Genetics

Core Messages

›Various components contribute to the development of AMD with up to 70% of AMD susceptibility likely to be ascribed to genetic factors.

›To date, several AMD-associated genes have been identified. Among these are two major loci, namely complement factor H (CFH) at 1q31 and ARMS2/HTRA1 at 10q26, together estimated to explain more than 50% of AMD cases.

›The discovery of CFH as a major risk factor, in concert with subsequently identified modifiers CFB, C3, and CFI, strongly suggests a major role of the alternative complement pathway in AMD pathogenesis.

›Recent advances in high-throughput technologies will help to discover additional AMDrelatedgenes,butputativelywithminorfrequency and effect sizes. This will further enhance our insight into disease-associated pathways and mechanisms underlying AMD pathology.

›Tailored to the genetic risk profile of individual patients, preventive measures or therapeutic treatments are within reach.

L.G. Fritsche • U. Friedrich • B.H.F. Weber (*) Institute of Human Genetics, University of Regensburg, Regensburg, Germany

e-mail: lars.fritsche@klinik.uni-regensburg.de; ulrike.friedrich@klinik.uni-regensburg.de; bweb@klinik.uni-regensburg.de

2.1Introduction

In 1885, age-related macular degeneration (AMD) was first described as a clinical entity, characterized by agedependent pigmentary and atrophic changes in the central area of the retina [1]. Today, AMD is known as one of the most frequent retinal disorders in industrialized countries and is a major cause of legal blindness [2]. Over the last decades, extraordinary efforts have been undertaken to elucidate the pathogenic mechanisms underlying the disease. A great number of clinical, pathological and experimental observations have been made, culminating in a first basic concept of disease pathology detailing the probable sequence of events in the process from a normally ageing retina to the diseased state [3]. Accordingly, AMD was recognized as a disorder associated with age and other environmental factors leading to age-related changes in the homeostatic equilibrium of physiological processes of the retinal pigment epithelium (RPE) and the retina. Following these early injuries, it has been suggested that the retina is susceptible to increased oxidative stress, leading to injuries of the RPE and possibly the choriocapillaries. Such injuries could then trigger chronic inflammatory responses leading to the formation of abnormal extracellular matrix (ECM). As a consequence, this damage could cause an altered diffusion of nutrients across Bruch’s membrane, a five-layered ECM structure between the RPE and the retina. This may ultimately lead to atrophy of photoreceptors, RPE and choriocapillaries, or to choroidal new vessel growth, characterizing the two late stages of AMD.

In the largest twin study available to date, genetic predisposition to AMD was firmly established, estimating heritability between 46% and 71% [4].

Initial studies to identify the underlying genetic factors date back to the late 1990s. However, this early work revealed only weak or rare risk variants [5–7]. It was not until 2005 that the first major AMD susceptibility gene, the complement factor H gene (CFH), was identified with effect sizes and frequencies of corresponding risk variants that were unprecedented in the field of complex diseases [8–11]. Moreover, CFH could be directly related to a biological pathway that had been suggested over 20 years before to play a role in AMD pathology, namely a chronic inflammatory component in the disease process [12]. Since then, a number of additional genes have been reliably associated with AMD and have revolutionized our understanding of disease etiology. The following chapter will summarize the current progress in AMD genetics and touch upon emerging perspectives and future challenges in AMD patient management.

2.2Identifying Risk Factors of a Common Disease

Since 2003, three parallel and synergistic developments have spurred the identification of genetic risk factors:

•The completion of the human genome sequence,

•The establishment and public availability of databases of common human variations, and

•The methodological progress in sequencing and genotyping.

The information on common single nucleotide

polymorphisms (SNPs) has now become directly and easily accessible for genotyping large numbers of samples without the need for prior characterization of these variants in timeand cost-intensive experiments.

Nevertheless, approaches to define AMD-associated genes have still to overcome a number of obstacles. For example, as expected for diseases with a genetically complex etiology, an obvious inheritance pattern is likely not evident in a given pedigree. In contrast to Mendelian diseases where generally a single gene variant exerts a strong effect, genetic risk factors in complex diseases commonly have weak effects that often act synergistically with other genetic or environmental factors in disease development. This may be further complicated by a situation where a single risk factor might not be sufficient or even required in the disease process.

Approaches to define AMD-associated genes mainly resort to case-control studies entailing the assessment of a disproportionate enrichment of risk variants in a group of patients versus healthy controls. Two synergistic working hypotheses are applied, known as the “common disease–common variant (CDCV) hypothesis” and “common disease–rare variant (CDRV) hypothesis.” In the former, common variants are thought to be a likely cause of a common disease. Consequently, the search focuses on the common variations in the human genome associated with AMD. In contrast, the CDRV hypothesis deals with rare genetic variants that by themselves have strong effects but are unlikely to have a major impact on the overall disease load as they affect only few individuals in the diseased patient population. However, as a substantial number of such variants might exist in one or several genes, their sum might collectively propel these genes to common susceptibility loci [13].

As common SNPs are more favorable for analysis as compared to rare “individual” variants, the CDCV hypothesis has received broader acceptance in the scientific community and thus has become the main focus of current research in complex diseases. The common risk variants are not only instrumental in dissecting the role of candidate genes, but are also successfully exploited in undirected approaches such as genome-wide association studies (GWAS). In the candidate gene studies interesting genes are selected based on a priori information, which might include:

•Functional properties of the gene products that possibly fit the observed pathology,

•A specific expression profile, or

•A known role in overlapping pathologies or animal models connected with the disease.

The candidate gene approach typically analyzes a

restricted number of SNPs within a genetic candidate locus in a large case control sample, and can provide enough statistical power to detect even uncommon and weak risk effects. In contrast, GWAS only became feasible with newly developed high throughput genotyping technologies allowing the analysis of thousands to millions of variants in a single experiment. A main advantage of such an approach is the neutrality in genetic marker selection and thus the possibility to identify associated genes that are not obviously suggested as candidate genes by their specific properties. However, the magnitude of analyzed markers also increases the probability of false-positive association signals by chance. Thus, corrections for multiple