Portal de Periódicos da CAPES

Analysis of SERPING1 and its association with age-related macular degeneration

2009; Wiley; Volume: 89; Issue: 2 Linguagem: Inglês

10.1111/j.1755-3768.2009.01788.x

1755-3768

James G. Carter, A Churchill,

Retinal and Optic Conditions

Editor, The molecular genetics of age-related macular degeneration (AMD) is a growing area of interest, with studies frequently producing more questions than answers. In 2008, Ennis et al. reported a new association locus; SERPING1, a serine protease inhibitor (serpin) of the C1 complex of the alternative complement pathway (Ennis et al. 2008). In that case–control study, DNA from 958 UK subjects (479 patients, 479 controls) and 500 US subjects (248 patients, 252 controls) were analysed across a panel of 32 genes. An association was found between AMD and several single nucleotide polymorphisms (SNPs) in the SERPING1 gene. However, a recent study was unable to replicate these findings in two independent US cohorts numbering 786 (476 AMD, 310 controls) and 1541 subjects (1241 AMD, 300 controls) (Park et al. 2009). To corroborate one or other of these two conflicting publications, we conducted our own analysis of the SERPING1 AMD-association locus. Exudative AMD (exAMD) subjects were recruited from dedicated clinics providing photodynamic therapy and anti-VEGF treatments at the Bristol Eye Hospital (n = 94). Patient ages ranged from 51 to 94 and had AMD, secondary to choroidal neovascular membranes, as demonstrated by fluorescein angiography. Sex distribution was 28.0% male and 72.0% female with mean ages of 76.0 and 78.4 , respectively. Control subjects (n = 95) were recruited from healthy volunteers accompanying patients and age-matched to within 5 years of ex-AMD cases. All controls underwent visual acuity testing, anterior segment and fundus examination. Sex distribution of control subjects was 33.7% male and 65.3% female, with mean ages of 72.6 and 73.3 , respectively (age range: 55–89 years). Both groups were in Hardy–Weinberg equilibrium. All subjects were genotyped for the SERPING1 polymorphism; rs2511989, IVS6-865 g>a. This locus showed strong association with AMD in the original published data (p = 5.4 × 10−6) (Ennis et al. 2008). Genotyping was determined by direct sequencing (PCR primers, forward: 5′-CTCA GCCTTGGCACTATTGA-3′; reverse: 5′-AGGCAGGAGAATTGCTTGAA-3′; anneal Tm 67°C; 1.5mm MgCl2; 394 bp). Data analysis was attained using statistical software SPSS v15 (SPSS Inc., Chicago, USA). Carriage of the G allele was raised in AMD group versus controls (61.2% versus 57.9%) as was carriage of the GG genotype (40.4% versus 30.5%), but neither was statistically significant (Pearson’s χ2 test; p = 0.52 and p = 0.16, respectively). We observed a higher percentage of GA heterozygotes in the control group (p = 0.07) (Table 1). We did not find a significant association between SERPING1 and exudative AMD in our study. It is important to realize that in smaller studies the chances of type II errors are increased; statistical power for this analysis, based on Ennis et al.’s frequency data, is 24–30%. However, our cohort is still sufficiently large to detect a significant association if similar frequencies as those reported by Ennis et al. had been observed, i.e. G allele frequencies of 63% versus 52% gives a significant (p < 0.05) result even with our smaller sample size. It is also important to note that Ennis et al. were assessing multiple AMD phenotypes, not just the more severe exudative cases, making direct comparisons unsuitable. However, Park et al. did carry out subgroup analysis, including ex-AMD, where no significant association was seen. There are several possible explanations for the conflicting findings between these three studies: different populations have been shown to vary in consistency of linkage disequilibrium, and this may influence the interpretation of tagged SNP results that proxy for other linked polymorphisms hindering direct comparisons (Klaver & Bergen 2008). It is also possible that another polymorphism within SERPING1 is more relevant to AMD than the one studied. For example, an additional association SNP identified by Ennis et al. (rs2511988) located 20bp upstream of the 3’ splice site for exon 7, adjacent to the branch point, could influence splicing efficiency (Kralovicova & Vorechovsky 2009). This polymorphism was not screened by Park et al. or ourselves. Furthermore, there are 18 SERPING1 variants not examined in any of the AMD studies that may have functional effects on coding or protein regulation (Klaver & Bergen 2008). Finally, of course it has to be mentioned that perhaps AMD is not associated with polymorphisms in SERPING1 at all. Ennis et al. have laid the groundwork, but much more work is needed to definitively identify the molecular markers or variants that influence functional efficacy of SERPING1 and to confirm or disprove the biological role that SERPING1 plays in the development of AMD.