Chromosome 10q26 locus and age-related macular degeneration: A progress update

Experimental Eye Research 119 (2014) 1e7

Contents lists available at ScienceDirect

Experimental Eye Research journal homepage: www.elsevier.com/locate/yexer

Review

Chromosome 10q26 locus and age-related macular degeneration: A progress update Gaofeng Wang* John P. Hussman Institute for Human Genomics, Dr. John T. Macdonald Foundation Department of Human Genetics, University of Miami Miller School of Medicine, 1501 N.W. 10th Avenue, BRB 525, M860, Miami, FL 33136, United States

a r t i c l e i n f o

a b s t r a c t

Article history: Received 11 September 2013 Accepted in revised form 18 November 2013 Available online 28 November 2013

Age-related macular degeneration (AMD) is the leading cause of late-onset central vision loss in developed countries. Both genetic and environmental factors contribute to the onset of AMD. Variation at a locus on chromosome 10q26 has been consistently associated with this disease and represents one of the two strongest genetic effects being identified in AMD. At least three genes are located within the bounds of the locus: pleckstrin homology domain containing family A member 1 (PLEKHA1), age-related maculopathy susceptibility 2 (ARMS2) and high-temperature requirement A serine peptidase 1 (HTRA1), all of which are associated with AMD. Due to the strong linkage disequilibrium (LD) across this region, statistical genetic analysis alone is incapable of distinguishing the effect of an individual gene in the locus. Uncertainty remains, however, in regards to which gene is responsible for the linkage and association of the locus with AMD. Investigating functional consequences of the associated variants and related genes tends to be essential to identifying the biologically responsible gene(s) underlying AMD. This review examines the recent progress and current uncertainty on the genetic and functional analyses of the 10q26 locus in AMD with a focus on ARMS2 and HTRA1. A discussion, which entails the possible multi-faceted approaches for pinpointing the gene(s) in the locus underlying the pathogenesis of AMD, is also included. Ó 2013 Elsevier Ltd. All rights reserved.

Keywords: age-related macular degeneration PLEKHA1 ARMS2 HTRA1 single nucleotide polymorphism indel gene annotation promoter transcript splice variants

1. Introduction AMD is the major cause of visual impairment and blindness among individuals over the age of 50 in developed countries. AMD affects the central region of the retina (the macula), which has the highest concentration of cone photoreceptors and is responsible for visual acuity. As a progressive disease, AMD is generally classified into early, intermediate and advanced stages. The advanced AMD is further distinguished as geographic atrophy (dry AMD) and neovascularization (wet AMD). The pathological changes accompanying AMD progression have been documented in detail (reviewed by Ding et al., 2009). However, the etiology and pathogenesis of AMD are not fully understood. AMD has a complex etiology that likely result from the interplay of several risk factors, both genetic and environmental. By genome-wide linkage screen, many groups have contributed to identifying several AMD susceptibility loci. Despite variable phenotype definitions and different analysis approaches, among a list of suggested chromosome regions, the loci at

* Tel.: þ1 305 243 5434; fax: þ1 305 243 2703. E-mail address: [email protected]. 0014-4835/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.exer.2013.11.009

1q31 and 10q26 were repeatedly linked to the disease and confirmed by a meta-analysis (Fisher et al., 2005). The two susceptibility loci have been repeatedly verified in most, if not all, association studies and have recently been reconfirmed in a metaanalysis by the international AMD Gene Consortium (Fritsche et al., 2013). Compared to the loci at 1q31 and 10q26, other associated loci have smaller effects on the risk of AMD. Recent reviews have provided insights with respect to the overall status of genetic studies on AMD (Gorin, 2012; Liu et al., 2012; Ambati and Fowler, 2012). This review particularly focuses on recent progress and current uncertainty around the genetic and functional analyses of the 10q26 locus. 2. Chromosome 10q26 locus d uncertainty on susceptibility gene(s) In 2005, using various approaches, four groups independently reported the CFH (complement factor H, MIM 134370) gene, located on 1q31, as the first major AMD susceptibility gene (Haines et al., 2005; Klein et al., 2005; Edwards et al., 2005; Hageman et al., 2005). Since then, virtually all studies have validated CFH Y402H as a major AMD susceptibility variant in

2

G. Wang / Experimental Eye Research 119 (2014) 1e7

Caucasians. Compared to the successful discovery of CFH Y402H, it has been difficult to identify with certainty the susceptibility variation(s) responsible for the linkage and association at the locus on 10q26, which originally generated the most significant linkage signal in a meta-analysis (Fisher et al., 2005). In 2005, by analyzing coding variants in the locus, Jakobsdottir et al. reported that PLEKHA1 (pleckstrin homology domain containing, family A, member 1, MIM 607772) and hypothetical gene LOC387715 were primarily responsible for the evidence of linkage and could be major contributors to AMD susceptibility (Jakobsdottir et al., 2005). By analyzing 93 single nucleotide polymorphisms (SNPs) spanning 22 Mb at the chromosome 10q26, Rivera et al. disclosed the significant association across a high linkage disequilibrium (LD) region harboring the two genes PLEKHA1 and LOC387715. The strongest association was centered over a coding SNP rs10490924 (amino acid substitute A69S in LOC387715), suggesting a role of LOC387715 in the pathogenesis of AMD (Rivera et al., 2005). Through the employment of linkage analysis methods followed by family-based and case-control association analyses of a total of 185 SNPs in the locus, Schmidt et al. showed that SNP rs10490924 in LOC387715 also generated the strongest association signal and a strong statistical interaction between the LOC387715 variant and a history of cigarette smoking (Schmidt et al., 2006). These studies suggested that LOC387715 could be the responsible gene in the locus (Fig. 1). Two studies soon followed that pointed to the gene HTRA1 (high-temperature requirement, MIM 602194), rather than its adjacent LOC387715, as the second major AMD susceptibility gene. In 2006, Yang et al. reported that SNP rs11200638 at gene HTRA1 was the most significantly associated variant by analyzing 15 variants including SNP rs10490924 of LOC387715 in a Caucasian casecontrol cohort. The preliminary functional analysis showed an effect of SNP rs11200638 on HTRA1 expression and the presence of immunoreactive HTRA1 protein in drusen. HTRA1 was suggested as the responsible gene at the chromosome 10q26 locus (Yang et al., 2006). In another genome-wide association study, SNP rs11200638 (odds ratio (OR) ¼ 10.00, P ¼ 8.24  1012) and rs10490924 (OR ¼ 11.14, P ¼ 2.4  1012) were almost equally

associated with AMD in a Chinese population. Nevertheless, the conclusion pointed to HTRA1 based on the assumption that transcripts of LOC387715 are not existed in the retina (Dewan et al., 2006). Interestingly, by evaluating 45 tag SNPs spanning genes PLEKHA1, LOC387715 and HTRA1 in a Caucasian case-control dataset, Kanda et al. demonstrated that SNP rs10490924 in LOC387715 alone could explain the bulk of the association in the 10q26 chromosomal region. The hypothetical LOC387715 was renamed as age-related maculopathy susceptibility 2 (ARMS2, MIM 611313) and variation at rs10490924 was again suggested as the second major AMD susceptibility allele (Kanda et al., 2007). Soon after, a complex deletioneinsertion polymorphism (indel; consisting of a 443 bp deletion and an adjacent 54 bp insertion) in the ARMS2 30 UTR and flanking region was identified by sequencing. Fritsche et al. reported a strong association of the indel with AMD and suggested a pathogenic role of the indel by decreasing the stability of ARMS2 transcripts (Fritsche et al., 2008). Many of the follow-up studies have validated the strong association of variants rs11200638, rs10490924 and the indel with the risk of AMD in different samples (Shastry et al., 2006; Conley et al., 2006; Ross et al., 2007; Francis et al., 2007; Seddon et al., 2007; Mori et al., 2007; Weger et al., 2007; Deangelis et al., 2008; Wang et al., 2009; Andreoli et al., 2009; Hadley et al., 2010; Wang et al., 2010a; Hayashi et al., 2010; Fuse et al., 2011; Sundaresan et al., 2012; Chakravarthy et al., 2013). However, these three significantly associated variants are in strong LD and generate virtually identical association test and odds ratios, indicating the limited ability of statistical genetic analysis to establish the most likely disease-associated variant. Overall, the most significantly associated haplotype in the 10q26 locus at least includes rs10490924 and the indel of ARMS2 and rs11200638 of HTRA1 (Fig. 1). There are three likely possibilities: (1) both ARMS2 and HTRA1 influence the risk of AMD; (2) only one gene (ARMS2 or HTRA1) contributes to AMD development; (3) another gene (such as PLEKHA1) or unannotated non-coding RNA derived from this locus confers susceptibility. It is imperative to recognize that only via analyzing functional consequences of the AMD-associated variants and

Chromosome 10

3′

3′

5′

3′

rs1049331 rs2293870

HTRA1

rs11200638

Indel

Exon 2

Exon 1

5′

ARMS2

rs10490924 (A69S)

PLEKHA1

rs3793917

5′′

10q26 locus

Exon 1 HTRA1

ARMS2 Coding exon

UTR

Intron or intergenic region

CpG island

Fig. 1. An overview of the 10q26 locus, showing the three genes and the most significantly associated variants.

G. Wang / Experimental Eye Research 119 (2014) 1e7

uncovering the function of ARMS2 and HTRA1 proteins in the retina can this genetic puzzle be solved. 3. HTRA1 d a plausible but inconsistent candidate Human HTRA1 gene was originally identified by detecting an extensive homology to the Escherichia coli HtrA family of serine protease that is essential for bacterial survival at high temperatures (Zumbrunn and Trueb, 1996). HTRA1 is ubiquitously expressed, but it was soon discovered either absent or significantly downregulated in various tumors. Further, overexpression of HTRA1 inhibits tumor growth, suggesting a tumor suppressor function for HTRA1 (Baldi et al., 2002; Chien et al., 2004, 2006). HTRA1 protein is known as a secreted protease that has the capacity to degrade numerous extracellular matrix proteins. Furthermore, HTRA1 protein regulates the signaling of transforming growth factor-beta (TGFb) family proteins, insulin-like growth factors (IGF) and fibroblast growth factors (FGF) mostly by physical proteineprotein interactions. Due to fact that the basic function of ARMS2 remains unknown, HTRA1 was generally regarded as a better functional candidate. Current researches have demonstrated the functional plausibleness of HTRA1 in the pathology and pathogenesis of AMD. However, whether the AMD-associated variants change the expression and/or function of HTRA1 in the retina remains inconsistent even controversial. These inconsistent results of the genetic variants on HTRA1 somewhat diminish the plausibility of HTRA1 as being a possible responsible gene at the locus. 3.1. Promoter variants and HTRA1 transcription Although variants such as rs3793917 in the promoter/intergenic region of HTRA1 have been repeatedly associated with the risk of AMD (Hadley et al., 2010; Richardson et al., 2010; Tian et al., 2012), most attention of functional analysis has been focused on rs11200638. SNP rs11200638 is located within a conserved CpG island and 497 bp upstream from the transcription start site of HTRA1 (Fig. 1). Compared to the major allele G of rs11200638, the AMD-associated minor allele A may disrupt the CG pattern in the region, which could result in an altered transcription of HTRA1. The risk genotype AA at rs11200638 was primarily correlated with a higher level of HTRA1 mRNA and protein by examining blood lymphocytes and retinal pigment epithelium (RPE) from a small number of patients and controls (Yang et al., 2006). Interestingly, this correlation was subsequently replicated in archived eye tissues and fresh placenta tissues by quantitative RT-PCR and immunohistochemistry (Chan et al., 2007; Tuo et al., 2008; Yang et al., 2010). Thus, an increased level of HTRA1 was suggested to play a potential role in the pathogenesis of AMD. However, several following studies, which utilized much larger numbers of samples, have repeatedly demonstrated that genotypes at rs11200638 or other AMD-associated variants in the locus are not correlated with the transcription level of HTRA1 in the retina or lymphocyte (Kanda et al., 2007, 2010; Chowers et al., 2008; Wang et al., 2010b; 2013a; Friedrich et al., 2011). Further discrediting the potential role of HTRA1, a recent systemic survey for potential biomarkers of AMD revealed many genes that are either over- or under-expressed in AMD-affected RPE/choroid. HTRA1 was not on the list of those differentially expressed genes (Newman et al., 2012). These discordant results of rs11200638 on HTRA1 expression call into question the proposed HTRA1-AMD functional association. Additionally, results from the in vitro analysis of HTRA1 promoter activity are not consistent either. DeWan et al. primarily observed that the A allele in luciferase constructs for rs11200638 generated slightly higher luciferase activity than the G allele constructs, but it was not statistically significant (Dewan et al., 2006).

3

Kanda et al. further analyzed the HTRA1 promoter and concluded that there was no difference in promoter activity between the two alleles at rs11200638 (Kanda et al., 2007). By cloning longer fragments of the HTRA1 promoter region, Yang et al. showed that the haplotype tagged by the indel and rs11200638 had an enhancing effect on the transcription of HTRA1 (Yang et al., 2010). However, Friedrich et al. did not verify this reported effect of the haplotype by cloning the similar fragment of HTRA1 promoter region (Friedrich et al., 2011). Overall, ex vivo and in vitro studies have unfortunately presented inconsistent and conflicting results surrounding the potential function of SNP rs11200638 and other variants on HTRA1 transcription. These inconsistencies remain one of the major controversies surrounding 10q26 locus in AMD and weaken the notion that an up-regulated transcription of HTRA1 is the AMD susceptibility factor. Future studies may be able to produce conclusive results by applying robust and unbiased approaches such as highthroughput sequencing on large numbers of human eye tissues. Until such an analysis is thoroughly conducted, it is premature to conclude or exclude HTRA1 as a responsible gene at the locus. 3.2. Synonymous variants and HTRA1 translation Two synonymous SNPs rs1049331 (A34A) and rs2293870 (G36G) in the exon 1 of HTRA1, which are located within the strong LD with rs11200638, have also been associated with the risk of AMD or sub-phenotypes of AMD (Deangelis et al., 2008; Tam et al., 2008; Andreoli et al., 2009). Although a synonymous SNP will not result in an amino acid change, there is a possibility that the two synonymous SNPs may alter HTRA1 translation by changing mRNA structure or tRNA preference as previously reported in other genes (Lavner and Kotlar, 2005; Chamary et al., 2006; Sauna and KimchiSarfaty, 2011). In silico analysis showed that SNPs rs1049331 and rs2293870 could only cause slight changes in the secondary structure of HTRA1 mRNA. Wang et al. observed no significant changes of HTRA1 mRNA and protein levels among haplotypes at SNPs rs1049331 and rs2293870 (Wang et al., 2013a). Interestingly, Jacobo et al. recently reported that variation at rs1049331 and rs2293870 could implicate a frequent-to-rare codon conversion that reduced the translation rate of HTRA1 mRNA. HTRA1 protein being produced from the SNP-containing mRNA appeared to be more susceptible to proteolysis and had a reduced binding capacity to IGF-1 (Jacobo et al., 2013). So far, all three possible consequences of HTRA1 expression, i.e. up-regulated, down-regulated and notchanged, caused by the AMD-associated variants within the strong LD have been presented in the literature. 3.3. HTRA1 and AMD modeling Current knowledge regarding HTRA1 protein suggests that HTRA1 can be a functionally plausible candidate relevant to AMD for multiple reasons. First, HTRA1 protein appears to deposit in drusen, the pathological hallmark of dry AMD (Yang et al., 2006; Chan et al., 2007). Secondly, loss-of-function mutations in HTRA1 cause familial ischemic cerebral small-vessel disease that is associated with intimal thickening collagen fibers and loss of vascular smooth-muscle cells (Hara et al., 2009; Shiga et al., 2011). Through inhibiting the signaling of TGFb family members, HTRA1 protein appears to have a role in regulating angiogenesis, which is the most crucial and distinguishing aspect of wet AMD (Oka et al., 2004; Launay et al., 2008). Although an up-regulated HTRA1 has not yet been fully established in AMD, animal modeling has been focused on overexpressing HTRA1 in rodents. One group observed a fragmented elastic layer in Bruch’s membrane when overexpressing mouse HTRA1 in mouse RPE and another group reported polypoidal

4

G. Wang / Experimental Eye Research 119 (2014) 1e7

choroidal vasculopathy after overexpressing human HTRA1 in mouse RPE (Vierkotten et al., 2011; Jones et al., 2011). Furthermore, vascular underdevelopment appeared in the retina of HTRA1 knockout mice (Zhang et al., 2012). The findings of these studies strongly suggest that HTRA1 could be involved in the pathogenesis of AMD. 3.4. Toward establishing HTRA1 as a susceptibility gene It can be obviously noted that HTRA1 is functionally plausible in AMD. However, the key questions to be answered in order to establish HTRA1 as a responsible gene at the locus have not yet been fully addressed, (1) Do AMD-associated promoter variants enhance HTRA1 transcription? (2) Is HTRA1 expressed higher in the retinas of AMD patients as compared to that of control groups? In order to resolve the inconsistent results, novel approaches should be implemented in future studies. First, multiple HTRA1 alternative splice transcripts are annotated in human genome databases, which potentially could be expressed in the retina and be translated into different protein isoforms. It is also possible that some novel splice transcripts in the retina or in the AMD-affected retina have not yet been identified. High-throughput sequencing-based analyses can help identify novel HTRA1 transcripts and allow a more accurate measurement of HTRA1 transcripts. Secondly, a large number of fresh human eye tissues especially from AMD-affected individuals are required to obtain adequate statistical power to determine the changes, especially minor changes in HTRA1, which could be accomplished in future studies. 4. ARMS2 d a plausible but function unknown candidate The most significantly associated haplotype includes rs10490924 (nonsynonymous change A69S) in ARMS2. This A69S coding variant could potentially change the function of ARMS2, which implicates a plausible candidate for AMD. However, compared to the relatively conserved HTRA1, ARMS2 is currently only annotated in genomes of humans and some higher primates (Francis et al., 2008; Pahl et al., 2012). Lack of homologous genes in non-primate model organisms such as mice makes it difficult to study ARMS2. Furthermore, in the protein databases such as UniProt, the ARMS2 protein is listed as being evidenced only at the level of transcripts. Due to the lack of a consensus sequence of ARM2 protein, the basic biological function of ARMS2 remains largely unknown. Overall, the role of ARMS2 in AMD pathogenesis remains an enigma. 4.1. Gene annotation For an accurate gene annotation, expressed sequence tags (EST) are required in addition to sequence similarity comparisons and computational predictions (Mount, 2000; Ashurst and Collins, 2003). However, in genome databases, all ARMS2 ESTs are from the placenta and HT1080 cell (a fibrosarcoma cell line) and no EST from the human retina is available. Part of ARMS2 transcripts has been observed in the retina by different groups (Rivera et al., 2005; Kanda et al., 2007; Fritsche et al., 2008; Wang et al., 2010a, b; Yang et al., 2010). Applying rapid amplification of cDNA ends (RACE) assays, Wang et al. recently found that the transcription start site for ARMS2 in the retina is at Chr10:124,213,959 (NCBI GRCh37), which is 220 bp upstream from the annotated site. The polyadenylation site of ARMS2 transcripts has been verified at Chr10:124,216,867 as annotated. By RT-PCR and sequencing, Wang et al. identified two ARMS2 transcript isoforms, the annotated transcript being designated as isoform-A and a novel transcript splice variant being designated as isoform-B. Both isoforms share

the same splice donor site GT at the genomic DNA position (chr10:124,214,541e124,214,542) (NCBI GRCh37). Isoform-A was spliced at the conventional acceptor site AG at chr10:124,216,421e 124,216,422. Isoform-B was spliced at an alternative acceptor site TG at chr10:124,216,608e124,216,609 (Fig. 2). TG dinucleotides have been reported as an alternative splice acceptor site in many genes (Szafranski et al., 2007). The sequence of one ARMS2 EST (BG194076) also suggests the same splice site as isoform-B (Harrington et al., 2001). Furthermore, it is also possible that some novel ARMS2 transcript splice variants other than isoform-B exists, but have yet to be identified in the retina. Highthroughput sequencing of transcriptomes of the retina will help illuminate a full picture of transcripts derived from the ARMS2 gene in future studies. 4.2. Coding variants and the risk of AMD In addition to the A69S variant, other common coding variants in ARMS2 also have been evaluated in the risk of AMD. BergeronSawitzke et al. first reported protective trends of SNPs rs10490923 (R3H) and rs2736911 (R38X, which introduces a premature stop codon) in AMD but the results were not statistically significant (Bergeron-Sawitzke et al., 2009). Yang et al. soon reported that R38X is inversely associated with AMD in larger multicohorts (Yang et al., 2010). This inverse association was somehow verified in a small Polish sample (90 wet AMD cases and 40 controls, P ¼ 0.053) (Teper et al., 2012). However, by analyzing all three common coding variants (R3H, R38X and A69S) of ARMS2, Wang et al. found that the inverse association of R3H or R38X with AMD is insignificant after adjustment for sex and age in a big dataset (1169 AMD cases and 707 controls) (Wang et al., 2013b). It is likely that the genetic association of R3H and R38X in AMD is either weak or insignificant. 4.3. Variants and the stability of ARMS2 transcripts Fritsche et al. first identified that the indel consisted a 443 bp deletion and an adjacent 54 bp insertion (Fritsche et al., 2008). Wang et al. found that this indel was even more complex and was composed of two side-by-side indels separated by 17 bp: (1) 9 bp deletion with 10 bp insertion; (2) 417 bp deletion with 27 bp insertion. Nevertheless, the indel or indels delete the original polyA signal of ARMS2 and insert two ATTTA fragments, which are transcribed to two AUUUA motif located in the 30 UTR that potentially can mediate mRNA degradation (Bolognani and Perrone-Bizzozero, 2008). Initially, the indel was correlated with a lower level of exogenous ARMS2 transcripts in cultured cells and endogenous ARMS2 protein in placentas where ARMS2 is highly expressed (Fritsche et al., 2008). A deficient expression of ARMS2 caused by the indel was thus suggested to functionally contribute to the onset of AMD. On the other side, the nonsense change R38X introduces a premature stop codon and could potentially lead to lower ARMS2 expression due to nonsense-mediated mRNA decay (Chang et al., 2007). It is reported that R38X is associated with a lower level of ARMS2 expression (Yang et al., 2010). A possible role of the decreased level of ARMS2 transcripts in AMD etiology is thus put in a paradoxical position. If both the indel and R38X, as reported, decrease the level of ARMS2 transcripts, it would be impossible to explain how on one hand, the indel confers the risk of AMD, yet R38X might be protective against the development of the disease. Subsequent attempts to replicate this effect of the indel and R38X on the stability of ARMS2 transcripts have been mixed. Some studies verified the effect of the indel (Yang et al., 2010; Friedrich et al., 2011), while others found that genotypes at the indel do not change the level of ARMS2 transcripts in human retinas (Kanda

G. Wang / Experimental Eye Research 119 (2014) 1e7

5

chr10: 124,216,609

ARMS2 gene Exon 1

Exon 2

5

3

Transcripts Isoform-A

Isoform-B

1057 nt

Proteins (predicted)

99

99 107 a.a

129 a.a

Isoform-A

Coding exon

UTR

870 nt

Isoform-B

Intron or intergenic region

Protein

Fig. 2. An overview of the alternative splice of ARMS2 transcripts. The sequence traces show the two splice sites that result in two transcripts isoforms, which predict two protein isoforms with different carboxyl tails.

et al., 2010; Wang et al., 2010a). It should be noted that the AUUUA motif alone may not be sufficient to enhance mRNA degradation even if the inserted AUUUA is transcribed (Lagnado et al., 1994). Furthermore, for R38X to execute “nonsense-mediated mRNA decay”, the precondition is that the annotated translational start site is correct, or at least in-frame with the annotated site. If that isn’t the case, SNP rs2736911 won’t result in a premature stop codon (R38X), and it will no longer be a nonsense variant. Minor et al. recently reported that allele X at R38X decreased exogenous ARMS2 transcripts in cultured cells possibly by accelerating the degradation of the transcripts. However, genotypes at R38X and the indel do not obviously affect the level of ARMS2 transcripts in human eye tissues possibly due to the complicated regulation of ARMS2 expression in vivo (Minor et al., 2013). From the published data, it still remains difficult to conclude or exclude with certainty whether the indel or R38X actually function in altering the stability of ARMS2 transcripts. 4.4. The function of ARMS2 protein d unclear The basic biological function of ARMS2 protein still remains unclear. Minor et al. recently reported that R38X has the capability to mediate transcript decay in cultured cells, suggesting that R38X could be a nonsense variant and that the annotated translational site may be correct. Furthermore, a unique peptide fragment sequence matching the N-terminus of the predicted ARMS2 protein was identified by proteome-wide screening (Hornbeck et al., 2012), suggesting that ARMS2 protein is translated from the annotated translational start site. The two known ARMS2 transcript isoforms predict two protein isoforms, one with 107 amino acids and the other 129 amino acids. The difference between these two protein isoforms exists at the carboxyl terminal. Currently, a

conclusively verified sequence of ARMS2 proteins has yet to be attained. Current studies on ARMS2 protein distribution in the retina and subcellular targeting are inconsistent and controversial. At least three distinct patterns of ARMS2 protein localization (mitochondrial, cytosolic and secreted) have been reported (Kanda et al., 2007; Fritsche et al., 2008; Wang et al., 2009; Kortvely et al., 2010). Using immunogen-tagged ARMS2 constructs, Kanda et al., reported that ARMS2 is located at mitochondrial outer membrane in transfected COS1 cells (Kanda et al., 2007). By immunohistochemistry, Fritsche et al. further showed that immunoreactive ARMS2 is only localized in the mitochondria of photoreceptors in human retina (Fritsche et al., 2008). Due to the important role of mitochondria in degenerative diseases, a mitochondrial ARMS2 is very attractive in constructing a hypothesis of its role in the pathogenesis of AMD. However, applying the reported experimental design and strategy, Wang et al. found that ARMS2 is largely a cytosolic protein, not a mitochondrial protein in retinal epithelial ARPE-19 cells and COS7 cells (Wang et al., 2009). Yeast-twohybridization (Y2H) is frequently used in the laboratory to screen physical binding partners of proteins. The resulted binding partners could suggest the subcellular localization and possible functions of the bait protein. Interestingly, using placental ARMS2 cDNA as bait in Y2H, Körtvely et al., reported that ARMS2 interacts mainly with extracellular matrix proteins, suggesting that ARMS2 is potentially a secreted protein. Further, the immunoreactive ARMS2 is mostly confined to choroid pillars, instead of inner segment of photoreceptors, in human eyes (Kortvely et al., 2010). Additionally, potential functions of ARMS2 have recently been explored in cultured cells (Xu et al., 2012; Cheng et al., 2013). Overall, the current findings on ARMS2 protein distribution in the retina and subcellular localization need further supporting evidence.

6

G. Wang / Experimental Eye Research 119 (2014) 1e7

4.5. Toward establishing ARMS2 as a susceptibility gene Although, the strongly associated coding variant A69S makes ARMS2 a plausible candidate in the locus, many questions have yet to be addressed, making it difficult to determine whether ARMS2 is a responsible gene. For example, (1) What is the sequence of endogenous ARMS2 proteins in the retina; (2) What is the relationship between ARMS2 expression and the pathological changes in AMD; (3) Does A69S variant change the function of ARMS2 and contribute to the susceptibility of AMD. A more accurate annotation of ARMS2 and its expression in the retina could be obtained by systemic genomic and proteomic approaches. Although the lack of homologous genes in non-primate animals makes it difficult to study ARMS2, viral assistant gene delivery could help model ARMS2 in non-primates. 5. Summary Due to the almost perfect LD in the chromosome 10q26 locus, statistical analysis alone cannot delineate the responsible gene for the linkage and association with AMD. There are mainly two possible candidate genes for this locus, HTRA1 and ARMS2. Much more is known about the function of HTRA1 but the inconsistent results of functional analysis for the associated variants on HTRA1 expression bring into question the plausibility of HTRA1 as the functionally responsible gene at the locus. Little is known about the function of ARMS2. Substantial inconsistency exists in the functional analysis of ARMS2 including that of the indel and R38X as well as ARMS2 protein distribution in the retina. Through application of non-biased and rigorous approaches, future functional studies on the locus will strive to determine the responsible gene(s) conferring the susceptibility to AMD. Acknowledgments I thank Jonathan Haines, William Scott and Margaret PericakVance for feedback on early versions of this review. I also thank Kevin Dickson, Emily Minor and Lili Tewes for assistance. This study was supported by a Bright Focus Foundation grant (M-2012048) and a James & Esther King Biomedical Research award (3KN08). References Ambati, J., Fowler, B.J., 2012. Mechanisms of age-related macular degeneration. Neuron 75, 26e39. Andreoli, M.T., et al., 2009. Comprehensive analysis of complement factor H and LOC387715/ARMS2/HTRA1 variants with respect to phenotype in advanced age-related macular degeneration. Am. J. Ophthalmol. 148, 869e874. Ashurst, J.L., Collins, J.E., 2003. Gene annotation: prediction and testing. Annu. Rev. Genomics Hum. Genet. 4, 69e88. Baldi, A., et al., 2002. The HtrA1 serine protease is down-regulated during human melanoma progression and represses growth of metastatic melanoma cells. Oncogene 21, 6684e6688. Bergeron-Sawitzke, J., et al., 2009. Multilocus analysis of age-related macular degeneration. Eur. J. Hum. Genet. 17, 1190e1199. Bolognani, F., Perrone-Bizzozero, N.I., 2008. RNA-protein interactions and control of mRNA stability in neurons. J. Neurosci. Res. 86, 481e489. Chakravarthy, U., et al., 2013. ARMS2 increases the risk of early and late age-related macular degeneration in the European eye study. Ophthalmology 120, 342e 348. Chamary, J.V., et al., 2006. Hearing silence: non-neutral evolution at synonymous sites in mammals. Nat. Rev. Genet. 2, 98e108. Chan, C.C., et al., 2007. Human HtrA1 in the archived eyes with age-related macular degeneration. Trans. Am. Ophthalmol. Soc. 105, 92e97. Discussion 97e8. Chang, Y.F., et al., 2007. The nonsense-mediated decay RNA surveillance pathway. Annu. Rev. Biochem. 76, 51e74. Cheng, Y., et al., 2013. Genetic and functional dissection of ARMS2 in age-related macular degeneration and polypoidal choroidal vasculopathy. PLoS One 8, e53665. Chien, J., et al., 2004. A candidate tumor suppressor HtrA1 is downregulated in ovarian cancer. Oncogene 23, 1636e1644.

Chien, J., et al., 2006. Serine protease HtrA1 modulates chemotherapy-induced cytotoxicity. J. Clin. Investig. 116, 1994e2004. Chowers, I., et al., 2008. Sequence variants in HTRA1 and LOC387715/ARMS2 and phenotype and response to photodynamic therapy in neovascular age-related macular degeneration in populations from Israel. Mol. Vis. 14, 2263e2271. Conley, Y.P., et al., 2006. CFH, ELOVL4, PLEKHA1 and LOC387715 genes and susceptibility to age-related maculopathy: AREDS and CHS cohorts and meta-analyses. Hum. Mol. Genet. 15, 3206e3218. Deangelis, M.M., et al., 2008. Alleles in the HtrA serine peptidase 1 gene alter the risk of neovascular age-related macular degeneration. Ophthalmology 115, 1209e1215. Dewan, A., et al., 2006. HTRA1 promoter polymorphism in wet age-related macular degeneration. Science 314, 989e992. Ding, X., et al., 2009. Molecular pathology of age-related macular degeneration. Prog. Retin. Eye Res. 28, 1e18. Edwards, A.O., et al., 2005. Complement factor H polymorphism and age-related macular degeneration. Science 308, 421e424. Fisher, S.A., et al., 2005. Meta-analysis of genome scans of age-related macular degeneration. Hum. Mol. Genet. 14, 2257e2264. Francis, P.J., et al., 2007. The LOC387715 gene, smoking, body mass index, environmental associations with advanced age-related macular degeneration. Hum. Hered. 63, 212e218. Francis, P.J., et al., 2008. Rhesus monkeys and humans share common susceptibility genes for age-related macular disease. Hum. Mol. Genet. 17, 2673e2680. Friedrich, U., et al., 2011. Risk- and non-risk-associated variants at the 10q26 AMD locus influence ARMS2 mRNA expression but exclude pathogenic effects due to protein deficiency. Hum. Mol. Genet. 20, 1387e1399. Fritsche, L.G., et al., 2008. Age-related macular degeneration is associated with an unstable ARMS2 (LOC387715) mRNA. Nat. Genet. 40, 892e896. Fritsche, L.G., et al., 2013. Seven new loci associated with age-related macular degeneration. Nat. Genet. 45, 433e439. Fuse, N., et al., 2011. Polymorphisms in ARMS2 (LOC387715) and LOXL1 genes in the Japanese with age-related macular degeneration. Am. J. Ophthalmol. 151, 550e 556. Gorin, M.B., 2012. Genetic insights into age-related macular degeneration: controversies addressing risk, causality, and therapeutics. Mol. Asp. Med. 33, 467e486. Hadley, D., et al., 2010. Analysis of six genetic risk factors highly associated with AMD in the region surrounding ARMS2 and HTRA1 on chromosome 10, region q26. Investig. Ophthalmol. Vis. Sci. 51, 2191e2196. Hageman, G.S., et al., 2005. A common haplotype in the complement regulatory gene factor H (HF1/CFH) predisposes individuals to age-related macular degeneration. Proc. Natl. Acad. Sci. U. S. A. 102, 7227e7232. Haines, J.L., et al., 2005. Complement factor H variant increases the risk of agerelated macular degeneration. Science 308, 419e421. Hara, K., et al., 2009. Association of HTRA1 mutations and familial ischemic cerebral small-vessel disease. N. Engl. J. Med. 360, 1729e1739. Harrington, J.J., et al., 2001. Creation of genome-wide protein expression libraries using random activation of gene expression. Nat. Biotechnol. 19, 440e445. Hayashi, H., et al., 2010. CFH and ARMS2 variations in age-related macular degeneration, polypoidal choroidal vasculopathy, and retinal angiomatous proliferation. Investig. Ophthalmol. Vis. Sci. 51, 5914e5919. Hornbeck, P.V., et al., 2012. PhosphoSitePlus: a comprehensive resource for investigating the structure and function of experimentally determined post-translational modifications in man and mouse. Nucl. Acids Res. 40, D261eD270. Jacobo, S.M., et al., 2013. Age-related macular degeneration-associated silent polymorphisms in HtrA1 impair its ability to antagonize insulin-like growth factor 1. Mol. Cell Biol. 33, 1976e1990. Jakobsdottir, J., et al., 2005. Susceptibility genes for age-related maculopathy on chromosome 10q26. Am. J. Hum. Genet. 77, 389e407. Jones, A., et al., 2011. Increased expression of multifunctional serine protease, HTRA1, in retinal pigment epithelium induces polypoidal choroidal vasculopathy in mice. Proc. Natl. Acad. Sci. U. S. A. 35, 14578e14583. Kanda, A., et al., 2007. A variant of mitochondrial protein LOC387715/ARMS2, not HTRA1, is strongly associated with age-related macular degeneration. Proc. Natl. Acad. Sci. U. S. A. 104, 16227e16232. Kanda, A., et al., 2010. Age-related macular degeneration-associated variants at chromosome 10q26 do not significantly alter ARMS2 and HTRA1 transcript levels in the human retina. Mol. Vis. 16, 1317e1323. Klein, R.J., et al., 2005. Complement factor H polymorphism in age-related macular degeneration. Science 308, 385e389. Kortvely, E., et al., 2010. ARMS2 is a constituent of the extracellular matrix providing a link between familial and sporadic age-related macular degenerations. Investig. Ophthalmol. Vis. Sci. 51, 79e88. Lagnado, C.A., et al., 1994. AUUUA is not sufficient to promote poly(A) shortening and degradation of an mRNA: the functional sequence within AU-rich elements may be UUAUUUA(U/A)(U/A). Mol. Cell Biol. 14, 7984e7995. Launay, S., et al., 2008. HtrA1-dependent proteolysis of TGF-beta controls both neuronal maturation and developmental survival. Cell Death Differ. 15, 1408e 1416. Lavner, Y., Kotlar, D., 2005. Codon bias as a factor in regulating expression via translation rate in the human genome. Gene 1, 127e138. Liu, M.M., et al., 2012. Genetic mechanisms and age-related macular degeneration: common variants, rare variants, copy number variations, epigenetics, and mitochondrial genetics. Hum. Genomics 6, 13.

G. Wang / Experimental Eye Research 119 (2014) 1e7 Minor, E.A., et al., 2013. AMD-associated variants at the chromosome 10q26 locus and the stability ofARMS2 transcripts. Investig. Ophthalmol. Vis. Sci. 54 (8), 5913e5919. Mori, K., et al., 2007. Association of the HTRA1 gene variant with age-related macular degeneration in the Japanese population. J. Hum. Genet. 52, 636e641. Mount, S.M., 2000. Genomic sequence, splicing, and gene annotation. Am. J. Hum. Genet. 67, 788e792. Newman, A.M., et al., 2012. Systems-level analysis of age-related macular degeneration reveals global biomarkers and phenotype-specific functional networks. Genome Med. 2, 16. Oka, C., et al., 2004. HtrA1 serine protease inhibits signaling mediated by Tgfbeta family proteins. Development 131, 1041e1053. Pahl, L., et al., 2012. Characterization of the 10q26-orthologue in rhesus monkeys corroborates a functional connection between ARMS2 and HTRA1. Exp. Eye Res. 98, 75e78. Richardson, A.J., et al., 2010. An intergenic region between the tagSNP rs3793917 and rs11200638 in the HTRA1 gene indicates association with age-related macular degeneration. Investig. Ophthalmol. Vis. Sci. 51, 4932e4936. Rivera, A., et al., 2005. Hypothetical LOC387715 is a second major susceptibility gene for age-related macular degeneration, contributing independently of complement factor H to disease risk. Hum. Mol. Genet. 14, 3227e3236. Ross, R.J., et al., 2007. The LOC387715 polymorphism and age-related macular degeneration: replication in three case-control samples. Investig. Ophthalmol. Vis. Sci. 48, 1128e1132. Sauna, Z.E., Kimchi-Sarfaty, C., 2011. Understanding the contribution of synonymous mutations to human disease. Nat. Rev. Genet. 10, 683e691. Schmidt, S., et al., 2006. Cigarette smoking strongly modifies the association of LOC387715 and age-related macular degeneration. Am. J. Hum. Genet. 78, 852e 864. Seddon, J.M., et al., 2007. Association of CFH Y402H and LOC387715 A69S with progression of age-related macular degeneration. JAMA 297, 1793e1800. Shastry, B.S., et al., 2006. Further support for the common variants in complement factor H (Y402H) and LOC387715 (A69S) genes as major risk factors for the exudative age-related macular degeneration. Ophthalmologica 220, 291e295. Shiga, A., et al., 2011. Cerebral small-vessel disease protein HTRA1 controls the amount of TGF-b1 via cleavage of proTGF-b1. Hum. Mol. Genet. 20, 1800e1810. Sundaresan, P., et al., 2012. Polymorphisms in ARMS2/HTRA1 and complement genes and age-related macular degeneration in India: findings from the INDEYE study. Investig. Ophthalmol. Vis. Sci. 53, 7492e7497. Szafranski, K., et al., 2007. Violating the splicing rules: TG dinucleotides function as alternative 30 splice sites in U2-dependent introns. Genome Biol. 8, R154.

7

Tam, P.O., et al., 2008. HTRA1 variants in exudative age-related macular degeneration and interactions with smoking and CFH. Investig. Ophthalmol. Vis. Sci. 49, 2357e2365. Tian, J., et al., 2012. Association of genetic polymorphisms and age-related macular degeneration in Chinese population. Investig. Ophthalmol. Vis. Sci. 53, 4262e 4269. Teper, S.J., et al., 2012. A69S and R38X ARMS2 and Y402H CFH gene polymorphisms as risk factors for neovascular age-related macular degeneration in Poland e a brief report. Med. Sci. Monit. 18, PR1ePR3. Tuo, J., et al., 2008. The HtrA1 promoter polymorphism, smoking, and age-related macular degeneration in multiple case-control samples. Ophthalmology 115, 1891e1898. Vierkotten, S., et al., 2011. Overexpression of HTRA1 leads to ultra structural changes in the elastic layer of Bruch’s membrane via cleavage of extracellular matrix components. PLoS One 6, e22959. Wang, G., et al., 2009. Localization of age-related macular degeneration-associated ARMS2 in cytosol, not mitochondria. Investig. Ophthalmol. Vis. Sci. 50, 3084e3090. Wang, G., et al., 2010a. Analysis of the indel at the ARMS2 30 UTR in age-related macular degeneration. Hum. Genet. 127, 595e602. Wang, G., et al., 2010b. Genotype at polymorphism rs11200638 is not correlated with HTRA1 expression level. Arch. Ophthalmol. 128, 1491e1493. Wang, G., et al., 2013a. Variants at chromosome 10q26 locus and the expression of HTRA1 in the retina. Exp. Eye Res. 112, 102e105. Wang, G., et al., 2013b. Coding variants in ARMS2 and the risk of age-related macular degeneration. JAMA Ophthalmol. 131, 804e805. Weger, M., et al., 2007. Association of the HTRA1 -625G>A promoter gene polymorphism with exudative age-related macular degeneration in a central European population. Mol. Vis. 13, 1274e1279. Xu, Y.T., et al., 2012. Age-related maculopathy susceptibility 2 participates in the phagocytosis functions of the retinal pigment epithelium. Int. J. Ophthalmol. 5, 125e132. Yang, Z., et al., 2006. A variant of the HTRA1 gene increases susceptibility to agerelated macular degeneration. Science 314, 992e993. Yang, Z., et al., 2010. Genetic and functional dissection of HTRA1 and LOC387715 in age-related macular degeneration. PLoS Genet. 6, e1000836. Zhang, L., et al., 2012. High temperature requirement factor A1 (HTRA1) gene regulates angiogenesis through transforming growth factor-b family member growth differentiation factor 6. J. Biol. Chem. 287, 1520e1526. Zumbrunn, J., Trueb, B., 1996. Primary structure of a putative serine protease specific for IGF-binding proteins. FEBS Lett. 398, 187e192.