also a significant risk factor for DR (data not shown).
Major/ minor allele
Logistic regression analysis
SNP
Inferred haplotype frequencies by EM algorithm of four SNPs were compared using permutation tests. As shown in Table 4, only the haplotype H2, which contains C at rs12150053 and A at rs12948385, increased significantly in both DR-positive patients compared with DR-negative patients, whereas haplotypes H1, H3 and H4, which contains T at rs12150053 and G at rs12948385, were decreased in those patients with DR.
Table 3 Case-control study of PEDF SNPs for diabetic retinopathy, diabetic proliferative retinopathy and diabetic macular edema
Case-control haplotype analysis of PEDF SNPs
p value
Case-control study of PEDF SNPs
Genotypes (11:12:22)
DME
All SNPs were in Hardy–Weinberg equilibrium. Pairwise D 0 and r2 were analyzed between four PEDF SNPs in patients with type 2 diabetes (Table 2). Strong linkage disequilibrium (r2 = 0.919) was observed between rs12150053 and rs12948385. Furthermore, weak linkage disequilibrium was observed between rs1136287 and the other three SNPs (r2 = 0.163–0.215).
Based on case-control association studies, 187 patients with DR (cases) and 229 patients without DR (controls), rs12150053 and rs12948385 were significantly associated with DR in the models of recessive, codominant and allele with C and A being a risk allele, respectively (Table 3), whereas no association with rs9913583 and rs1136289 (Met72Thr) was observed (data not shown). The association with rs12150053 or rs12948385 was stronger in the patients with DME, although this association was weaker and was not significant in those with PDR (Table 3).
2.54 (1.44–4.46) 1.75 (0.16–19.56) — 2.05 (1.23–3.30)
Linkage disequilibrium between SNPs
424
H. Iizuka et al. / Biochemical and Biophysical Research Communications 361 (2007) 421–426
Table 4 Case-control haplotype analysis of PEDF SNPs for diabetic retinopathy Haplotype
SNP
DR
rs12150053
rs12948385
rs9913583
rs1136287
()
(+)
p value
H1 H2 H3 H4
T C T T
G A G G
C C C A
C T T T
0.550 0.139 0.171 0.141
0.492 0.208 0.156 0.144
NS 0.0052 NS NS
DR, diabetic retinopathy; NS, not significant. p values are based on 10,000 permutations.
VEGF rs2010963 and PEDF rs12150053 in susceptibility to DR was evaluated. With a multivariate logistic regression analysis, a significant interaction between the VEGF and PEDF SNPs was observed (p = 0.0158, Table 6). The OR of the combined genotypes (VEGF rs2010963/PEDF rs12150053) is 7.06, that is greater than the product of the ORs of the individual genotypes (1.94 · 2.40 = 4.66), indicating the existence of a synergic effect of the VEGF and PEDF polymorphisms toward DR susceptibility. Discussion The present study evaluated a potential association of PEDF polymorphisms with DR in Japanese patients with type 2 diabetes. The results of the case-control studies demonstrated that two SNPs (rs12150053 and rs12948385) in the promoter region of PEDF were significantly associated with the presence of DR. Using logistic regression analysis, including several clinical characteristics such as disease duration, the genotypes of the SNPs were also significantly associated with an increased risk of DR. In addition, Table 5 Multivariate logistic regression analysis for the association with diabetic retinopathy Variable
OR (95% CI)
p value
Female sex Age at onset Duration of disease Systolic blood pressure HbA1c Cholesterol Insulin therapy PEDF rs12150053a
0.82 1.01 1.11 1.01 1.01 1.00 2.68 2.40
NS NS 2.56 · 109 0.0359 NS NS 0.0002 0.0004
(0.52–1.30) (0.99–1.03) (1.07–1.15) (1.00–1.03) (0.91–1.14) (1.00–1.01) (1.60–4.48) (1.48–3.91)
a Dominant model (TC or CC vs TT). OR, odds ratio; NS, not significant.
although we presented uncorrected p values, even with a conservative Bonferroni correction for multiple testing, the associations between the PEDF SNPs and DR remained significant. Previously, PEDF was analyzed as a candidate gene for two ocular diseases, Leber’s congenital amaurosis [26] and retinitis pigmentosa [27], but PEDF was excludes as a common cause of either of the disease. Yamagishi et al. analyzed the PEDF Met72Thr polymorphism (rs1136289) in 143 diabetic patients (10 type 1 and 133 type 2), but there were no significant correlations between PEDF Met72Thr and diabetic microangiopathy including DR [28]. Although the sample size of the study was small, the data was consistent with the present study, which showed no significant association between the Met72Thr SNP and DR. This is the first report that demonstrates a significant relation between PEDF polymorphisms and disease susceptibility. A significant association with DR was observed in rs12150053 and rs12948385 SNPs, which were in tight linkage disequilibrium each other. Case-control haplotype analysis suggested the primary role of these SNPs in disease susceptibility. Since these SNPs are located in the promoter region of PEDF and other variations were not found in the nearby region after screening of 18 patients, it is tempting to speculate that decreased PEDF expression in the retina induced by these SNPs, both or alone, contributes to the development of DR. The PEDF promoter region contains several putative binding sites for transcription elements and two Alu repetitive sequences [29]. Neither of the SNPs was not included in the putative transcription factor binding sites. We are now conducting experiments on the influence of the SNPs on the regulation of PEDF transcription and expression of the protein. Similarly to the observations in the VEGF C-634G [7], the PEDF SNPs were not specifically associated with the development of PDR. Although PEDF is a strong anti-
Table 6 Interaction between VEGF rs2010963 and PEDF r12150053 in the association with diabetic retinopathy SNP
DR () (n = 187)
DR (+) (n = 229)
OR (95% CI)a
p value
VEGF rs2010963 (CC vs CG or GG) PEDF rs12150053 (TC or CC vs TT) VEGF rs2010963 + PEDF rs12150053 (CC/TC or CC/CC vs others)b
15.7% (36) 27.1% (62) 3.9% (9)
21.9% (41) 39.6% (74) 13.9% (26)
1.94 (1.09–3.46) 2.40 (1.48–3.91) 7.06 (2.70–18.50)
0.0239 0.0004 6.95 · 105
a Odds ratios were adjusted by multivariate logistic regression analysis including sex, age at onset, duration, systolic blood pressure, HbA1c, cholesterol, and insulin therapy. b Interaction between VEGF and PEDF SNPs is significant by multivariate logistic regression analysis (p = 0.0158).
H. Iizuka et al. / Biochemical and Biophysical Research Communications 361 (2007) 421–426
angiogenic factor in the eye, which could contribute to the development of PDR, its neuroprotective effect may explain the association of the PEDF SNPs and the development of DR [16,17,19,30]. Furthermore, PEDF was recently shown to reduce the VEGF-induced vascular leakage, probably through the anti-inflammatory effect of PEDF [21,31]. Therefore, PEDF may also contribute to the development of DME. In fact, the present study showed that the association of the PEDF SNPs was stronger in DR patients complicated by DME. However, the number of patients with PDR or DME was not large; especially the number of PDR patients might have been insufficient in this study to detect a significant association. Two genes, VEGF and PEDF, are now found to be associated with DR. The present study assessed a possible interactive effect of VEGF rs2010963 and PEDF rs12150053 using logistic regression analysis. The patients with the combined risk genotypes of rs2010963 and rs12150053 have an increased risk of developing DR. The risk is significantly interactive, thus indicating a synergistic effect of the VEGF and PEDF SNPs. Accumulating evidence suggests that VEGF and PEDF have a reciprocal relation in the eye, and their imbalance, VEGF increase and PEDF decrease, may play an important role in the development of DR [3,21]. It is therefore possible that polymorphisms in VEGF and PEDF SNPs, leading to an imbalance of VEGF and PEDF expression, may jointly contribute to the development of DR. However, further genetic studies with larger sample size and functional studies may be necessary to clarify the exact role of the PEDF SNPs, and their possible interaction with VEGF SNPs in the development of diabetic retinal complications. Finally, based on the neuroprotective and anti-angiogenic functions, PEDF may play a role in the pathogenesis of neurodegenerative diseases and angiogenic disorders, such as Alzheimer’s disease, Parkinson’s disease, neovascular age-related macular degeneration and tumors. Furthermore, serum PEDF levels have recently been reported to be elevated in patients with metabolic syndrome [32]. Therefore, it may be worthwhile to analyze the possible association between PEDF polymorphisms and these diseases. Acknowledgments We thank Ritsuko Doki for her valuable technical assistance. This work was supported in part by grants from the Japanese Diabetes Foundation, the Medical Research Center, Saitama Medical University, and the Japanese Ministry of Education, Culture, Sports, Science and Technology. References [1] M. Porta, F. Bandello, Diabetic retinopathyA clinical update, Diabetologia 45 (2002) 1617–1634. [2] T.A. Ciulla, A.G. Amador, B. Zinman, Diabetic retinopathy and diabetic macular edema: pathophysiology, screening, and novel therapies, Diabetes Care 26 (2003) 2653–2664. [3] R.N. Frank, Diabetic retinopathy, N. Engl. J. Med. 350 (2004) 48–58.
425
[4] K.M. Warpeha, U. Chakravarthy, Molecular genetics of microvascular disease in diabetic retinopathy, Eye 17 (2003) 305–311. [5] C.L. Hanis, D. Hallman, Genetics of diabetic retinopathy, Curr. Diab. Rep. 6 (2006) 155–161. [6] K. Uhlmann, P. Kovacs, Y. Boettcher, H.P. Hammes, R. Paschke, Genetics of diabetic retinopathy, Exp. Clin. Endocrinol. Diabetes 114 (2006) 275–294. [7] T. Awata, K. Inoue, S. Kurihara, T. Ohkubo, M. Watanabe, K. Inukai, I. Inoue, S. Katayama, A common polymorphism in the 5 0 untranslated region of the VEGF gene is associated with diabetic retinopathy in type 2 diabetes, Diabetes 51 (2002) 1635–1639. [8] T. Awata, S. Kurihara, N. Takata, T. Neda, H. Iizuka, T. Ohkubo, M. Osaki, M. Watanabe, Y. Nakashima, K. Inukai, I. Inoue, I. Kawasaki, K. Mori, S. Yoneya, S. Katayama, Functional VEGF C634G polymorphism is associated with development of diabetic macular edema and correlated with macular retinal thickness in type 2 diabetes, Biochem. Biophys. Res. Commun. 333 (2005) 679–685. [9] D. Lambrechts, E. Storkebaum, M. Morimoto, J. Del-Favero, F. Desmet, S.L. Marklund, S. Wyns, V. Thijs, J. Andersson, I. van Marion, A. Al-Chalabi, S. Bornes, R. Musson, V. Hansen, L. Beckman, R. Adolfsson, H.S. Pall, H. Prats, S. Vermeire, P. Rutgeerts, S. Katayama, T. Awata, N. Leigh, L. Lang-Lazdunski, M. Dewerchin, C. Shaw, L. Moons, R. Vlietinck, K.E. Morrison, W. Robberecht, C. Van Broeckhoven, D. Collen, P.M. Andersen, P. Carmeliet, VEGF is a modifier of amyotrophic lateral sclerosis in mice and humans and protects motoneurons against ischemic death, Nat. Genet. 34 (2003) 383–394. [10] L. Boiardi, B. Casali, D. Nicoli, E. Farnetti, Q. Chen, P. Macchioni, M.G. Catanoso, L. Pulsatelli, R. Meliconi, C. Salvarani, Vascular endothelial growth factor gene polymorphisms in giant cell arteritis, J. Rheumatol. 30 (2003) 2160–2164. [11] H.S. Young, A.M. Summers, M. Bhushan, P.E. Brenchley, C.E. Griffiths, Single-nucleotide polymorphisms of vascular endothelial growth factor in psoriasis of early onset, J. Invest. Dermatol. 122 (2004) 209–215. [12] C. Salvarani, L. Boiardi, B. Casali, I. Olivieri, F. Cantini, F. Salvi, R. Malatesta, R. La Corte, G. Triolo, A. Ferrante, D. Filippini, G. Paolazzi, P. Sarzi-Puttini, D. Nicoli, E. Farnetti, Q. Chen, L. Pulsatelli, Vascular endothelial growth factor gene polymorphisms in Behcet’s disease, J. Rheumatol. 31 (2004) 1785–1789. [13] A. Vannay, G. Dunai, I. Banyasz, M. Szabo, R. Vamos, A. Treszl, J. Hajdu, T. Tulassay, B. Vasarhelyi, Association of genetic polymorphisms of vascular endothelial growth factor and risk for proliferative retinopathy of prematurity, Pediatr. Res. 57 (2005) 396–398. [14] J.L. Haines, N. Schnetz-Boutaud, S. Schmidt, W.K. Scott, A. Agarwal, E.A. Postel, L. Olson, S.J. Kenealy, M. Hauser, J.R. Gilbert, M.A. Pericak-Vance, Functional candidate genes in agerelated macular degeneration: significant association with VEGF, VLDLR, and LRP6, Invest. Ophthalmol. Vis. Sci. 47 (2006) 329–335. [15] F.I. Errera, L.H. Canani, M.E. Silva, E. Yeh, W. Takahashi, K.G. Santos, K.E. Souto, B. Tschiedel, I. Roisenberg, J.L. Gross, M.R. Passos-Bueno, Functional vascular endothelial growth factor 634G>C SNP is associated with proliferative diabetic retinopathy: a case-control study in a Brazilian population of European ancestry, Diabetes Care 30 (2007) 275–279. [16] P.G. Gettins, M. Simonovic, K. Volz, Pigment epithelium-derived factor (PEDF), a serpin with potent anti-angiogenic and neurite outgrowth-promoting properties, Biol. Chem. 383 (2002) 1677–1682. [17] J. Tombran-Tink, C.J. Barnstable, PEDF: a multifaceted neurotrophic factor, Nat. Rev. Neurosci. 4 (2003) 628–636. [18] C.J. Barnstable, J. Tombran-Tink, Neuroprotective and antiangiogenic actions of PEDF in the eye: molecular targets and therapeutic potential, Prog. Retin. Eye Res. 23 (2004) 561–577. [19] J. Tombran-Tink, The neuroprotective and angiogenesis inhibitory serpin, PEDF: new insights into phylogeny, function, and signaling, Front Biosci. 10 (2005) 2131–2149. [20] G.L. King, K. Suzuma, Pigment-epithelium-derived factor—a key coordinator of retinal neuronal and vascular functions, N. Engl. J. Med. 342 (2000) 349–351.
426
H. Iizuka et al. / Biochemical and Biophysical Research Communications 361 (2007) 421–426
[21] S.X. Zhang, J.J. Wang, G. Gao, C. Shao, R. Mott, J.X. Ma, Pigment epithelium-derived factor (PEDF) is an endogenous antiinflammatory factor, FASEB J. 20 (2006) 323–325. [22] N. Ogata, J. Tombran-Tink, M. Nishikawa, T. Nishimura, Y. Mitsuma, T. Sakamoto, M. Matsumura, Pigment epithelium-derived factor in the vitreous is low in diabetic retinopathy and high in rhegmatogenous retinal detachment, Am. J. Ophthalmol. 132 (2001) 378–382. [23] B.O. Boehm, G. Lang, O. Volpert, P.M. Jehle, A. Kurkhaus, S. Rosinger, G.K. Lang, N. Bouck, Low content of the natural ocular anti-angiogenic agent pigment epithelium-derived factor (PEDF) in aqueous humor predicts progression of diabetic retinopathy, Diabetologia 46 (2003) 394–400. [24] J.A. Doll, V.M. Stellmach, N.P. Bouck, A.R. Bergh, C. Lee, L.P. Abramson, M.L. Cornwell, M.R. Pins, J. Borensztajn, S.E. Crawford, Pigment epithelium-derived factor regulates the vasculature and mass of the prostate and pancreas, Nat. Med. 9 (2003) 774–780. [25] T. Awata, T. Neda, H. Iizuka, S. Kurihara, T. Ohkubo, N. Takata, M. Osaki, M. Watanabe, Y. Nakashima, T. Sawa, K. Inukai, I. Inoue, M. Shibuya, K. Mori, S. Yoneya, S. Katayama, Endothelial nitric oxide synthase gene is associated with diabetic macular edema in type 2 diabetes, Diabetes Care 27 (2004) 2184– 2190. [26] R. Koenekoop, A.L. Pina, M. Loyer, J. Davidson, J. Robitaille, I. Maumenee, J. Tombran-Tink, Four polymorphic variations in the PEDF gene identified during the mutation screening of patients with Leber congenital amaurosis, Mol. Vis. 5 (1999) 10.
[27] R. Goliath, J. Tombran-Tink, I.R. Rodriquez, G. Chader, R. Ramesar, J. Greenberg, The gene for PEDF, a retinal growth factor is a prime candidate for retinitis pigmentosa and is tightly linked to the RP13 locus on chromosome 17p13.3, Mol. Vis. 2 (1996) 5. [28] S. Yamagishi, S. Amano, Y. Inagaki, T. Okamoto, Y. Koda, M. Soejima, H. Kimura, Pigment epithelium-derived factor Met72Thr polymorphism in patients with diabetic microangiopathy, Int. J. Clin. Pharmacol. Res. 22 (2002) 67–71. [29] J. Tombran-Tink, K. Mazuruk, I.R. Rodriguez, D. Chung, T. Linker, E. Englander, G.J. Chader, Organization, evolutionary conservation, expression and unusual Alu density of the human gene for pigment epithelium-derived factor, a unique neurotrophic serpin, Mol. Vis. 2 (1996) 11. [30] U. Barnas, A. Schmidt, A. Illievich, H.P. Kiener, D. Rabensteiner, A. Kaider, R. Prager, H. Abrahamian, K. Irsigler, G. Mayer, Evaluation of risk factors for the development of nephropathy in patients with IDDM: insertion/deletion angiotensin converting enzyme gene polymorphism, hypertension and metabolic control, Diabetologia 40 (1997) 327–331. [31] H. Liu, J.G. Ren, W.L. Cooper, C.E. Hawkins, M.R. Cowan, P.Y. Tong, Identification of the antivasopermeability effect of pigment epithelium-derived factor and its active site, Proc. Natl. Acad. Sci. USA 101 (2004) 6605–6610. [32] S. Yamagishi, H. Adachi, A. Abe, T. Yashiro, M. Enomoto, K. Furuki, A. Hino, Y. Jinnouchi, K. Takenaka, T. Matsui, K. Nakamura, T. Imaizumi, Elevated serum levels of pigment epithelium-derived factor in the metabolic syndrome, J. Clin. Endocrinol. Metab. 91 (2006) 2447–2450.
Promoter polymorphisms of the pigment epithelium-derived factor gene are associated with diabetic retinopathy Hiroyuki Iizuka a, Takuya Awata a,b,*, Masataka Osaki b, Tamotsu Neda b, Susumu Kurihara b, Kiyoaki Inoue b, Kouichi Inukai b, Sho Kabasawa c, Keisuke Mori c, Shin Yoneya c, Shigehiro Katayama b b
a Division of RI Laboratory, Biomedical Research Center, Saitama Medical University, Saitama, Japan Division of Endocrinology and Diabetes, Department of Medicine, Saitama Medical University, Saitama, Japan c Department of Ophthalmology, Saitama Medical University, Saitama, Japan
Received 23 June 2007 Available online 16 July 2007
Abstract Pigment epithelium-derived factor (PEDF or SERPINF1), a neuroprotective and anti-angiogenic factor, may play an important role in the pathogenesis of diabetic retinopathy (DR). In 416 patients with type 2 diabetes, four polymorphisms in the PEDF SNPs were identified, rs12150053 and rs12948385 in the promoter region, rs9913583 in the 5 0 -untranslated region, and rs1136287 (Met72Thr) in exon 3. Based on case-control studies, rs12150053 and rs12948385, but not rs9913583 and rs1136287, were significantly associated with DR. A logistic regression analysis revealed that the TC or CC genotype of rs12150053 was a significant risk factor for DR (odds ratio 2.40, p = 0.0004). The GA or AA genotype of rs12948385 was also a significant risk factor for DR. In addition, a significant interaction between the vascular endothelial growth factor (VEGF) and PEDF SNPs in the susceptibility to DR was found. These results demonstrate that the PEDF gene, in cooperation with the VEGF gene, may contribute to the development of DR. 2007 Elsevier Inc. All rights reserved. Keywords: Pigment epithelium-derived factor gene; Diabetic retinopathy; Diabetic macular edema; Vascular endothelial growth factor; Interaction
Diabetic retinopathy (DR), which frequently occurs in patients with both type 1 and type 2 diabetes, is a central cause of visual impairment in adults. Therapeutic options for the severe form of this condition (proliferative diabetic retinopathy, PDR) and diabetic macular edema (DME), which may occur at any stage of DR, are limited [1–3]. Therefore, it is important to identify the genetic factors that determine susceptibility to DR, to gain insight into Abbreviations: DR, diabetic retinopathy; PDR, proliferative diabetic retinopathy; DME, diabetic macular edema; VEGF, vascular endothelial growth factor; PEDF, pigment epithelium-derived factor; SNP, single nucleotide polymorphism; NPDR, non-proliferative diabetic retinopathy; OR, odds ratio. * Corresponding author. Address: Division of Endocrinology and Diabetes, Department of Medicine, Saitama Medical University, Saitama, Japan. Fax: +81 49 278 1834. E-mail address: [email protected] (T. Awata). 0006-291X/$ - see front matter 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2007.07.025
its pathogenesis and genetic risk factors for this condition. Most genetic studies have focused on the logical candidate genes for DR, such as aldose reductase and advanced glycation end products receptor, and several studies reported an association of these genes with DR, although further examination of these factors is required [4–6]. Previously, we have found that the C-634G SNP (rs2010963) in the 5 0 -untranslated region of the vascular endothelial growth factor (VEGF) gene was significantly associated with both DR and DME [7,8]; we have also provided evidence supporting the functional significance of the SNP [7–9]. Subsequently, rs2010963 was also found to be associated with giant cell arteritis [10], psoriasis [11], Behcet’s disease [12], proliferative retinopathy of prematurity [13], age-related macular degeneration [14] and PDR [15]. In the present study, we evaluated pigment epitheliumderived factor (PEDF or SERPINF1) as a potential
422
H. Iizuka et al. / Biochemical and Biophysical Research Communications 361 (2007) 421–426
susceptibility gene for DR. PEDF is widely expressed in the nervous system including the retina and in most tissues of the body, and it is an important neuromodulatory factor that can provide effective neuroprotection and a primary endogenous angiogenic inhibitor in the eye [16–20]. It is suggested that PEDF, as a counterpart of VEGF, is an antiinflammatory factor that may be actively involved in the pathogenesis of DR [3,21]. A decrease of PEDF levels was actually observed in the vitreous and aqueous humor from patients with DR [22,23]. Furthermore, PEDF knockout mice showed several retinal abnormalities, such as morphological alterations and increased microvessel density [24]. Therefore, PEDF is an attractive candidate gene for DR. Materials and methods Subjects. The present study included 416 type 2 diabetes patients. All of the subjects were unrelated and Japanese. The patients were recruited from the patients admitted to the Saitama Medical University Hospital in Saitama prefecture, Japan (age range: 15–83 yr). Written informed consent was obtained from each individual, and the study was approved by the Ethical Committee of Saitama Medical University and all procedures were conducted in accordance with the principles of the Declaration of Helsinki. Diagnosis, classification of diabetes and measurement of clinical features were carried out as previously reported [25]. Ophthalmologic evaluation. The funduscopic characteristics of patients were determined and graded by trained retinal specialists (S.Y. and K.M.), based on a grading scale similar to the international clinical diabetic retinopathy/macular edema disease severity scale (http://www.aao.org/aao/ education/library/international_DR.cfm), as reported previously [25]. Based on the severity grading, fluorescein angiography was performed on those patients with severe non-proliferative (NPDR) or PDR to confirm the funduscopic observations. Genotyping of the PEDF gene polymorphisms. The human PEDF gene consists of eight exons [19]. Eighteen patients were screened for variations in the promoter region and all exons and their boundary regions of the Table 1 Clinical characteristics of type 2 diabetic patients
n (male:female) Age (yr) Age at diagnosis (yr) Duration of disease (yr) Systolic BP (mmHg) Diastolic BP (mmHg) HbA1c (%) Cholesterol (mg/dl) Triglyceride (mg/dl) Insulin therapy
DR ()
DR (+)
p value
229 (112:117) 55 ± 15 48 ± 14 7±7 134 ± 20 78 ± 15 8.8 ± 2.3 196 ± 42 153 ± 118 24%
187 (96:91) 58 ± 11 46 ± 12 13 ± 8 138 ± 22 77 ± 14 8.9 ± 2.0 197 ± 51 148 ± 90 44%
NS 0.0066 NS <0.0001 0.0231 NS NS NS NS <0.0001
Data are n, means ± SD or %. BP, blood pressure; DR, diabetic retinopathy; NS, not significant.
PEDF gene by directly sequencing DNA fragments produced by PCR. This yielded the identification of 13 polymorphisms in the PEDF genes. All of these polymorphisms were included in the NCBI dbSNP database (http://www.ncbi.nlm.nih.gov/projects/SNP/index.html). Among them, we further analyzed two SNPs in the promoter region (rs12150053 and rs12948385), one in the 5 0 -untranslated region (rs9913583), and a nonsynonymous SNP in exon 3 (rs1136287; Glu298Asp) because they might affect the gene expression or the protein function; seven SNPs in introns and two synonymous SNP (Thr130Thr and Tyr321Tyr) were excluded in the present study. The genotyping of each polymorphism was carried out using polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) analysis. The PCR primers for rs12150053, rs12948385, rs9913583 and rs1136287 were 5 0 -TTGAGACAAGCGTGA CCAATGT-3 0 (forward) and 5 0 -TTCCCGTCACCATAACC-3 0 (reverse); 5 0 -CTACAGGTGCGCGCCAAC-3 0 (forward) and 5 0 -CTAAGGCGCA CACATTTGCACA-3 0 (reverse); 5 0 -ATCACAGAATCCAAATTGAGT GCAG-3 0 (forward) and 5 0 -TCTTTTCTGCTCCCTGGAGTGCCCC-3 0 (reverse); and 5 0 -GACAGTCCCTGTGCATCTCTGCGAG-3 0 (forward) and 5 0 -TCAGCCACGTTTACGCAGAGGCCAC-3 0 (reverse), respectively. The rs12150053 C allele, the rs12948385 A allele, the rs9913583 A allele, and the rs1136287 C allele result in the gain of a BsrI site, a HincII site, a BglII site, and a BssSI site, respectively. After digestion by an appropriate restriction enzyme described above, the PCR products were separated by electrophoreses on a 2.5–3.0% agarose gel and then visualized using ethidium bromide staining. Genotyping of the VEGF gene polymorphism and interaction between VEGF and PEDF SNPs. The VEGF C-634G SNP (rs2010963) was genotyped as described previously [8]. The interaction of the VEGF rs2010963 and the PEDF rs12150053 was evaluated using a logistic regression analysis; i.e. the VEGF rs2010963 (CC = 1, CG or GG = 0; recessive model), the PEDF rs12150053 (TC or CC = 1, TT = 0; dominant model) and the product of them (CC/TC or CC/CC = 1, others = 0), as an interactive term, were included together as explanatory variables in a multivariate logistic regression model, and then the interactive effects were considered to be represented by the increased OR for the product. Statistical analysis. Continuous clinical data were compared using the unpaired Student’s t-test, and categorical clinical data were compared using Fisher’s exact test. Multivariate logistic regression analyses were performed to assess the independent role of the genotype of SNPs and other variables, including sex, age at onset, duration of diabetes, systolic blood pressure, HbA1c level, cholesterol level, and treatment of diabetes (categorical variable: insulin therapy or no insulin therapy). StatView Ver.5 (SAS Institute; Cary, NC) was used for these tests. Hardy–Weinberg equilibrium, linkage disequilibrium, case-control study, and case-control haplotype analysis for multiple loci were assessed using the SNPAlyze Ver.6 software (Dynacom, Yokohama, Japan). Statistical significance was defined as p < 0.05.
Results Clinical characteristics Among the 416 type 2 diabetic patients, 229 patients had no evidence of retinopathy and the remaining 187 had DR.
Table 2 Pairwise linkage disequilibrium between PEDF SNPs SNP
rs12150053 rs12948385 rs9913583 rs1136287
Major/minor allele
T/C G/A C/A C/T
Position from transcription start site
Position from translation start site
790 358 +72 +8028
5740 5308 4879 +3078
r2 (upper)/D 0 (lower) rs12150053
rs12948385 0.919
0.974 1 0.962
1 0.907
rs9913583
rs1136287
0.035 0.037
0.215 0.198 0.163
0.934
0.0067 0.0081
0.0027
NS
423
0.0045 0.0048
NS
Table 1 summarizes the clinical characteristics of the diabetic patients included in the study. Among 187 patients with DR, 101 patients had non-proliferative retinopathy, and 86 had PDR; DME was present in 66 patients.
0.0015
p value
H. Iizuka et al. / Biochemical and Biophysical Research Communications 361 (2007) 421–426
NS NS — 1.33 (0.84–2.12) 0.0229 0.0106
NS
1.73 (1.14–2.60) 2.50 (0.62–10.12) — 1.61 (1.13–2.31) Codominant Allele
Recessive
112:69:6 G/A rs12948385
165:61:3
Dominant
Codominant Allele
Recessive
Dominant 167:60:2
NS
1.53 (0.90–2.58) –– 54:32:0 0.0120
0.0110 0.0057
NS
0.0085
1.76 (1.17–2.67) 3.76 (0.75–18.86) — 1.68 (1.17–2.41) 113:68:6
OR (95% CI)
DR, diabetic retinopathy; PDR, proliferative diabetic retinopathy; DME, macular edema; OR, odds ratio; NS, not significant. p values are vs DR ().
2.43 (1.38–4.26) 1.16 (0.12–11.33) — 1.95 (1.21–3.12) NS NS
NS NS
55:31:0
— 1.35 (0.85–2.17)
NS
1.52 (0.90–2.57) —
NS
34:31:1
34:31:1
OR (95% CI) OR (95% CI) Genotypes (11:12:22) Genotypes (11:12:22)
p value
PDR DR (+)
Case
Model in carriers of minor allele Control
T/C
Previous studies demonstrated that the C-634G SNP (rs2010963) of the VEGF gene was associated with DR [7,8]. In the present study, a possible interaction between
rs12150053
Interaction between VEGF and PEDF in susceptibility to DR
DR () Genotypes (11:12:22)
To further assess the relationship of the PEDF polymorphisms with DR, multivariate logistic regression analysis was applied, including the patients’ clinical characteristics and the PEDF genotype. As shown in Table 5, the TC or CC genotype of rs12150053 was a significant risk factor for DR (dominant model, OR 2.40, p = 0.0004). The GA or AA genotype of rs12948385, which was in strong linkage disequilibrium with rs12150053, was also a significant risk factor for DR (data not shown).
Major/ minor allele
Logistic regression analysis
SNP
Inferred haplotype frequencies by EM algorithm of four SNPs were compared using permutation tests. As shown in Table 4, only the haplotype H2, which contains C at rs12150053 and A at rs12948385, increased significantly in both DR-positive patients compared with DR-negative patients, whereas haplotypes H1, H3 and H4, which contains T at rs12150053 and G at rs12948385, were decreased in those patients with DR.
Table 3 Case-control study of PEDF SNPs for diabetic retinopathy, diabetic proliferative retinopathy and diabetic macular edema
Case-control haplotype analysis of PEDF SNPs
p value
Case-control study of PEDF SNPs
Genotypes (11:12:22)
DME
All SNPs were in Hardy–Weinberg equilibrium. Pairwise D 0 and r2 were analyzed between four PEDF SNPs in patients with type 2 diabetes (Table 2). Strong linkage disequilibrium (r2 = 0.919) was observed between rs12150053 and rs12948385. Furthermore, weak linkage disequilibrium was observed between rs1136287 and the other three SNPs (r2 = 0.163–0.215).
Based on case-control association studies, 187 patients with DR (cases) and 229 patients without DR (controls), rs12150053 and rs12948385 were significantly associated with DR in the models of recessive, codominant and allele with C and A being a risk allele, respectively (Table 3), whereas no association with rs9913583 and rs1136289 (Met72Thr) was observed (data not shown). The association with rs12150053 or rs12948385 was stronger in the patients with DME, although this association was weaker and was not significant in those with PDR (Table 3).
2.54 (1.44–4.46) 1.75 (0.16–19.56) — 2.05 (1.23–3.30)
Linkage disequilibrium between SNPs
424
H. Iizuka et al. / Biochemical and Biophysical Research Communications 361 (2007) 421–426
Table 4 Case-control haplotype analysis of PEDF SNPs for diabetic retinopathy Haplotype
SNP
DR
rs12150053
rs12948385
rs9913583
rs1136287
()
(+)
p value
H1 H2 H3 H4
T C T T
G A G G
C C C A
C T T T
0.550 0.139 0.171 0.141
0.492 0.208 0.156 0.144
NS 0.0052 NS NS
DR, diabetic retinopathy; NS, not significant. p values are based on 10,000 permutations.
VEGF rs2010963 and PEDF rs12150053 in susceptibility to DR was evaluated. With a multivariate logistic regression analysis, a significant interaction between the VEGF and PEDF SNPs was observed (p = 0.0158, Table 6). The OR of the combined genotypes (VEGF rs2010963/PEDF rs12150053) is 7.06, that is greater than the product of the ORs of the individual genotypes (1.94 · 2.40 = 4.66), indicating the existence of a synergic effect of the VEGF and PEDF polymorphisms toward DR susceptibility. Discussion The present study evaluated a potential association of PEDF polymorphisms with DR in Japanese patients with type 2 diabetes. The results of the case-control studies demonstrated that two SNPs (rs12150053 and rs12948385) in the promoter region of PEDF were significantly associated with the presence of DR. Using logistic regression analysis, including several clinical characteristics such as disease duration, the genotypes of the SNPs were also significantly associated with an increased risk of DR. In addition, Table 5 Multivariate logistic regression analysis for the association with diabetic retinopathy Variable
OR (95% CI)
p value
Female sex Age at onset Duration of disease Systolic blood pressure HbA1c Cholesterol Insulin therapy PEDF rs12150053a
0.82 1.01 1.11 1.01 1.01 1.00 2.68 2.40
NS NS 2.56 · 109 0.0359 NS NS 0.0002 0.0004
(0.52–1.30) (0.99–1.03) (1.07–1.15) (1.00–1.03) (0.91–1.14) (1.00–1.01) (1.60–4.48) (1.48–3.91)
a Dominant model (TC or CC vs TT). OR, odds ratio; NS, not significant.
although we presented uncorrected p values, even with a conservative Bonferroni correction for multiple testing, the associations between the PEDF SNPs and DR remained significant. Previously, PEDF was analyzed as a candidate gene for two ocular diseases, Leber’s congenital amaurosis [26] and retinitis pigmentosa [27], but PEDF was excludes as a common cause of either of the disease. Yamagishi et al. analyzed the PEDF Met72Thr polymorphism (rs1136289) in 143 diabetic patients (10 type 1 and 133 type 2), but there were no significant correlations between PEDF Met72Thr and diabetic microangiopathy including DR [28]. Although the sample size of the study was small, the data was consistent with the present study, which showed no significant association between the Met72Thr SNP and DR. This is the first report that demonstrates a significant relation between PEDF polymorphisms and disease susceptibility. A significant association with DR was observed in rs12150053 and rs12948385 SNPs, which were in tight linkage disequilibrium each other. Case-control haplotype analysis suggested the primary role of these SNPs in disease susceptibility. Since these SNPs are located in the promoter region of PEDF and other variations were not found in the nearby region after screening of 18 patients, it is tempting to speculate that decreased PEDF expression in the retina induced by these SNPs, both or alone, contributes to the development of DR. The PEDF promoter region contains several putative binding sites for transcription elements and two Alu repetitive sequences [29]. Neither of the SNPs was not included in the putative transcription factor binding sites. We are now conducting experiments on the influence of the SNPs on the regulation of PEDF transcription and expression of the protein. Similarly to the observations in the VEGF C-634G [7], the PEDF SNPs were not specifically associated with the development of PDR. Although PEDF is a strong anti-
Table 6 Interaction between VEGF rs2010963 and PEDF r12150053 in the association with diabetic retinopathy SNP
DR () (n = 187)
DR (+) (n = 229)
OR (95% CI)a
p value
VEGF rs2010963 (CC vs CG or GG) PEDF rs12150053 (TC or CC vs TT) VEGF rs2010963 + PEDF rs12150053 (CC/TC or CC/CC vs others)b
15.7% (36) 27.1% (62) 3.9% (9)
21.9% (41) 39.6% (74) 13.9% (26)
1.94 (1.09–3.46) 2.40 (1.48–3.91) 7.06 (2.70–18.50)
0.0239 0.0004 6.95 · 105
a Odds ratios were adjusted by multivariate logistic regression analysis including sex, age at onset, duration, systolic blood pressure, HbA1c, cholesterol, and insulin therapy. b Interaction between VEGF and PEDF SNPs is significant by multivariate logistic regression analysis (p = 0.0158).
H. Iizuka et al. / Biochemical and Biophysical Research Communications 361 (2007) 421–426
angiogenic factor in the eye, which could contribute to the development of PDR, its neuroprotective effect may explain the association of the PEDF SNPs and the development of DR [16,17,19,30]. Furthermore, PEDF was recently shown to reduce the VEGF-induced vascular leakage, probably through the anti-inflammatory effect of PEDF [21,31]. Therefore, PEDF may also contribute to the development of DME. In fact, the present study showed that the association of the PEDF SNPs was stronger in DR patients complicated by DME. However, the number of patients with PDR or DME was not large; especially the number of PDR patients might have been insufficient in this study to detect a significant association. Two genes, VEGF and PEDF, are now found to be associated with DR. The present study assessed a possible interactive effect of VEGF rs2010963 and PEDF rs12150053 using logistic regression analysis. The patients with the combined risk genotypes of rs2010963 and rs12150053 have an increased risk of developing DR. The risk is significantly interactive, thus indicating a synergistic effect of the VEGF and PEDF SNPs. Accumulating evidence suggests that VEGF and PEDF have a reciprocal relation in the eye, and their imbalance, VEGF increase and PEDF decrease, may play an important role in the development of DR [3,21]. It is therefore possible that polymorphisms in VEGF and PEDF SNPs, leading to an imbalance of VEGF and PEDF expression, may jointly contribute to the development of DR. However, further genetic studies with larger sample size and functional studies may be necessary to clarify the exact role of the PEDF SNPs, and their possible interaction with VEGF SNPs in the development of diabetic retinal complications. Finally, based on the neuroprotective and anti-angiogenic functions, PEDF may play a role in the pathogenesis of neurodegenerative diseases and angiogenic disorders, such as Alzheimer’s disease, Parkinson’s disease, neovascular age-related macular degeneration and tumors. Furthermore, serum PEDF levels have recently been reported to be elevated in patients with metabolic syndrome [32]. Therefore, it may be worthwhile to analyze the possible association between PEDF polymorphisms and these diseases. Acknowledgments We thank Ritsuko Doki for her valuable technical assistance. This work was supported in part by grants from the Japanese Diabetes Foundation, the Medical Research Center, Saitama Medical University, and the Japanese Ministry of Education, Culture, Sports, Science and Technology. References [1] M. Porta, F. Bandello, Diabetic retinopathyA clinical update, Diabetologia 45 (2002) 1617–1634. [2] T.A. Ciulla, A.G. Amador, B. Zinman, Diabetic retinopathy and diabetic macular edema: pathophysiology, screening, and novel therapies, Diabetes Care 26 (2003) 2653–2664. [3] R.N. Frank, Diabetic retinopathy, N. Engl. J. Med. 350 (2004) 48–58.
425
[4] K.M. Warpeha, U. Chakravarthy, Molecular genetics of microvascular disease in diabetic retinopathy, Eye 17 (2003) 305–311. [5] C.L. Hanis, D. Hallman, Genetics of diabetic retinopathy, Curr. Diab. Rep. 6 (2006) 155–161. [6] K. Uhlmann, P. Kovacs, Y. Boettcher, H.P. Hammes, R. Paschke, Genetics of diabetic retinopathy, Exp. Clin. Endocrinol. Diabetes 114 (2006) 275–294. [7] T. Awata, K. Inoue, S. Kurihara, T. Ohkubo, M. Watanabe, K. Inukai, I. Inoue, S. Katayama, A common polymorphism in the 5 0 untranslated region of the VEGF gene is associated with diabetic retinopathy in type 2 diabetes, Diabetes 51 (2002) 1635–1639. [8] T. Awata, S. Kurihara, N. Takata, T. Neda, H. Iizuka, T. Ohkubo, M. Osaki, M. Watanabe, Y. Nakashima, K. Inukai, I. Inoue, I. Kawasaki, K. Mori, S. Yoneya, S. Katayama, Functional VEGF C634G polymorphism is associated with development of diabetic macular edema and correlated with macular retinal thickness in type 2 diabetes, Biochem. Biophys. Res. Commun. 333 (2005) 679–685. [9] D. Lambrechts, E. Storkebaum, M. Morimoto, J. Del-Favero, F. Desmet, S.L. Marklund, S. Wyns, V. Thijs, J. Andersson, I. van Marion, A. Al-Chalabi, S. Bornes, R. Musson, V. Hansen, L. Beckman, R. Adolfsson, H.S. Pall, H. Prats, S. Vermeire, P. Rutgeerts, S. Katayama, T. Awata, N. Leigh, L. Lang-Lazdunski, M. Dewerchin, C. Shaw, L. Moons, R. Vlietinck, K.E. Morrison, W. Robberecht, C. Van Broeckhoven, D. Collen, P.M. Andersen, P. Carmeliet, VEGF is a modifier of amyotrophic lateral sclerosis in mice and humans and protects motoneurons against ischemic death, Nat. Genet. 34 (2003) 383–394. [10] L. Boiardi, B. Casali, D. Nicoli, E. Farnetti, Q. Chen, P. Macchioni, M.G. Catanoso, L. Pulsatelli, R. Meliconi, C. Salvarani, Vascular endothelial growth factor gene polymorphisms in giant cell arteritis, J. Rheumatol. 30 (2003) 2160–2164. [11] H.S. Young, A.M. Summers, M. Bhushan, P.E. Brenchley, C.E. Griffiths, Single-nucleotide polymorphisms of vascular endothelial growth factor in psoriasis of early onset, J. Invest. Dermatol. 122 (2004) 209–215. [12] C. Salvarani, L. Boiardi, B. Casali, I. Olivieri, F. Cantini, F. Salvi, R. Malatesta, R. La Corte, G. Triolo, A. Ferrante, D. Filippini, G. Paolazzi, P. Sarzi-Puttini, D. Nicoli, E. Farnetti, Q. Chen, L. Pulsatelli, Vascular endothelial growth factor gene polymorphisms in Behcet’s disease, J. Rheumatol. 31 (2004) 1785–1789. [13] A. Vannay, G. Dunai, I. Banyasz, M. Szabo, R. Vamos, A. Treszl, J. Hajdu, T. Tulassay, B. Vasarhelyi, Association of genetic polymorphisms of vascular endothelial growth factor and risk for proliferative retinopathy of prematurity, Pediatr. Res. 57 (2005) 396–398. [14] J.L. Haines, N. Schnetz-Boutaud, S. Schmidt, W.K. Scott, A. Agarwal, E.A. Postel, L. Olson, S.J. Kenealy, M. Hauser, J.R. Gilbert, M.A. Pericak-Vance, Functional candidate genes in agerelated macular degeneration: significant association with VEGF, VLDLR, and LRP6, Invest. Ophthalmol. Vis. Sci. 47 (2006) 329–335. [15] F.I. Errera, L.H. Canani, M.E. Silva, E. Yeh, W. Takahashi, K.G. Santos, K.E. Souto, B. Tschiedel, I. Roisenberg, J.L. Gross, M.R. Passos-Bueno, Functional vascular endothelial growth factor 634G>C SNP is associated with proliferative diabetic retinopathy: a case-control study in a Brazilian population of European ancestry, Diabetes Care 30 (2007) 275–279. [16] P.G. Gettins, M. Simonovic, K. Volz, Pigment epithelium-derived factor (PEDF), a serpin with potent anti-angiogenic and neurite outgrowth-promoting properties, Biol. Chem. 383 (2002) 1677–1682. [17] J. Tombran-Tink, C.J. Barnstable, PEDF: a multifaceted neurotrophic factor, Nat. Rev. Neurosci. 4 (2003) 628–636. [18] C.J. Barnstable, J. Tombran-Tink, Neuroprotective and antiangiogenic actions of PEDF in the eye: molecular targets and therapeutic potential, Prog. Retin. Eye Res. 23 (2004) 561–577. [19] J. Tombran-Tink, The neuroprotective and angiogenesis inhibitory serpin, PEDF: new insights into phylogeny, function, and signaling, Front Biosci. 10 (2005) 2131–2149. [20] G.L. King, K. Suzuma, Pigment-epithelium-derived factor—a key coordinator of retinal neuronal and vascular functions, N. Engl. J. Med. 342 (2000) 349–351.
426
H. Iizuka et al. / Biochemical and Biophysical Research Communications 361 (2007) 421–426
[21] S.X. Zhang, J.J. Wang, G. Gao, C. Shao, R. Mott, J.X. Ma, Pigment epithelium-derived factor (PEDF) is an endogenous antiinflammatory factor, FASEB J. 20 (2006) 323–325. [22] N. Ogata, J. Tombran-Tink, M. Nishikawa, T. Nishimura, Y. Mitsuma, T. Sakamoto, M. Matsumura, Pigment epithelium-derived factor in the vitreous is low in diabetic retinopathy and high in rhegmatogenous retinal detachment, Am. J. Ophthalmol. 132 (2001) 378–382. [23] B.O. Boehm, G. Lang, O. Volpert, P.M. Jehle, A. Kurkhaus, S. Rosinger, G.K. Lang, N. Bouck, Low content of the natural ocular anti-angiogenic agent pigment epithelium-derived factor (PEDF) in aqueous humor predicts progression of diabetic retinopathy, Diabetologia 46 (2003) 394–400. [24] J.A. Doll, V.M. Stellmach, N.P. Bouck, A.R. Bergh, C. Lee, L.P. Abramson, M.L. Cornwell, M.R. Pins, J. Borensztajn, S.E. Crawford, Pigment epithelium-derived factor regulates the vasculature and mass of the prostate and pancreas, Nat. Med. 9 (2003) 774–780. [25] T. Awata, T. Neda, H. Iizuka, S. Kurihara, T. Ohkubo, N. Takata, M. Osaki, M. Watanabe, Y. Nakashima, T. Sawa, K. Inukai, I. Inoue, M. Shibuya, K. Mori, S. Yoneya, S. Katayama, Endothelial nitric oxide synthase gene is associated with diabetic macular edema in type 2 diabetes, Diabetes Care 27 (2004) 2184– 2190. [26] R. Koenekoop, A.L. Pina, M. Loyer, J. Davidson, J. Robitaille, I. Maumenee, J. Tombran-Tink, Four polymorphic variations in the PEDF gene identified during the mutation screening of patients with Leber congenital amaurosis, Mol. Vis. 5 (1999) 10.
[27] R. Goliath, J. Tombran-Tink, I.R. Rodriquez, G. Chader, R. Ramesar, J. Greenberg, The gene for PEDF, a retinal growth factor is a prime candidate for retinitis pigmentosa and is tightly linked to the RP13 locus on chromosome 17p13.3, Mol. Vis. 2 (1996) 5. [28] S. Yamagishi, S. Amano, Y. Inagaki, T. Okamoto, Y. Koda, M. Soejima, H. Kimura, Pigment epithelium-derived factor Met72Thr polymorphism in patients with diabetic microangiopathy, Int. J. Clin. Pharmacol. Res. 22 (2002) 67–71. [29] J. Tombran-Tink, K. Mazuruk, I.R. Rodriguez, D. Chung, T. Linker, E. Englander, G.J. Chader, Organization, evolutionary conservation, expression and unusual Alu density of the human gene for pigment epithelium-derived factor, a unique neurotrophic serpin, Mol. Vis. 2 (1996) 11. [30] U. Barnas, A. Schmidt, A. Illievich, H.P. Kiener, D. Rabensteiner, A. Kaider, R. Prager, H. Abrahamian, K. Irsigler, G. Mayer, Evaluation of risk factors for the development of nephropathy in patients with IDDM: insertion/deletion angiotensin converting enzyme gene polymorphism, hypertension and metabolic control, Diabetologia 40 (1997) 327–331. [31] H. Liu, J.G. Ren, W.L. Cooper, C.E. Hawkins, M.R. Cowan, P.Y. Tong, Identification of the antivasopermeability effect of pigment epithelium-derived factor and its active site, Proc. Natl. Acad. Sci. USA 101 (2004) 6605–6610. [32] S. Yamagishi, H. Adachi, A. Abe, T. Yashiro, M. Enomoto, K. Furuki, A. Hino, Y. Jinnouchi, K. Takenaka, T. Matsui, K. Nakamura, T. Imaizumi, Elevated serum levels of pigment epithelium-derived factor in the metabolic syndrome, J. Clin. Endocrinol. Metab. 91 (2006) 2447–2450.