Abstract

Pseudoxanthoma elasticum (PXE) is a heritable disorder affecting the skin, eyes and cardiovascular system. It is caused by mutations in the ABCC6 gene and its clinical picture is highly variable. PXE often leads to severe visual impairment due to the development of choroidal neovascularisation (CNV). CNV in PXE-associated retinopathy is believed to be mediated by the action of vascular endothelial growth factor (VEGF). The objective of the present study was to evaluate a possible impact of variations in the VEGFA gene on ocular manifestations of PXE. For this purpose, we evaluated the distribution of 10 single nucleotide polymorphisms (SNPs) in the promoter and coding region of the VEGFA gene in DNA samples from 163 German patients affected by PXE and in 163 healthy control subjects. Haplotype analysis of SNPs c.-1540A>C, c.-460C>T, c.-152G>A, c.405C>G, c.674C>T, c.1032C>T, c.4618C>T and c.5092C>A revealed that the haplotype CTGGCCCC was associated with PXE (OR 2.05, 95% CI 1.33–3.15, Pcorrected = 0.01). Furthermore, five SNPs showed significant association with severe retinopathy. The most significant single SNP association was c.-460C>T (OR 3.83, 95% CI 2.01–7.31, Pcorrected = 0.0003). Logistic regression analysis identified the c.-460T and the c.674C alleles as independent risk factors for development of severe retinopathy. Our findings suggest an involvement of VEGF in the pathogenesis of ocular PXE manifestations. VEGF gene polymorphisms might prove useful as prognostic markers for the development of PXE-associated retinopathy and permit earlier therapeutic intervention in order to prevent loss of central vision, one of the most devastating consequences of this disease.

INTRODUCTION

Pseudoxanthoma elasticum (PXE; Groenblad–Strandberg syndrome, OMIM 177850 and 264800) is an autosomally heritable disease with considerable phenotypic heterogeneity ( 1 , 2 ). PXE is characterized by ectopic mineralization and fragmentation of elastic fibers in the skin, the vessel walls and the Bruch membrane in the eye. It is caused by mutations in the ABCC6 gene ( 3 ). The protein encoded by this gene is a member of the superfamily of ATP-binding cassette transporters and belongs to the MRP subfamily, which is involved in multi-drug resistance. The precise physiological function of ABCC6 and its physiological substrate are as yet unknown, although it has been suggested that it might serve as an export pump facilitating the removal of certain metabolites from hepatocytes. Until today, as many as 188 different PXE-associated mutations in ABCC6 have been described ( 4 ). PXE is believed to be a result of impaired ABCC6 transporter activity due to mutations leading to limited protein functionality or complete lack of gene product as a result of nonsense-mediated decay. Epigenetic mechanisms have also been implied in PXE etiology ( 5 ). ABCC6 is mainly expressed in the liver and kidneys, while little or no protein is found in the tissues affected by PXE. On the basis of this finding, it was concluded that PXE is at least partly a metabolic disease ( 6 ).

The clinical picture of PXE is highly variable, with age at disease onset and the number and magnitude of its symptoms differing considerably between patients ( 7 ). This is even true in the case of affected siblings bearing the same ABCC6 genotype ( 8 ). On the basis of these observations, it has been speculated that other genes, called modifier genes, as well as environmental factors might contribute to the expression and severity of PXE. The relevance of this approach could be confirmed by the identification of modifier genes for PXE in recent years ( 9–11 ). Those genes are involved in established or assumed pathomechanisms of PXE, namely the biosynthesis of glycosaminoglycans, regulation of biological calcification and response to oxidative stress.

Virtually all patients affected by PXE develop ocular manifestations, which in many cases are the severest consequences of the disease ( 12 ). Common findings are peau d'orange, angioid streaks and choroidal neovascularisation (CNV). In the course of the disease, many patients suffer from severe visual impairment due to the development of CNV complicating angioid streaks. CNV in PXE-associated retinopathy is believed to be mediated by the action of vascular endothelial growth factor (VEGF), thus resembling the neovascular form of age-related macular degeneration (AMD) and proliferative diabetic retinopathy (DR).

AMD is the leading cause of new-onset blindness in the western world ( 13 ). It is responsible for a loss of central vision in ∼28% of individuals over 75 years of age ( 14 , 15 ). The neovascular (‘wet’) form of AMD is characterized by subretinal neovascularisation with subsequent haemorrhages or leakage which may result in a sudden loss of vision ( 16 ). These symptoms show a high degree of similarity to the ocular changes observed in PXE ( 17 ). Both diseases lead to modifications of subretinal structures resulting in the development of CNV. AMD has a complex etiology that involves multiple genes, as well as risk factors like smoking and hypertension ( 16 , 18 ). Association of several single markers and haplotypes in the VEGFA gene has been reported for AMD.

VEGF is a highly conserved homodimeric glycoprotein and a potent mediator of angiogenesis ( 19 ). In addition, VEGF is able to increase microvascular permeability, dilate arteries and chemotactically attract monocytes ( 20 ). The VEGFA gene can be alternatively spliced into different isoforms ( 19 ). VEGF has been implicated in a number of diseases, especially those with an angiogenic basis. It is elevated in patients with DR, as well as in those with AMD ( 21 , 22 ). The VEGFA gene is highly polymorphic ( 23 , 24 ). Some variations in VEGFA have been shown to influence VEGF expression ( 20 , 23 , 24 ). Recently, association of polymorphisms in the VEGFA gene with DR and AMD has been shown ( 21 , 25–28 ). The parallels between those retinopathies and the ocular pathology seen in PXE lead us to the hypothesis that the VEGFA gene may also be associated with retinopathy in PXE. Therefore, the objective of the present study was to evaluate a possible association of variations in the VEGFA gene with PXE. For this purpose, we determined the allele frequencies of 10 polymorphisms in the promoter and coding region of the VEGFA gene in a cohort of patients suffering from PXE and a control cohort comprising healthy individuals. We further investigated a possible impact of those variations on the severity of the ocular phenotype seen in PXE patients.

RESULTS

Association of single markers with PXE

Ten single nucleotide polymorphisms (SNPs) in the promoter and coding region of the VEGFA gene were genotyped in DNA samples from 163 patients affected by PXE and in 163 healthy controls. The allelic distribution for the polymorphisms is shown in Table  1 . No single marker was significantly associated with either the PXE group or the control group. All genotype distributions were within the Hardy–Weinberg equilibrium.

Table 1.

Allele frequencies of VEGFA polymorphisms in PXE patients and controls

Polymorphism RS number Allele  Cases a ( n = 326)   Controls a ( n = 326)  OR (95% CI) P -value  
c.-1540A>C rs699947 A 158 (48.5) 164 (50.3) 1.08 (0.79–1.46) 0.638 
C 168 (51.5) 162 (49.7) 
c.-460C>T rs833061 C 160 (49.1) 172 (52.8) 1.16 (0.85–1.58) 0.347 
T 166 (50.9) 154 (47.2) 
c.-152G>A rs13207351 G 164 (50.3) 156 (47.9) 1.10 (0.81–1.50) 0.531 
A 162 (49.7) 170 (52.1) 
c.405C>G rs2010963 C 91 (27.9) 109 (33.4) 1.30 (0.93–1.81) 0.126 
G 235 (72.1) 217 (66.6) 
c.674C>T rs1413711 C 168 (51.5) 161 (49.4) 1.09 (0.80–1.48) 0.584 
T 158 (48.5) 165 (50.6) 
c.1032C>T rs25648 C 261 (80.1) 273 (83.7) 1.28 (0.86–1.92) 0.222 
T 65 (19.9) 53 (16.3) 
c.4618C>T rs735286 C 238 (73.0) 222 (68.1) 1.26 (0.89–1.76) 0.169 
T 88 (27.0) 104 (31.9) 
c.5092C>A rs2146323 C 223 (68.4) 203 (62.3) 1.31 (0.95–1.81) 0.100 
A 103 (31.6) 123 (37.7) 
c.9109C>T rs3025020 C 233 (71.5) 215 (66.0) 1.27 (0.91–1.77) 0.128 
T 93 (28.5) 111 (34.0) 
c.9162C>T rs3025021 C 214 (65.6) 210 (64.4) 1.04 (0.75–1.43) 0.743 
T 112 (34.4) 116 (35.6) 
Polymorphism RS number Allele  Cases a ( n = 326)   Controls a ( n = 326)  OR (95% CI) P -value  
c.-1540A>C rs699947 A 158 (48.5) 164 (50.3) 1.08 (0.79–1.46) 0.638 
C 168 (51.5) 162 (49.7) 
c.-460C>T rs833061 C 160 (49.1) 172 (52.8) 1.16 (0.85–1.58) 0.347 
T 166 (50.9) 154 (47.2) 
c.-152G>A rs13207351 G 164 (50.3) 156 (47.9) 1.10 (0.81–1.50) 0.531 
A 162 (49.7) 170 (52.1) 
c.405C>G rs2010963 C 91 (27.9) 109 (33.4) 1.30 (0.93–1.81) 0.126 
G 235 (72.1) 217 (66.6) 
c.674C>T rs1413711 C 168 (51.5) 161 (49.4) 1.09 (0.80–1.48) 0.584 
T 158 (48.5) 165 (50.6) 
c.1032C>T rs25648 C 261 (80.1) 273 (83.7) 1.28 (0.86–1.92) 0.222 
T 65 (19.9) 53 (16.3) 
c.4618C>T rs735286 C 238 (73.0) 222 (68.1) 1.26 (0.89–1.76) 0.169 
T 88 (27.0) 104 (31.9) 
c.5092C>A rs2146323 C 223 (68.4) 203 (62.3) 1.31 (0.95–1.81) 0.100 
A 103 (31.6) 123 (37.7) 
c.9109C>T rs3025020 C 233 (71.5) 215 (66.0) 1.27 (0.91–1.77) 0.128 
T 93 (28.5) 111 (34.0) 
c.9162C>T rs3025021 C 214 (65.6) 210 (64.4) 1.04 (0.75–1.43) 0.743 
T 112 (34.4) 116 (35.6) 

a Alleles n (%).

Association of single markers with severe retinopathy

Results of ophthalmological examination were available for 97 of 163 PXE patients in this study. These patients were separated into subgroups depending on the severity of ocular manifestations as described in Materials and Methods. Analysis showed a significant association ( P < 0.05) with severe ocular involvement for eight of ten markers (Table  2 ). After correction for multiple testing, c.-1540A>C (rs699947), c.-460C>T (rs833061), c.-152G>A (rs13207351), c.674C>T (rs1413711) and c.1032C>T (rs25648) remained significantly associated with severe retinopathy.

Table 2.

Allele frequencies of VEGFA polymorphisms in patient subgroups

Polymorphism RS number Allele  Group A a ( n = 62)   Group B a ( n = 132)  OR (95% CI) Pcorrected 
c.-1540A>C rs699947 A 40 (64.5) 81 (61.4) 2.89 (1.54–5.41) 0.011 
C 22 (35.5) 51 (38.6) 
c.-460C>T rs833061 C 43 (69.4) 49 (37.1) 3.83 (2.01–7.31) 0.0003 
T 19 (30.6) 83 (62.9) 
c.-152G>A rs13207351 G 19 (30.6) 82 (62.1) 3.71 (1.95–7.07) 0.0004 
A 43 (69.4) 50 (37.9) 
c.405C>G rs2010963 C 11 (17.7) 46 (34.8) 2.48 (1.18–5.22) 0.147 
G 51 (82.3) 86 (65.2) 
c.674C>T rs1413711 C 21 (33.9) 82 (62.1) 3.21 (1.70–6.02) 0.004 
T 41 (66.1) 50 (37.9) 
c.1032C>T rs25648 C 42 (67.7) 114 (86.4) 3.02 (1.46/6.25) 0.023 
T 20 (32.3) 18 (13.6) 
c.4618C>T rs735286 C 51 (82.3) 88 (66.7) 2.32 (1.10–4.89) 0.246 
T 11 (17.7) 44 (33.3) 
c.5092C>A rs2146323 C 36 (58.1) 99 (75.0) 2.17 (1.14–4.12) 0.168 
A 26 (41.9) 33 (25.0) 
c.9109C>T rs3025020 C 40 (64.5) 97 (73.5) 1.52 (0.80–2.91) 1.000 
T 22 (35.5) 35 (26.5) 
c.9162C>T rs3025021 C 45 (72.6) 85 (64.4) 1.42 (0.73–2.75) 1.000 
T 17 (27.4) 47 (35.6) 
Polymorphism RS number Allele  Group A a ( n = 62)   Group B a ( n = 132)  OR (95% CI) Pcorrected 
c.-1540A>C rs699947 A 40 (64.5) 81 (61.4) 2.89 (1.54–5.41) 0.011 
C 22 (35.5) 51 (38.6) 
c.-460C>T rs833061 C 43 (69.4) 49 (37.1) 3.83 (2.01–7.31) 0.0003 
T 19 (30.6) 83 (62.9) 
c.-152G>A rs13207351 G 19 (30.6) 82 (62.1) 3.71 (1.95–7.07) 0.0004 
A 43 (69.4) 50 (37.9) 
c.405C>G rs2010963 C 11 (17.7) 46 (34.8) 2.48 (1.18–5.22) 0.147 
G 51 (82.3) 86 (65.2) 
c.674C>T rs1413711 C 21 (33.9) 82 (62.1) 3.21 (1.70–6.02) 0.004 
T 41 (66.1) 50 (37.9) 
c.1032C>T rs25648 C 42 (67.7) 114 (86.4) 3.02 (1.46/6.25) 0.023 
T 20 (32.3) 18 (13.6) 
c.4618C>T rs735286 C 51 (82.3) 88 (66.7) 2.32 (1.10–4.89) 0.246 
T 11 (17.7) 44 (33.3) 
c.5092C>A rs2146323 C 36 (58.1) 99 (75.0) 2.17 (1.14–4.12) 0.168 
A 26 (41.9) 33 (25.0) 
c.9109C>T rs3025020 C 40 (64.5) 97 (73.5) 1.52 (0.80–2.91) 1.000 
T 22 (35.5) 35 (26.5) 
c.9162C>T rs3025021 C 45 (72.6) 85 (64.4) 1.42 (0.73–2.75) 1.000 
T 17 (27.4) 47 (35.6) 

a Alleles n (%).

Multivariate single marker analysis

We further assessed the relation between VEGFA polymorphisms and PXE-associated retinopathy by logistic regression analysis, including VEGFA alleles and clinical features given in Table  3 as covariates, and using forward and backward variable selection analysis (Table  4 ). Owing to high linkage disequilibrium (LD) ( r2 = 0.96) between c.-460C>T and c.-152G>A, the latter was excluded. Age, blood pressure and carriage of the c.-460T and the c.674C allele were retained, both in the final forward and in the final backward models. Analysis showed that both the c.-460T and the c.674C allele had a significantly increased risk for development of severe visual impairment (OR 6.54, 95% CI 2.67–15.9, P = 0.00004 and OR 5.05, 95% CI 2.12–12.0, P = 0.0002, respectively), indicating that both are independent risk factors for development of severe retinopathy in PXE.

Table 3.

Characteristics of ocular phenotype groups

Characteristics  Group A a  Group B b P -value  
n 31 66 NA 
Males (females) 9 (22) 21 (45) 0.819 
Age, years c 37.0 (14–70) 53.4 (20–82) <0.001 
Age at onset c 23.5 (8–53) 34.6 (7–66) 0.002 
Duration c 13.3 (2–47) 18.8 (2–51) 0.038 
Hypertension d 2 (6.5%) 27 (40.9%) 0.0003 
Current smokers d 12 (38.7%) 13 (19.7%) 0.080 
Characteristics  Group A a  Group B b P -value  
n 31 66 NA 
Males (females) 9 (22) 21 (45) 0.819 
Age, years c 37.0 (14–70) 53.4 (20–82) <0.001 
Age at onset c 23.5 (8–53) 34.6 (7–66) 0.002 
Duration c 13.3 (2–47) 18.8 (2–51) 0.038 
Hypertension d 2 (6.5%) 27 (40.9%) 0.0003 
Current smokers d 12 (38.7%) 13 (19.7%) 0.080 

NA, not applicable.

a Subjects with no or only mild retinopathy but without impairment of vision, defined as either without fundus findings or with a finding of peau d'orange and/or angioid streaks.

b Subjects with severe retinopathy accompanied by visual impairment, defined as visual acuity worse than 20/20 in at least one eye.

c Mean value (age range).

d Total number (percentage).

Table 4.

Odds ratios adjusted by logistic regression for association with the severe visual impairment group

Variable Odds ratio (95% CI) P -value  
Age 1.10 (1.06–1.14)  1.84×10 −7 
Female sex 1.31 (0.52–3.29) 0.564 
Hypertension 7.97 (2.13–29.9) 0.0021 
Current smoker 0.63 (0.26–1.54) 0.310 
c.-460T genotype 6.97 (2.81–17.3) 0.00004 
c.674C genotype 5.21 (2.18–12.5) 0.00021 
Variable Odds ratio (95% CI) P -value  
Age 1.10 (1.06–1.14)  1.84×10 −7 
Female sex 1.31 (0.52–3.29) 0.564 
Hypertension 7.97 (2.13–29.9) 0.0021 
Current smoker 0.63 (0.26–1.54) 0.310 
c.-460T genotype 6.97 (2.81–17.3) 0.00004 
c.674C genotype 5.21 (2.18–12.5) 0.00021 

Marker–marker LD

Haplotype analysis using Haploview yielded two distinct blocks of LD in the VEGFA gene, as previously described by other groups ( 18 , 27 ) (Fig.  1 ). Highest pairwise r2 (0.963) was observed between c.-460C>T (rs833061) and c-152G>A (rs13207351) in the first block.

Figure 1.

Linkage disequilibrium structure as estimated by Haploview. LD structure was analyzed for a total of 652 alleles from patients and controls. LD patterns and haplotype blocks were defined according to the ‘spine of LD’ setting in Haploview software, based on each end marker of a block having a D ′ value of more than 0.8 (33). A standard color scheme is used to display LD pattern, with black for perfect LD ( r2 = 1), white for no LD ( r2 = 0) and shades of gray for intermediate LD (0 < r2 < 1).

Figure 1.

Linkage disequilibrium structure as estimated by Haploview. LD structure was analyzed for a total of 652 alleles from patients and controls. LD patterns and haplotype blocks were defined according to the ‘spine of LD’ setting in Haploview software, based on each end marker of a block having a D ′ value of more than 0.8 (33). A standard color scheme is used to display LD pattern, with black for perfect LD ( r2 = 1), white for no LD ( r2 = 0) and shades of gray for intermediate LD (0 < r2 < 1).

Association of haplotypes with PXE

Haplotypes estimated by Haploview are shown in Table  5 . Block 1 haplotype CTGGCCCC was associated with the patient group (OR 2.05, 95% CI 1.33–3.15, Pcorrected = 0.013), whereas the haplotype ACAGTCCA was associated with the control group (OR 0.56, 95% CI 0.37–0.86, Pcorrected = 0.066), although it did not reach statistical significance after correction for multiple testing.

Table 5.

Estimated VEGFA haplotypes in PXE patients and controls

Haplotype Frequency Counts PXE, controls Frequencies PXE, controls OR (95% CI) Pcorrected 
Block 1 
 CTGCCCTC 0.289 86:238, 102:224 0.265, 0.312 0.79 (0.57–1.12) 1.000 
 ACAGCTCA 0.165 41:283, 67:259 0.126, 0.205 0.56 (1.17–2.72) 0.066 
 CTGGCCCC 0.165 69:255, 38:288 0.212, 0.118 2.05 (1.33–3.15) 0.013 
 ACAGTTCA 0.157 57:268, 46:280 0.174, 0.140 1.30 (0.85–1.98) 1.000 
 ACAGCTCC 0.146 50:274, 45:281 0.154, 0.138 1.14 (0.74–1.76) 1.000 
 CCAGCCCC 0.013 4:320, 4:322 0.014, 0.012 1.01 (0.25–4.06) 1.000 
 CTGCCCCC 0.011 4:320, 3:323 0.012, 0.010 1.35 (0.30–6.06) 1.000 
Block 2 
 CT 0.346 112:212, 113:213 0.344, 0.348 1.00 (0.72–1.38) 1.000 
 CC 0.343 122:202, 102:224 0.375, 0.312 1.33 (0.96–1.84) 0.889 
 TC 0.306 91:233, 108:218 0.279, 0.3332 0.79 (0.56–1.10) 1.000 
Haplotype Frequency Counts PXE, controls Frequencies PXE, controls OR (95% CI) Pcorrected 
Block 1 
 CTGCCCTC 0.289 86:238, 102:224 0.265, 0.312 0.79 (0.57–1.12) 1.000 
 ACAGCTCA 0.165 41:283, 67:259 0.126, 0.205 0.56 (1.17–2.72) 0.066 
 CTGGCCCC 0.165 69:255, 38:288 0.212, 0.118 2.05 (1.33–3.15) 0.013 
 ACAGTTCA 0.157 57:268, 46:280 0.174, 0.140 1.30 (0.85–1.98) 1.000 
 ACAGCTCC 0.146 50:274, 45:281 0.154, 0.138 1.14 (0.74–1.76) 1.000 
 CCAGCCCC 0.013 4:320, 4:322 0.014, 0.012 1.01 (0.25–4.06) 1.000 
 CTGCCCCC 0.011 4:320, 3:323 0.012, 0.010 1.35 (0.30–6.06) 1.000 
Block 2 
 CT 0.346 112:212, 113:213 0.344, 0.348 1.00 (0.72–1.38) 1.000 
 CC 0.343 122:202, 102:224 0.375, 0.312 1.33 (0.96–1.84) 0.889 
 TC 0.306 91:233, 108:218 0.279, 0.3332 0.79 (0.56–1.10) 1.000 

Impact of haplotypes on ocular manifestations

Haploview analysis was also performed for ocular manifestation subgroups; the results are given in Table  6 . Block 1 haplotype ACAGTTCA was markedly more frequent in the mild ocular involvement group (OR 2.78, 95% CI 1.32–5.88, Pcorrected = 0.051), even though the association was only borderline significant after correction for multiple testing.

Table 6.

Estimated haplotypes in patient subgroups

Haplotype Frequency Counts, Group A, B Frequencies, Group A, B OR (95% CI) Pcorrected 
Block 1 
 CTGCCCTC 0.284 44:88, 11:51 0.333, 0.177 2.32 (1.10–4.89) 0.246 
 CTGGCCCC 0.206 33:99, 7:55 0.250, 0.113 2.62 (1.09–6.31) 0.276 
 ACAGTTCA 0.180 17:115, 18:44 0.127, 0.293 0.36 (0.17–0.76) 0.051 
 ACAGCTCC 0.160 17:115, 14:48 0.130, 0.223 0.51 (0.23–1.11) 0.997 
 ACAGCTCA 0.103 13:119, 7:55 0.098, 0.113 0.86 (0.32–2.27) 1.000 
 CCAGCCCC 0.016 1:131, 2:60 0.006, 0.039 0.23 (0.02–2.58) 0.997 
 CTGCCCCC 0.010 2:130, 0:62 0.015, 0.000 2.40 (0.11–50.7) 1.000 
Block 2 
 CC 0.376 50:82, 23:39 0.379, 0.371 1.03 (0.55–1.93) 1.000 
 CT 0.330 47:85, 17:45 0.356, 0.274 1.46 (0.76–2.84) 1.000 
 TC 0.294 35:97, 22:40 0.265, 0.355 0.66 (0.34–1.25) 1.000 
Haplotype Frequency Counts, Group A, B Frequencies, Group A, B OR (95% CI) Pcorrected 
Block 1 
 CTGCCCTC 0.284 44:88, 11:51 0.333, 0.177 2.32 (1.10–4.89) 0.246 
 CTGGCCCC 0.206 33:99, 7:55 0.250, 0.113 2.62 (1.09–6.31) 0.276 
 ACAGTTCA 0.180 17:115, 18:44 0.127, 0.293 0.36 (0.17–0.76) 0.051 
 ACAGCTCC 0.160 17:115, 14:48 0.130, 0.223 0.51 (0.23–1.11) 0.997 
 ACAGCTCA 0.103 13:119, 7:55 0.098, 0.113 0.86 (0.32–2.27) 1.000 
 CCAGCCCC 0.016 1:131, 2:60 0.006, 0.039 0.23 (0.02–2.58) 0.997 
 CTGCCCCC 0.010 2:130, 0:62 0.015, 0.000 2.40 (0.11–50.7) 1.000 
Block 2 
 CC 0.376 50:82, 23:39 0.379, 0.371 1.03 (0.55–1.93) 1.000 
 CT 0.330 47:85, 17:45 0.356, 0.274 1.46 (0.76–2.84) 1.000 
 TC 0.294 35:97, 22:40 0.265, 0.355 0.66 (0.34–1.25) 1.000 

DISCUSSION

Although it is known that PXE is caused by mutations in the ABCC6 gene, there seems to be no obvious link between ABCC6 deficiency and the symptoms observed in patients. The mechanisms by which ABCC6 deficiency results in the activation of pathways leading to the manifestation of PXE symptoms remain to be defined. The clinical course and phenotype of PXE are highly variable ( 7 , 8 ), thus suggesting the involvement of other factors, e.g. modifier genes. Candidates for a role as a modifier in PXE are those genes that are involved in pathogenic pathways leading to the development of PXE manifestations. Any genetic variation in genes regulating these pathways may be connected to PXE pathogenesis. Indeed, in recent years, secondary genetic factors acting as modifier genes were shown to play a role in the severity and course of PXE ( 9–11 ). On the basis of the results of the current study, we propose that VEGF acts as a modifier in PXE as well. Furthermore, we present evidence for a correlation between VEGFA genotype and the ocular PXE phenotype.

Apart from the development of calcified lesions in the dermis, retinopathy is the most prominent clinical feature of PXE and also the severest consequence in most cases. Virtually all patients develop ocular manifestations in the course of the disease ( 12 ). The earliest ophthalmological findings in many cases are peau d'orange and angioid streaks, which represent ruptures of the Bruch membrane occurring due to alteration of elastic fibers. Severe visual impairment is often caused by the development of CNV that is favored by breaks in the Bruch membrane. CNV has been reported to occur in 72–86% of PXE cases ( 29 ). The prognosis for visual acuity is crucially dependent on whether CNV emerges or not. Even though breaks in the Bruch membrane favor the growth of fibrovascular tissue originating from the choroid through the break, defects in the Bruch membrane are not automatically associated with CNV ( 30 ). An additional requirement is the action of growth factors that drive the formation of new vessels, and among them VEGF is the most important. VEGF has been implied in a variety of pathologies with an angiogenic basis, among them proliferative DR and neovascular AMD. The phenotypic similarities these conditions share with PXE-associated retinopathy and the crucial role of VEGF in CNV development suggest a potential role for VEGF in PXE susceptibility. Supporting evidence comes from the successful application of the anti-VEGF antibody bevacizumab in the treatment of PXE-associated retinopathy ( 12 ). In the present study, we analyzed polymorphisms in the promoter and coding region of the VEGF gene. Haplotype analysis revealed one haplotype to be significantly more frequent among PXE patients. This haplotype contains the c.-460T and the c.-152G alleles, both of which were present in the promoter haplotype Churchill et al . ( 26 ) identified to be associated with AMD.

We further studied the impact of VEGFA genotype on the course of PXE-associated retinopathy by dividing the study cohort according to severity of visual impairment. Analysis clearly shows an association of several VEGFA polymorphisms and the extent of ocular manifestation. As older patients tend to have more pronounced ocular manifestations than younger ones, we performed logistic regression analysis in order to correct for age and other potential confounding variables. Using forward and backward variable selection analyses, we identified carriage of the c.-460T and the c.674C allele as independent risk factors for the development of severe retinopathy.

Churchill et al . ( 26 ) found the CC genotype of c.674C>T to be a potential risk factor for AMD. They also described an association with certain haplotypes in the promoter and coding region of VEGFA with AMD. Interestingly, the promoter haplotype contained the c.-460T, the c.-152G and the c.-405C alleles. All of these were also more frequent in the group of PXE patients with visual impairment in our study. Therefore, it seems that carriage of the alleles c.-460T and c.674C we identified as independent risk factors for severe retinopathy also confers to the development of AMD, thus confirming parallels between PXE-associated retinopathy and AMD and pointing to a comparable role for VEGF in both pathologies.

On the basis of the evidence for an involvement of VEGF in PXE-associated retinopathy, it seems justified to assume that polymorphisms leading to altered VEGF expression may modify the course of retinopathy. In the case of AMD, VEGF has been reported to be elevated in patients suffering from the disease ( 22 ). Increased expression may be partly attributable to genetic factors. Indeed, some polymorphisms could be shown to have an impact on VEGF expression. Buraczynska et al . ( 25 ). reported higher VEGF serum concentration in DR patients carrying a deletion in the VEGF promoter in homozygous state which had previously been demonstrated to be in perfect LD with the SNP c.-1540A>C we examined in our study. The corresponding allele c.-1540C was significantly more common in the group of PXE patients with severe visual impairment. Lambrechts et al . ( 31 ) showed association of two haplotypes containing both the c.-1540A (−2578 in their study) and the c.405G allele with lower circulating VEGF levels in patients with amyotrophic lateral sclerosis. This is in concordance with two haplotypes containing these alleles that were identified in our study cohort and were associated with the mild-ocular-involvement group. Evidence thus suggests altered expression of VEGF as the mechanism by which polymorphisms confer to the increased susceptibility to CNV seen in PXE patients carrying alleles associated with increased VEGF production. Unfortunately, no data is available regarding the effect of the polymorphisms c.-460C>T and c.674C>T, which we identified as independent risk factors, on VEGF expression. However, since c.-460C>T is in strong LD with c.-1540A>C, carriers of the risk allele c.-460T will in most cases also be carriers of the c.-1540C allele that seems to lead to enhanced VEGF expression. It should be mentioned in this context that the analysis of VEGF expression is complicated by the fact that several isoforms exist, some of them demonstrating anti-angiogenic properties ( 19 ). It is obvious that the ratio of anti- to pro-angiogenic forms of VEGF must be taken into account when discussing altered VEGF expression.

There are some limitations to our study. Since PXE is a rare disease, cohort size is relatively small and therefore, the statistical power is limited. We have attempted to reduce false-positive results due to multiple testing by using strict Bonferroni correction criteria. This conservative correction assumes independence among markers and tends to underestimate association, so that the associations of single SNP markers and haplotypes we have found in our study are unlikely to be by chance alone. Nevertheless, a further study in a different cohort is needed to confirm our results. Moreover, we have not yet analyzed the effect of VEGFA polymorphisms on VEGF levels in the vitreous fluids of PXE patients because of a lack of suitable samples.

In summary, the results we present in this study represent strong evidence for an association of VEGFA polymorphisms with PXE. Furthermore, we established a correlation between the VEGFA genotype and the ocular PXE phenotype. Our findings strongly suggest an involvement of VEGF in the pathogenesis of ocular PXE manifestations. Association of VEGF with AMD as well as PXE-associated retinopathy is likely explained by an alteration of VEGF expression due to carriage of mutant alleles. It can be speculated that the underlying pathomechanisms are basically the same and that the progress of retinopathy is somehow accelerated in PXE due to certain features of this disease. VEGF gene polymorphisms might prove useful as a prognostic marker for development of PXE-associated retinopathy. Additionally, they may permit earlier therapeutic intervention in order to prevent loss of central vision, one of the most devastating consequences of this disease, given the fact that treatment is most efficient when the disease has not progressed far.

MATERIALS AND METHODS

Patient characteristics

The study cohort comprised 163 German patients with PXE from 150 non-consanguineous families with apparently autosomal recessive or sporadic mode of inheritance of the PXE phenotype. The diagnosis of PXE in all patients was consistent with the reported consensus criteria ( 32 ). The status of the PXE patients was determined by the presence of ocular findings and dermal lesions and histologically confirmed by the observation of calcified elastic fibers in skin biopsies after von Kossa staining. The biopsy samples were taken from lesional skin. All members of the study were thoroughly questioned about their personal diseases, organ involvements and family history by one medical specialist in order to minimize inter-observer variability. Blood samples from 163 Westphalian blood donors were used as controls for analysis of the polymorphisms. Blood donors did not suffer from any known disease. Especially, they neither had any visual impairment nor were they diabetics. Clinical characteristics of patients and controls are summarized in Table  7 . To assess the impact of VEGFA polymorphisms on ocular involvement, we further subdivided the PXE cohort into two groups based on visual impairment due to PXE-associated retinopathy. Group A comprised subjects with no or only mild retinopathy but without impairment of vision and was defined as either without fundus findings or with a finding of peau d'orange and/or angioid streaks. Group B comprised subjects with severe retinopathy accompanied by visual impairment, defined as visual acuity worse than 20/20 in at least one eye. Subgroup characteristics are given in Table  3 . The study was approved by the institutional review board; all patients gave their informed consent.

Table 7.

Characteristics of PXE patients and healthy controls

Characteristics PXE patients Blood donors 
n 163 163 
Males (females) 48 (115) 48 (115) 
Age, years a 48.5 ± 15.3 47.6 ± 14.4 
Age at onset a 30.9 ± 16.0 NA 
Characteristics PXE patients Blood donors 
n 163 163 
Males (females) 48 (115) 48 (115) 
Age, years a 48.5 ± 15.3 47.6 ± 14.4 
Age at onset a 30.9 ± 16.0 NA 

NA, not applicable.

a Mean ± SD.

DNA preparation

Genomic DNA was extracted from 200 µl EDTA blood using the QIAamp blood kit (Qiagen) according to the manufacturer's instructions.

Polymerase chain reaction

Polymerase chain reaction (PCR) primers were designed using the published sequence (GenBank accession no. NG_008732). PCR was performed in 50 µl reaction volume, containing ∼65 ng genomic DNA, 25 pmol of each primer (Biomers), 1.5 U HotStar Taq DNA polymerase (Qiagen) in 1× Reaction Buffer supplied with the enzyme and 0.25 m m of each dNTP (Promega). The PCR conditions were as follows: initial denaturation at 95°C for 15 min, 35 cycles of denaturation at 94°C for 1 min, annealing for 1 min, extension at 72°C for 1 min and final extension at 72°C for 15 min. The primer sequences, annealing temperatures and sizes of the PCR products are summarized in Table  8 .

Table 8.

Primer sequences

Polymorphism Upper primer Lower primer Annealing temperature (°C) 
c.-1540A>C GGCCTTAGGACACCATA AGGAAGCAGCTTGGAA 61 
c.-460C>T TGCGTGTGGGGTTGAGTG(C/T) GGCTCTGCGGACGCTCAGTGA 59/63 
c.-152G>A TGTCCGCACGTAACCTCA GATCCTCCCCGCTACCAG 64 
c.405C>G CGACGGCTTGGGGAGATTGCTCT CAGGTCACTCACTTTGCCCCGGTC 65 
c.674C>T CGCAAGTTCCTCAGACCC ACCCATTCCCATGACACC 60 
c.1032C>T GGTGTGCGCAGACAGTGCTCCAG CACCCAAGACAGCAGAAAGTTCATGGTCTC 65 
c.4618C>T ATGCCCACCACCTTCTCA CCTACGTCCTCCCCACAAC 62 
c.5092C>A ACGTTAGATTTTGGAAGGA(A/C) GGATGGCTTGAAGATGTAC 59/61 
c.9109C>T GGGCTGTGAATGACTGGA GTGGGCTAAAGTAGGGTGTG 62 
c.9162C>T GGGCTGTGAATGACTGGA GTGGGCTAAAGTAGGGTGTG 62 
Polymorphism Upper primer Lower primer Annealing temperature (°C) 
c.-1540A>C GGCCTTAGGACACCATA AGGAAGCAGCTTGGAA 61 
c.-460C>T TGCGTGTGGGGTTGAGTG(C/T) GGCTCTGCGGACGCTCAGTGA 59/63 
c.-152G>A TGTCCGCACGTAACCTCA GATCCTCCCCGCTACCAG 64 
c.405C>G CGACGGCTTGGGGAGATTGCTCT CAGGTCACTCACTTTGCCCCGGTC 65 
c.674C>T CGCAAGTTCCTCAGACCC ACCCATTCCCATGACACC 60 
c.1032C>T GGTGTGCGCAGACAGTGCTCCAG CACCCAAGACAGCAGAAAGTTCATGGTCTC 65 
c.4618C>T ATGCCCACCACCTTCTCA CCTACGTCCTCCCCACAAC 62 
c.5092C>A ACGTTAGATTTTGGAAGGA(A/C) GGATGGCTTGAAGATGTAC 59/61 
c.9109C>T GGGCTGTGAATGACTGGA GTGGGCTAAAGTAGGGTGTG 62 
c.9162C>T GGGCTGTGAATGACTGGA GTGGGCTAAAGTAGGGTGTG 62 

Allele-specific PCR

The polymorphisms c.-460C>T and c.5092C>A were analyzed using allele-specific PCR. PCR was performed twice as described above using primers with the specific sequence for each allele. Agarose gel electrophoresis was applied to detect the amplification products.

Restriction fragment length polymorphism analysis

The obtained DNA fragments were digested at 37°C with 5 U of either Bgl II (c.-1540A>C), Dde I (c.-152G>A and c.1032C>T), Ava II (c.405C>G), Bme 1580I (c.674C>T), Mly I (c.4618C>T) or Msp I (c.9109C>T and c.9162C>T) overnight and subsequently separated on a 1.5% agarose gel. All restriction endonucleases were purchased from New England Biolabs.

Statistical analysis

Allele and haplotype frequencies were compared between cases and controls using Fisher's exact test. Correction for multiple testing was performed using the Bonferroni method. The χ 2 -test was used to examine whether the genotype distributions were within the Hardy–Weinberg equilibrium. P -values of less than 0.05 were considered significant after Bonferroni correction. All tests were performed using SPSS 15.0 (SPSS Inc.).

LD structure and identification of haplotype blocks

Determination of LD and haplotype blocks and frequencies were performed using Haploview 4.0 ( 33 ). Haplotype blocks were defined according to the ‘spine of LD’ setting in the Haploview software, on the basis of each end marker of a block having a D ′ value of more than 0.8.

ACKNOWLEDGEMENTS

We are grateful to all the PXE patients and their relatives, whose cooperation made this study possible. Furthermore, we would like to thank Peter Hof, chairman of the Selbsthilfe für PXE Erkrankte Deutschlands e.V ., and the members of the clinical ambulance for PXE at the Bethesda hospital in Freudenberg, Germany. We would like to thank Ulf Diekmann and Leif Krückemeier for their excellent technical assistance and Sarah L. Kirkby for her linguistic advice.

Conflict of Interest statement . None declared.

REFERENCES

1
Chassaing
N.
Martin
L.
Calvas
P.
Le Bert
M.
Hovnanian
A.
Pseudoxanthoma elasticum: a clinical, pathophysiological and genetic update including 11 novel ABCC6 mutations
J. Med. Genet.
 , 
2005
, vol. 
42
 (pg. 
881
-
892
)
2
Ladewig
M.S.
Götting
C.
Szliska
C.
Issa
P.C.
Helb
H.M.
Bedenicki
I.
Scholl
H.P.N.
Holz
F.G.
Pseudoxanthoma elasticum: Genetische Grundlagen, Manifestationen und Therapeutische Ansätze
Der Ophthalmologe
 , 
2006
, vol. 
103
 (pg. 
537
-
553
)
3
Le Saux
O.
Urban
Z.
Tschuch
C.
Csiszar
K.
Bacchelli
B.
Quaglino
D.
Pasquali-Ronchetti
I.
Pope
F.M.
Richards
A.
Terry
S.
, et al.  . 
Mutations in a gene encoding an ABC transporter cause pseudoxanthoma elasticum
Nat. Genet.
 , 
2000
, vol. 
25
 (pg. 
223
-
227
)
4
Plomp
A.S.
Florijn
R.J.
ten Brink
J.
Castle
B.
Kingston
H.
Martin-Santiago
A.
Gorgels
T.G.M.F.
de Jong
P.T.V.M.
Bergen
A.A.B.
ABCC6 mutations in pseudoxanthoma elasticum: an update including eight novel ones
Mol. Vis.
 , 
2007
, vol. 
14
 (pg. 
118
-
124
)
5
Ratajewski
M.
Van de Ven
W.J.M.
Bartosz
G.
Pulaski
L.
The human pseudoxanthoma elasticum gene ABCC6 is transcriptionally regulated by PLAG family transcription factors
Hum. Genet.
 , 
2008
, vol. 
124
 (pg. 
451
-
463
)
6
Le Saux
O.
Bunda
S.
Van Mart
C.M.
Douet
V.
Got
L.
Martin
L.
Hinek
A.
Serum factors from pseudoxanthoma elasticum patients alter elastic fiber formation in vitro
J. Invest. Dermatol.
 , 
2006
, vol. 
126
 (pg. 
1497
-
1505
)
7
Hu
X.
Plomp
A.S.
van Soest
S.
Wijnholds
J.
de Jong
P.T.V.M.
Bergen
A.A.B.
Pseudoxanthoma elasticum: a clinical, histopathological, and molecular update
Surv. Ophthalmol.
 , 
2003
, vol. 
48
 (pg. 
424
-
438
)
8
Chassaing
N.
Martin
L.
Mazereeuw
J.
Barrie
L.
Nizard
S.
Novel ABCC6 mutations in pseudoxanthoma elasticum
J. Invest. Dermatol.
 , 
2004
, vol. 
122
 (pg. 
608
-
613
)
9
Hendig
D.
Arndt
M.
Szliska
C.
Kleesiek
K.
Götting
C.
SPP1 promotor polymorphisms: identification of the first modifier gene for pseudoxanhtoma elasticum
Clin. Chem.
 , 
2007
, vol. 
53
 (pg. 
829
-
836
)
10
Schön
S.
Schulz
V.
Prante
C.
Hendig
D.
Szliska
C.
Kuhn
J.
Kleesiek
K.
Götting
C.
Polymorphisms in the xylosyltransferase genes cause higher serum XT-I activity in patients with pseudoxanthoma elasticum (PXE) and are involved in a severe disease course
J. Med. Genet.
 , 
2006
, vol. 
43
 (pg. 
745
-
749
)
11
Zarbock
R.
Hendig
D.
Szliska
C.
Kleesiek
K.
Götting
C.
Pseudoxanthoma elasticum: genetic variations in antioxidant genes are risk factors for early disease onset
Clin. Chem.
 , 
2007
, vol. 
53
 (pg. 
1734
-
1740
)
12
Finger
R.P.
Charbel Issa
P.
Ladewig
M.
Holz
F.G.
Scholl
H.P.N.
Intravitreal bevacizumab for choroidal neovascularisation associated with pseudoxanthoma elasticum
Br. J. Ophthalmol.
 , 
2008
, vol. 
92
 (pg. 
483
-
487
)
13
Beatty
S.
Koh
H.-H.
Henson
D.
Boulton
M.
The role of oxidative stress in the pathogenesis of age-related macular degeneration
Surv. Ophthalmol.
 , 
2000
, vol. 
45
 (pg. 
115
-
134
)
14
Klein
R.
Klein
B.E.
Linton
K.L.
Prevalence of age-related maculopathy. The Beaver Dam Eye study
Ophthalmology
 , 
1992
, vol. 
99
 (pg. 
933
-
943
)
15
Vingerling
J.R.
Dielemans
I.
Hofman
A.
Grobbee
D.E.
Hijmering
M.
Kramer
C.F.L.
de Jong
P.T.V.M.
The prevalence of age-related maculopathy in the Rotterdam study
Ophthalmology
 , 
1995
, vol. 
102
 (pg. 
205
-
210
)
16
Spaide
R.F.
Armstrong
D.
Browne
R.
Choroidal neovascularization in age related macular degeneration—what is the cause?
Retina
 , 
2003
, vol. 
23
 (pg. 
595
-
614
)
17
Götting
C.
Hendig
D.
Zarbock
R.
Szliska
C.
Kleesiek
K.
Complement factor H variant p.Y402H in pseudoxanthoma elasticum patients
Genet. Test.
 , 
2008
, vol. 
12
 (pg. 
1
-
6
)
18
Haines
J.L.
Schnetz-Boutaud
N.
Schmidt
S.
Scott
W.K.
Agarwal
A.
Postel
E.A.
Olson
L.
Kenealy
S.J.
Hauser
M.
Gilbert
J.R.
, et al.  . 
Functional candidate genes in age-related macular degeneration: significant association with VEGF, VLDLR, and LRP6
Invest. Ophthalmol. Vis. Sci.
 , 
2006
, vol. 
47
 (pg. 
329
-
335
)
19
Ferrara
N.
Davis-Smyth
T.
The biology of vascular endothelial growth factor
Endocr. Rev.
 , 
1997
, vol. 
18
 (pg. 
4
-
25
)
20
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
, vol. 
37
 (pg. 
443
-
448
)
21
Awata
T.
Inoue
K.
Kurihara
S.
Ohkubo
T.
Watanabe
M.
Inukai
K.
Inoue
I.
Katayama
S.
A common polymorphism in the 5′-untranslated region of the VEGF gene is associated with diabetic retinopathy in type 2 diabetes
Diabetes
 , 
2002
, vol. 
51
 (pg. 
1635
-
1639
)
22
Kliffen
M.
Sharma
H.S.
Mooy
C.M.
Kerkvliet
S.
de Jong
P.T.V.M.
Increased expression of angiogenic growth factors in age-related maculopathy
Br. J. Ophthalmol.
 , 
1997
, vol. 
81
 (pg. 
154
-
162
)
23
Watson
C.J.
Webb
N.J.A.
Bottomley
M.J.
Brenchley
P.E.C.
Identification of polymorphisms within the vascular endothelial growth factor (VEGF) gene: correlation with variation in VEGF protein production
Cytokine
 , 
2000
, vol. 
12
 (pg. 
1232
-
1235
)
24
Stevens
A.
Soden
J.
Brenchley
P.E.C.
Ralph
S.
Ray
D.W.
Haplotype analysis of the polymorphic human vascular endothelial growth factor gene promoter
Cancer Res.
 , 
2003
, vol. 
63
 (pg. 
812
-
816
)
25
Buraczynska
M.
Ksiazek
P.
Baranowicz-Gaszczyk
I.
Jozwiak
L.
Association of the VEGF gene polymorphism with diabetic retinopathy in type 2 diabetes patients
Nephrol. Dial. Transplant.
 , 
2007
, vol. 
22
 (pg. 
827
-
832
)
26
Churchill
A.J.
Carter
J.G.
Lovell
H.C.
Ramsden
C.
Turner
S.J.
Yeung
A.
Escardo
J.
Atan
D.
VEGF polymorphisms are associated with neovascular age-related macular degeneration
Hum. Mol. Genet.
 , 
2006
, vol. 
15
 (pg. 
2955
-
2961
)
27
Al-Kateb
H.
Mirea
L.
Xie
X.
Sun
L.
Liu
M.
Chen
H.
Bull
S.B.
Boright
A.P.
Paterson
A.D.
Multiple variants in vascular endothelial growth factor (VEGFA) are risk factors for time to severe retinopathy in type 1 diabetes: the DCCT/EDIC Genetics study
Diabetes
 , 
2007
, vol. 
56
 (pg. 
2161
-
2168
)
28
Churchill
A.J.
Carter
J.G.
Ramsden
C.
Turner
S.J.
Yeung
A.
Brenchley
P.E.C.
Ray
D.W.
VEGF polymorphisms are associated with severity of diabetic retinopathy
Invest. Ophthalmol. Vis. Sci.
 , 
2008
, vol. 
49
 (pg. 
3611
-
3616
)
29
Lommatzsch
A.
Management of choroidal vascularisation
Br. J. Ophthalmol.
 , 
2008
, vol. 
92
 (pg. 
445
-
446
)
30
Spraul
C.W.
Lang
G.E.
Grossniklaus
H.E.
Lang
G.K.
Charakteristika von Drusen und Veränderungen der Bruch-Membran in Augen mit altersabhängiger Makuladegeneration: Histologische Studie
Ophthalmologe
 , 
1998
, vol. 
95
 (pg. 
73
-
79
)
31
Lambrechts
D.
Storkebaum
E.
Morimoto
M.
Del-Favero
J.
Desmet
F.
Marklund
S.L.
Wyns
S.
Thijs
V.
Andersson
J.
Marion
I.
, et al.  . 
VEGF is a modifier of amyotrophic lateral sclerosis in mice and humans and protects motoneurons against ischemic cell death
Nat. Genet.
 , 
2003
, vol. 
34
 (pg. 
383
-
394
)
32
Lebwohl
M.
Neldner
K.H.
Pope
F.M.
De Paepe
A.
Christiano
A.M.
Boyd
C.D.
Uitto
J.
McKusick
V.A.
Classification of pseudoxanthoma elasticum: report of a consensus conference
J. Am. Acad. Dermatol.
 , 
1994
, vol. 
30
 (pg. 
103
-
107
)
33
Barrett
J.C.
Fry
B.
Maller
J.
Daly
M.J.
Haploview: analysis and visualization of LD and haplotype maps
Bioinformatics
 , 
2005
, vol. 
21
 (pg. 
263
-
265
)