CFH, ARMS2, HTRA1, and Complement Dysregulation in Macular Degeneration

CFH, ARMS2, and HTRA1 represent the two major genetic susceptibility loci that together explain more than 60% of the population-attributable heritable risk of age-related macular degeneration (AMD), the leading cause of irreversible central blindness in older adults across industrialized populations. In 2005, one of the first landmark genome-wide association studies (GWAS) in human genetics identified a common coding polymorphism (rs1061170, p.Tyr402His or Y402H) in the complement factor H (CFH) gene within the Regulator of Complement Activation (RCA) cluster on chromosome 1q31.3, shortly followed by the discovery of a second high-penetrance linkage disequilibrium (LD) block spanning ARMS2 (age-related maculopathy susceptibility 2) and the promoter of HTRA1 (high-temperature requirement A serine peptidase 1) on chromosome 10q26.13.

Pathobiologically, AMD originates at the retinal pigment epithelium (RPE)–Bruch's membrane–choriocapillaris interface of the central macula, where lipoprotein- and complement-rich extracellular deposits termed drusen accumulate between the RPE basal lamina and the inner collagenous layer of Bruch's membrane. When CFH Y402H impairs host polyanion recognition on Bruch's membrane and ARMS2–HTRA1 variants accelerate elastic lamina proteolysis, chronic alternative complement pathway (C3bBb) overactivation and membrane attack complex (MAC, C5b-9) deposition trigger RPE geographic atrophy (dry/advanced atrophic AMD) or vascular endothelial growth factor (VEGF)-driven choroidal neovascularization (CNV, wet/exudative AMD).

Architecture of the 20 Sushi Domains (SCR1–SCR20) and C3b Inactivation

Encoded by the CFH gene on chromosome 1q31.3 (23 exons, 1,231 amino acids, ~155 kDa plasma glycoprotein circulating at ~200–500 μg/mL), Complement Factor H (CFH) is the primary fluid-phase and surface-bound negative regulator of the alternative complement pathway. Unlike the classical and lectin pathways (which require antibodies or microbial mannose motifs), the alternative pathway undergoes continuous spontaneous "tick-over" hydrolysis of the internal thioester bond of plasma C3 at a basal rate of ~0.5% to 1.0% per hour, generating C3(H₂O) and nascent C3b that covalently attaches to hydroxyl and amine groups on both microbial and host cell surfaces. If unchecked, surface-bound C3b binds Factor B (CFB), which is cleaved by Factor D (CFD) into the active C3bBb alternative pathway C3 convertase—an enzymatic positive-feedback loop that cleaves thousands of additional C3 molecules within minutes.

To protect acellular host extracellular matrices such as Bruch's membrane and the renal glomerular basement membrane (which lack membrane-bound regulators like CD46, CD55, and CD59), CFH folds into a flexible string of 20 homologous Short Consensus Repeat (SCR) domains (also called complement control protein, CCP, or sushi domains, each ~60 amino acids stabilized by two invariant disulfide bonds):

1. N-Terminal Regulatory Core (SCR1–SCR4): Binds directly to C3b, accelerating the irreversible dissociation (decay-accelerating activity) of the catalytic Bb fragment from the C3bBb convertase and serving as the mandatory cofactor for the serine protease Complement Factor I (CFI) to cleave C3b into inactive iC3b.
2. Internal Surface-Recognition Hub (SCR6–SCR8, Containing Tyr402 in SCR7): Mediates tissue-specific anchoring to sulfated glycosaminoglycans (heparan sulfate, dermatan sulfate) in Bruch's membrane, C-reactive protein (CRP), and lipid peroxidation adducts (malondialdehyde, MDA) generated by photoreceptor outer-segment photo-oxidation.
3. C-Terminal Host-Cell Discriminator (SCR19–SCR20): Binds simultaneously to C3d/thioester domains and host sialic acid (Neu5Ac) and heparan sulfate chains on endothelial surfaces (where missense mutations cause atypical hemolytic uremic syndrome, aHUS).

How the Tyr402His (Y402H) Variant Unleashes Subretinal Inflammation

The rs1061170 (c.1204T>C, p.Tyr402His or Y402H) variant substitutes a positively charged histidine (His402) for a neutral aromatic tyrosine (Tyr402) within a surface-exposed cationic loop of SCR7 (which is shared by both full-length CFH and the truncated 7-SCR alternative splice isoform FHL-1, the predominant isoform capable of diffusing passively across the porous collagenous meshwork of Bruch's membrane). Structural crystallography and surface plasmon resonance (SPR) demonstrate that the His402 substitution alters the electrostatic spacing of SCR7, reducing the binding affinity of CFH and FHL-1 for Bruch's membrane heparan sulfate proteoglycans by ~50% to 70% and abolishing high-affinity neutralization of pro-inflammatory malondialdehyde (MDA) lipid peroxidation epitopes by >75%. Consequently, in Y402H carriers (OR ≈ 2.5–2.8 for heterozygotes; OR ≈ 5.5–7.4 for HH homozygotes), unanchored Bruch's membrane becomes coated with active C3bBb convertases, releasing anaphylatoxins C3a and C5a (which recruit subretinal microglia/macrophages) and assembling cytolytic C5b-9 membrane attack complexes that lyse choriocapillaris endothelial cells and overlying RPE monolayers. Surrounding CFH within the Chr1q31.3 RCA locus, a common 84-kb non-homologous allelic recombination deletion removing CFHR3 and CFHR1 (ΔCFHR3–CFHR1) confers strong protection against AMD (OR ≈ 0.35–0.48), because FHR-1 and FHR-3 share SCR19–20 homology with CFH and competitively displace CFH from C3b without possessing SCR1–4 cofactor activity (de-regulation).

HTRA1 Serine Peptidase Overexpression and Bruch's Membrane Elastolysis

The second major AMD susceptibility locus on chromosome 10q26.13 comprises a tight ~15-kb block of near-complete linkage disequilibrium (r² > 0.95) containing the coding missense variant rs10490924 (p.Ala69Ser or A69S) and an indel (del443ins54) in the primate-specific gene ARMS2, paired with the promoter regulatory variant rs11200638 (−625G>A) upstream of HTRA1. Whereas ARMS2 encodes a small 107-amino-acid polypeptide localized to the outer mitochondrial membrane and subretinal microglia (where A69S and the 3' UTR indel destabilize ARMS2 mRNA and impair non-canonical surface opsonization of apoptotic debris), functional chromatin conformation capture and human donor RPE transcriptomics demonstrate that the risk haplotype disrupts a cis-regulatory silencer/enhancer element controlling HTRA1, driving a 2- to 3-fold upregulation of secreted HTRA1 serine protease at the RPE–Bruch's membrane boundary.

HTRA1 is a 480-amino-acid homotrimeric, ATP-independent trypsin-like serine peptidase (His220–Asp250–Ser328 catalytic triad) equipped with an N-terminal insulin-like growth factor-binding domain (IGFBP) and a C-terminal PDZ substrate-gating domain. In the subretinal space, excess HTRA1 proteolytically degrades critical structural components of Bruch's membrane, including fibulin-3 (EFEMP1), fibulin-5 (FBLN5), fibronectin, decorin, vitronectin, and the endogenous anti-angiogenic glycoprotein thrombospondin-1 (THBS1). Furthermore, HTRA1 cleaves transforming growth factor-β (TGF-β) family precursors and alpha-1-antitrypsin (SERPINA1). This enzymatic digestion fragments the central elastic lamina of Bruch's membrane, lowering the mechanical barrier to invading choriocapillaris sprouts and explaining why the ARMS2–HTRA1 risk haplotype confers a particularly steep risk for neovascular (wet) AMD and polypoidal choroidal vasculopathy (PCV) (OR ≈ 7.5–10.0 in homozygotes). Across individuals homozygous for both CFH Y402H and ARMS2–HTRA1 risk alleles, the combined odds ratio for late-stage AMD reaches ~40- to 50-fold.

The Paradoxical APOE4 Protective Inversion in Subretinal Drusen

A striking mechanistic discovery of AMD genetics is the complete directional inversion of apolipoprotein E (APOE) alleles compared to Alzheimer's disease and cardiovascular atherosclerosis. In Alzheimer's disease, the APOE ε4 allele (Cys112Arg) increases risk 3- to 15-fold while APOE ε2 (Arg158Cys) is protective. In AMD, the opposite holds: APOE ε4 is significantly protective (OR ≈ 0.65–0.72 per allele), whereas APOE ε2 increases AMD risk (OR ≈ 1.25–1.40). Biophysical transport models of Bruch's membrane explain this paradox: the RPE continuously secretes large ApoB/ApoE-containing cholesteryl-ester lipoprotein particles basolaterally toward the choriocapillaris. Because ApoE4 carries a net +2 positive charge relative to ApoE2 and forms smaller, more compact lipid complexes, ApoE4 particles traverse the negatively charged heparan-sulfate proteoglycan sieve of aged Bruch's membrane with lower hydraulic resistance, whereas larger ApoE2-rich lipoproteins become entrapped, coalescing into soft drusen.

Pegcetacoplan (C3) and Avacincaptad Pegol (C5) Intravitreal Therapy

While wet neovascular AMD is effectively managed by intravitreal anti-VEGF biologics (ranibizumab, aflibercept, faricimab), dry geographic atrophy (GA)—affecting over 5 million individuals worldwide—lacked any disease-modifying therapy until the clinical validation of direct intravitreal complement inhibition in 2023. Two FDA-approved therapeutics directly target the complement cascade identified by CFH and C3 genetics:

1. Pegcetacoplan (Syfovre): A pegylated bicyclic 15-amino-acid compstatin derivative (43.5 kDa) that binds directly to C3 and C3b, sterically blocking both C3 cleavage by C3 convertases and the assembly of new C3bBb amplification loops. In the Phase 3 OAKS and DERBY trials, monthly or every-other-month intravitreal pegcetacoplan slowed the rate of geographic atrophy lesion growth by 18% to 22% over 24 months (with up to ~36% slowing in extrafoveal lesions during months 18–24).
2. Avacincaptad pegol (Izervay): A pegylated 39-nucleotide chemically modified RNA aptamer (~56 kDa) that binds and inhibits Complement C5, preventing its cleavage into pro-inflammatory C5a and the membrane-lytic C5b (C5b-9 MAC) subunit while preserving upstream C3b/iC3b opsonization (GATHER1 and GATHER2 Phase 3 trials: 14% to 27% reduction in GA lesion growth rate at 12 months).

Why does the CFH Y402H variant specifically damage Bruch's membrane in the eye while leaving most other organs unaffected?

Most vascularized human tissues protect their cell membranes against alternative complement C3b tick-over using membrane-anchored regulators (CD46, CD55, and CD59). However, Bruch's membrane is an acellular extracellular matrix that relies exclusively on soluble CFH and its truncated splice isoform FHL-1 anchoring to sulfated heparan sulfate chains via domain SCR7. The Y402H substitution specifically weakens SCR7 binding to ocular heparan sulfate and photo-oxidized malondialdehyde (MDA) adducts without disrupting SCR19–20 endothelial sialic acid binding.

How does the Chr10q26 ARMS2–HTRA1 risk allele promote choroidal neovascularization (wet AMD)?

The Chr10q26.13 risk haplotype (rs10490924 / rs11200638) increases expression of the secreted serine protease HTRA1 by 2- to 3-fold in the retinal pigment epithelium (RPE). Excess HTRA1 cleaves structural extracellular matrix proteins (fibulin-3/EFEMP1, fibulin-5, and fibronectin) that maintain the elastic lamina of Bruch's membrane, while simultaneously degrading the endogenous anti-angiogenic protein thrombospondin-1 (THBS1), allowing choroidal blood vessels to breach Bruch's membrane and invade the subretinal space.

Why is the APOE ε4 allele protective against macular degeneration despite being the strongest genetic risk factor for Alzheimer's disease?

In the retina, RPE cells clear phagocytosed photoreceptor lipids by secreting ApoE-coated lipoprotein particles across Bruch's membrane into the choroidal circulation. Because ApoE4 (Cys112Arg) carries two additional positive charges compared to ApoE2 (Arg158Cys) and forms smaller, higher-curvature particles, it passes more readily through the negatively charged proteoglycan matrix and experiences higher receptor-mediated clearance, preventing the sub-RPE lipid accumulation that nucleates soft drusen (OR ≈ 0.68).

1. Klein, R. J., Zeiss, C., Chew, E. Y., et al. (2005). "Complement factor H polymorphism in age-related macular degeneration." Science, 308(5720), 385–389. https://doi.org/10.1126/science.1109557
2. Weismann, D., Hartvigsen, K., Lauer, N., et al. (2011). "Complement factor H binds malondialdehyde epitopes and protects from oxidative stress." Nature, 478(7367), 76–81. https://doi.org/10.1038/nature10449
3. Fritsche, L. G., Igl, W., Bailey, J. N. C., et al. (2016). "A large genome-wide association study of age-related macular degeneration highlights contributions of rare and common variants." Nature Genetics, 48(2), 134–143. https://doi.org/10.1038/ng.3448
4. Heier, J. S., Lad, E. M., Holz, F. G., et al. (2023). "Pegcetacoplan for the treatment of geographic atrophy secondary to age-related macular degeneration (OAKS and DERBY): two multicentre, randomised, double-masked, sham-controlled, phase 3 trials." The Lancet, 402(10411), 1434–1448. https://doi.org/10.1016/S0140-6736(23)01520-901520-9)