Abstract
Primary angle-closure glaucoma (PACG) is a common form of glaucoma in the Far East. Its defining feature is iridocorneal angle closure. In addition to PACG, indications of angle closure are included in the diagnostic criteria of related conditions primary angle-closure suspect (PACS) and primary angle closure (PAC). To the best of our knowledge, a causative gene for iridocorneal angle closure in humans has not been identified. This study aimed to identify the genetic cause of iridocorneal angle closure in a pedigree with at least 10 individuals diagnosed with PACS, PAC or PACG. Results of linkage analysis, segregation analysis of 44 novel variations, whole exome sequencing of 10 individuals, screenings of controls and bioinformatics predictions identified a mutation in COL18A1 that encodes collagen type XVIII as the most likely cause of angle closure in the pedigree. The role of COL18A1 in the etiology of Knobloch syndrome (KS) that is consistently accompanied by optic anomalies, available functional data on the encoded protein and the recognized role of collagens and the extracellular matrix in glaucoma pathogenesis supported the proposed role of the COL18A1 mutation in the pedigree. Subsequent identification of other COL18A1 mutations in PACS affected individuals of two unrelated families further supported that COL18A1 may affect angle closure. These PACS individuals were parents and grandparents of KS-affected children. In conclusion, a gene that affects angle closure in humans, a critical feature of PACG, has been identified. The findings also reinforce the importance of collagens in eye features and functions.
Introduction
Glaucoma is a heterogeneous group of optic neuropathies characterized by a specific pattern of optic nerve degeneration and visual field loss that is usually accompanied by increased intraocular pressure (IOP) (1). It is the major cause of irreversible blindness worldwide (2). Primary glaucoma is classified into three major forms based on the anatomy of the anterior chamber drainage angle of the eye and age of onset: primary congenital glaucoma (PCG), primary open-angle glaucoma (POAG) and primary angle-closure glaucoma (PACG) (1). In glaucoma patients with increased IOP, the increase is thought to be due mainly to impaired drainage of aqueous humor from the anterior chamber (3). In PCG, the most severe and the least-common form of glaucoma, an anatomical defect (trabeculodysgenesis) in the trabecular meshwork (TM) is thought to affect poor drainage (4). As the angle between the iris root and the posterior TM remains open in POAG patients, increased IOP in these individuals is thought to be due to TM anomalies that hinder the flow of aqueous humor. In contrast, eyes with PACG reveal a closed-anterior chamber angle due to contact between the peripheral iris and the posterior TM. This closure obstructs outflow of the aqueous humor. With regards to genetics, PCG is usually a monogenetic Mendelian disease for which three causative genes, cytochrome p4501B1 (CYP1B1) (OMIM 601771), latent transforming growth factor beta binding protein 2 (LTBP2) (OMIM 602091) and TEK receptor tyrosine kinase (TEK) (OMIM 600221), have been identified (5–8). POAG and PACG are generally considered complex multifactorial disorders (9). Several POAG-causing genes, including myocilin (MYOC) (OMIM 601652), optineurin (OPTN) (OMIM 602432), WD repeat containing protein 36 (WDR36) (OMIM 609669), neurotrophin-4 (NTF4) (OMIM 162662) and TANK-binding kinase 1 (TBK1) (OMIM 604834) have been identified (10–14), but mutations in these account for disease in <10% of patients. Contrary to PCG and POAG, a definitive PACG causing gene has not been identified in humans. However, high prevalence of this condition in Asians, observations on familial clustering and results of sib-pair studies suggest that genetic factors are involved in the etiology of PACG (15–18). Association studies have implicated possible roles for several genes including collagen type XI alpha 1 (COL11A1) (OMIM 120280), pleckstrin homology domain containing, family A member 7 (PLEKHA7) (OMIM 612686), ATP binding cassette subfamily C member 5 (ABCC5) (OMIM 605251), protein-L-isoaspartate (D-aspartate) O-methyltransferase domain containing (PCMTD1)–C2H2C-type zinc finger (ST18) (OMIM 617155), ependymin related 1 (EPDR1), choline O-acetyltransferase (CHAT) (OMIM 118490), GLIS family zinc finger 3 (GLIS3) (OMIM 610192), fermitin family member 2 (FERMT2) (OMIM 607746) and dolichyl-phosphate mannosyltransferase subunit 2, regulatory (DPM2) (OMIM 603564)– family with sequence similarity 102 member A (FAM102A) (OMIM 610891) encoding genes (19–21). Results of candidate gene screenings have also suggested possible roles for some genes including genes that code for matrix metalloproteinase-9 (MMP-9) (OMIM 120361), hepatocyte growth factor (HGF) (OMIM 142409), methylene tetrahydrofolate reductase (MTHFR) (OMIM 607093), nitric oxide synthase 3 (eNOS) (OMIM 163729) and heat shock protein family A member 1A (HSP70-1A) (OMIM 140550) (22–26). However, these findings are not generally considered definitive, and putative roles of genes suggested by some studies were not confirmed in independent studies (27–29). Taken together, a PACG-causing gene or a strongly associated locus has not yet been identified (17).
Failure to find a strong PACG-associated locus likely reflects the complexity of the disease, meaning that multiple genes that affect various anatomical and functional features that associate with PACG contribute to its etiology. Factors known to be associated with PACG include hyperoptic refractive error, shallow anterior chamber, thick crystalline lens, short axial length, small corneal diameter and narrow iridocorneal angle (30–32). Clearly, some of these also associate with each other. Ultimately, the defining feature of PACG in individuals with glaucomatous optic nerve damage is angle closure. In addition to PACG, indications of angle closure are included in the diagnostic criteria of conditions known as primary angle-closure suspect (PACS) and primary angle closure (PAC) (see Methods section) (33). Angle closure evidences in these conditions in the absence of glaucomatous optic neuropathy.
The focus of this study is genetic analysis of a pedigree-labeled ANG200 in which angle closure is prevalent. The individuals were diagnosed with PACS, PAC or PACG. We report the presence of a mutation in COL18A1 in all the individuals diagnosed with these conditions. We found supporting evidence for the role of COL18A1 in pathogenesis of angle closure by observation of COL18A1 mutations in PACS-affected individuals of two additional unrelated families.
Results
Family and clinical data
The ANG200 pedigree with at least 1 members who presented with varying degrees of iridocorneal angle closure is shown in Figure 1. Their phenotypic data are presented in Table 1. Certain features of pedigree ANG200 are to be noted. First, nearly all individuals in generation III had been or were affected with cataract and/or a degree of angle closure. The high incidence of cataract in these individuals may be related to the fact that they were elderly, aged 75 years or older. Cataract was not observed in individuals of generation IV whose ages range from 41 to 69 years. The angle-closure status of many individuals in the pedigree was unknown to them until eye examinations were performed after they were recruited into the study. Age of onset of the condition is unknown in all affected individuals. Rather unexpectedly, many individuals in generations III and IV, including ANG200 IV-5–IV-9, have remained unwed.
ANG200 pedigree. Black-filled shapes: individuals with various degrees of ridocorneal angle closure and diagnosed with PACS, PAC or PACG (see Table 1); grey-filled shape: glaucoma diagnosis based on presence of glaucomatous optic nerve damage, but angle closure not assessable due to prior intracapsular cataract extraction eyes resulting from surgery; hatched shapes: reported by family members to have been blind in one or both eyes at time of death; open shapes of generations I–IV: reported by family members not to have had ocular problems; open shapes of generation V: normal eye exam. Arrow, proband; *, individuals whose whole genome SNP data were included in the linkage analysis, ° individuals for whom exome sequencing was performed. Clinical diagnoses and age at time of examination (years) are indicated; NEE, normal eye examination. Genotypes of individuals ascertained by direct sequencing are presented: N, normal COL18A1 allele; M, mutated COL18A1 allele.
Phenotypic and genotypic data on pedigree ANG200 individuals with mutation in COL18A1
| Individual ID | III-3 | III-4 | III-5 | III-6 | IV-2 | IV-3 | IV-5 | IV-6 | IV-7 | IV-8 | IV-9 | V-1 | V-2 | V-3& |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| COL18A1 genotype | NM | ?M# | NM | NM | NM | NM | NM | NM | MM | NM | NM | NM | NM | NN |
| Clinical examination | ||||||||||||||
| Sex | F | M | M | F | M | F | F | M | F | M | F | F | M | F |
| Age at examination (yrs) | 82 | 77 | 75 | 69 | 66 | 56 | 49 | 53 | 48 | 41 | 37 | 35 | 40 | |
| Angle1 | ||||||||||||||
| circumference OD/OS | * | 360°/VLE | 270°/ 270° | 360°/ 360° | 360°/ 360° | 270°/ 270° | 270°/ 270° | 360°/ 360° | 360°/ 360° | 360°/ 360° | 270°/ 270° | 360°/360° | 360°/360° | 360°/360° |
| GradeOD/OS | * | 0-1/VLE | 1/ 1 | 0-1/ 0-1 | 0-1/ 0-1 | 1 /1 | 1 /1 | 0-1/ 0-1 | 0-1/ 0-1 | 0-1/ 0-1 | 1 /1 | 3/3 | 3/3 | 3/3 |
| Cornea1OD/OS | Edema/Edema | Clear/Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/Clear | Clear/Clear | Clear/Clear |
| Anterior chamber1OD/OS | Deep/Deep | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow clear | Deep/Deep | Deep/Deep | Deep/Deep |
| Lens1OD/OS | Aphakic/ Aphakic | Severe cataract/ VLE | Severe cataract/ moderate cataract | Severe cataract (OU) | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/Clear | Clear/Clear | Clear/Clear |
| Vertical C/D ratio1OD/OS | 0.80/ 0.80 | 0.40/VLE | 0.25/ 0.25 | 0.45/ 0.45 | 0.55/ 0.60 | 0.40/ 0.40 | 0.50/ 0.50 | 0.50/ 0.50 | 0.30/ 0.30 | 0.40/ 0.45 | 0.60/ 0.60 | 0.25/0.25 | 0.50/0.60 | 0.30/0.30 |
| IOP2(mmHg)OD/OS | 144/144 | 25/VLE | 26/28 | 12/12 | 18/18 | 18/19 | 16/16 | 17/17 | 29/30 | 23/23 | 14/17 | 12/12 | 11/13 | |
| RefractionOD/OS | +13.75/ no reflex | -1.25 -1.25x65/ VLE | +2.25 -0.75x135/ +3.50 -0.25x60 | Plano -2.75/ +1.50 | +1.50 -0.75x90/ +1.75-0.50x85 | +0.75 -1.00x80/ +0.25 -1.00x180 | Plano -.25/ +0.25 -0.50x50 | Plano -0.50x3/ plano -0.25x180 | +1.25 -0.50x85/ +1.25 -0.25x85 | -0.75 -0.50x62/ -0.75 -1.00x95 | Plano/ plano -0.75x170 | Plano/Plano | Plano/Plano | -3.50/ -3.50 |
| Individual ID | III-3 | III-4 | III-5 | III-6 | IV-2 | IV-3 | IV-5 | IV-6 | IV-7 | IV-8 | IV-9 | V-1 | V-2 | V-3& |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| COL18A1 genotype | NM | ?M# | NM | NM | NM | NM | NM | NM | MM | NM | NM | NM | NM | NN |
| Clinical examination | ||||||||||||||
| Sex | F | M | M | F | M | F | F | M | F | M | F | F | M | F |
| Age at examination (yrs) | 82 | 77 | 75 | 69 | 66 | 56 | 49 | 53 | 48 | 41 | 37 | 35 | 40 | |
| Angle1 | ||||||||||||||
| circumference OD/OS | * | 360°/VLE | 270°/ 270° | 360°/ 360° | 360°/ 360° | 270°/ 270° | 270°/ 270° | 360°/ 360° | 360°/ 360° | 360°/ 360° | 270°/ 270° | 360°/360° | 360°/360° | 360°/360° |
| GradeOD/OS | * | 0-1/VLE | 1/ 1 | 0-1/ 0-1 | 0-1/ 0-1 | 1 /1 | 1 /1 | 0-1/ 0-1 | 0-1/ 0-1 | 0-1/ 0-1 | 1 /1 | 3/3 | 3/3 | 3/3 |
| Cornea1OD/OS | Edema/Edema | Clear/Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/Clear | Clear/Clear | Clear/Clear |
| Anterior chamber1OD/OS | Deep/Deep | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow clear | Deep/Deep | Deep/Deep | Deep/Deep |
| Lens1OD/OS | Aphakic/ Aphakic | Severe cataract/ VLE | Severe cataract/ moderate cataract | Severe cataract (OU) | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/Clear | Clear/Clear | Clear/Clear |
| Vertical C/D ratio1OD/OS | 0.80/ 0.80 | 0.40/VLE | 0.25/ 0.25 | 0.45/ 0.45 | 0.55/ 0.60 | 0.40/ 0.40 | 0.50/ 0.50 | 0.50/ 0.50 | 0.30/ 0.30 | 0.40/ 0.45 | 0.60/ 0.60 | 0.25/0.25 | 0.50/0.60 | 0.30/0.30 |
| IOP2(mmHg)OD/OS | 144/144 | 25/VLE | 26/28 | 12/12 | 18/18 | 18/19 | 16/16 | 17/17 | 29/30 | 23/23 | 14/17 | 12/12 | 11/13 | |
| RefractionOD/OS | +13.75/ no reflex | -1.25 -1.25x65/ VLE | +2.25 -0.75x135/ +3.50 -0.25x60 | Plano -2.75/ +1.50 | +1.50 -0.75x90/ +1.75-0.50x85 | +0.75 -1.00x80/ +0.25 -1.00x180 | Plano -.25/ +0.25 -0.50x50 | Plano -0.50x3/ plano -0.25x180 | +1.25 -0.50x85/ +1.25 -0.25x85 | -0.75 -0.50x62/ -0.75 -1.00x95 | Plano/ plano -0.75x170 | Plano/Plano | Plano/Plano | -3.50/ -3.50 |
(Continued)
Phenotypic and genotypic data on pedigree ANG200 individuals with mutation in COL18A1
| Individual ID | III-3 | III-4 | III-5 | III-6 | IV-2 | IV-3 | IV-5 | IV-6 | IV-7 | IV-8 | IV-9 | V-1 | V-2 | V-3& |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| COL18A1 genotype | NM | ?M# | NM | NM | NM | NM | NM | NM | MM | NM | NM | NM | NM | NN |
| Clinical examination | ||||||||||||||
| Sex | F | M | M | F | M | F | F | M | F | M | F | F | M | F |
| Age at examination (yrs) | 82 | 77 | 75 | 69 | 66 | 56 | 49 | 53 | 48 | 41 | 37 | 35 | 40 | |
| Angle1 | ||||||||||||||
| circumference OD/OS | * | 360°/VLE | 270°/ 270° | 360°/ 360° | 360°/ 360° | 270°/ 270° | 270°/ 270° | 360°/ 360° | 360°/ 360° | 360°/ 360° | 270°/ 270° | 360°/360° | 360°/360° | 360°/360° |
| GradeOD/OS | * | 0-1/VLE | 1/ 1 | 0-1/ 0-1 | 0-1/ 0-1 | 1 /1 | 1 /1 | 0-1/ 0-1 | 0-1/ 0-1 | 0-1/ 0-1 | 1 /1 | 3/3 | 3/3 | 3/3 |
| Cornea1OD/OS | Edema/Edema | Clear/Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/Clear | Clear/Clear | Clear/Clear |
| Anterior chamber1OD/OS | Deep/Deep | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow clear | Deep/Deep | Deep/Deep | Deep/Deep |
| Lens1OD/OS | Aphakic/ Aphakic | Severe cataract/ VLE | Severe cataract/ moderate cataract | Severe cataract (OU) | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/Clear | Clear/Clear | Clear/Clear |
| Vertical C/D ratio1OD/OS | 0.80/ 0.80 | 0.40/VLE | 0.25/ 0.25 | 0.45/ 0.45 | 0.55/ 0.60 | 0.40/ 0.40 | 0.50/ 0.50 | 0.50/ 0.50 | 0.30/ 0.30 | 0.40/ 0.45 | 0.60/ 0.60 | 0.25/0.25 | 0.50/0.60 | 0.30/0.30 |
| IOP2(mmHg)OD/OS | 144/144 | 25/VLE | 26/28 | 12/12 | 18/18 | 18/19 | 16/16 | 17/17 | 29/30 | 23/23 | 14/17 | 12/12 | 11/13 | |
| RefractionOD/OS | +13.75/ no reflex | -1.25 -1.25x65/ VLE | +2.25 -0.75x135/ +3.50 -0.25x60 | Plano -2.75/ +1.50 | +1.50 -0.75x90/ +1.75-0.50x85 | +0.75 -1.00x80/ +0.25 -1.00x180 | Plano -.25/ +0.25 -0.50x50 | Plano -0.50x3/ plano -0.25x180 | +1.25 -0.50x85/ +1.25 -0.25x85 | -0.75 -0.50x62/ -0.75 -1.00x95 | Plano/ plano -0.75x170 | Plano/Plano | Plano/Plano | -3.50/ -3.50 |
| Individual ID | III-3 | III-4 | III-5 | III-6 | IV-2 | IV-3 | IV-5 | IV-6 | IV-7 | IV-8 | IV-9 | V-1 | V-2 | V-3& |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| COL18A1 genotype | NM | ?M# | NM | NM | NM | NM | NM | NM | MM | NM | NM | NM | NM | NN |
| Clinical examination | ||||||||||||||
| Sex | F | M | M | F | M | F | F | M | F | M | F | F | M | F |
| Age at examination (yrs) | 82 | 77 | 75 | 69 | 66 | 56 | 49 | 53 | 48 | 41 | 37 | 35 | 40 | |
| Angle1 | ||||||||||||||
| circumference OD/OS | * | 360°/VLE | 270°/ 270° | 360°/ 360° | 360°/ 360° | 270°/ 270° | 270°/ 270° | 360°/ 360° | 360°/ 360° | 360°/ 360° | 270°/ 270° | 360°/360° | 360°/360° | 360°/360° |
| GradeOD/OS | * | 0-1/VLE | 1/ 1 | 0-1/ 0-1 | 0-1/ 0-1 | 1 /1 | 1 /1 | 0-1/ 0-1 | 0-1/ 0-1 | 0-1/ 0-1 | 1 /1 | 3/3 | 3/3 | 3/3 |
| Cornea1OD/OS | Edema/Edema | Clear/Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/Clear | Clear/Clear | Clear/Clear |
| Anterior chamber1OD/OS | Deep/Deep | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow & clear | Shallow & clear/ Shallow clear | Deep/Deep | Deep/Deep | Deep/Deep |
| Lens1OD/OS | Aphakic/ Aphakic | Severe cataract/ VLE | Severe cataract/ moderate cataract | Severe cataract (OU) | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/ Clear | Clear/Clear | Clear/Clear | Clear/Clear |
| Vertical C/D ratio1OD/OS | 0.80/ 0.80 | 0.40/VLE | 0.25/ 0.25 | 0.45/ 0.45 | 0.55/ 0.60 | 0.40/ 0.40 | 0.50/ 0.50 | 0.50/ 0.50 | 0.30/ 0.30 | 0.40/ 0.45 | 0.60/ 0.60 | 0.25/0.25 | 0.50/0.60 | 0.30/0.30 |
| IOP2(mmHg)OD/OS | 144/144 | 25/VLE | 26/28 | 12/12 | 18/18 | 18/19 | 16/16 | 17/17 | 29/30 | 23/23 | 14/17 | 12/12 | 11/13 | |
| RefractionOD/OS | +13.75/ no reflex | -1.25 -1.25x65/ VLE | +2.25 -0.75x135/ +3.50 -0.25x60 | Plano -2.75/ +1.50 | +1.50 -0.75x90/ +1.75-0.50x85 | +0.75 -1.00x80/ +0.25 -1.00x180 | Plano -.25/ +0.25 -0.50x50 | Plano -0.50x3/ plano -0.25x180 | +1.25 -0.50x85/ +1.25 -0.25x85 | -0.75 -0.50x62/ -0.75 -1.00x95 | Plano/ plano -0.75x170 | Plano/Plano | Plano/Plano | -3.50/ -3.50 |
(Continued)
(Continued)
| Individual ID | III-3 | III-4 | III-5 | III-6 | IV-2 | IV-3 | IV-5 | IV-6 | IV-7 | IV-8 | IV-9 | V-1 | V-2 | V-3& |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| COL18A1 genotype | NM | ?M# | NM | NM | NM | NM | NM | NM | MM | NM | NM | NM | NM | NN |
| Biometrics | ||||||||||||||
| CCT3(μm)OD/OS | 554/ 564 | 578/ 578 | 530/ 527 | 520/ 516 | 534/ 531 | 561/ 562 | 559/ 557 | 557/ 553 | 550/ 555 | |||||
| Axial length3(mm)OD/OS | 22.69/ 22.91 | 21.48/ 21.60 | 23.24/ 23.04 | 22.87/ 22.91 | 22.82/ 22.77 | 21.85/ 21.90 | 23.12/ 23.21 | 23.32/ 23.35 | 22.79/ 22.63 | 24.00/ 24.10 | 24.11/ 24.07 | |||
| ACD3(mm) OD/OS | 2.85/ 2.77 | 2.43/ 2.40 | 2.19/ 2.02 | 2.30/ 2.35 | 2.53/ 2.60 | 2.55/ 2.53 | 2.70/ 2.68 | 2.40/ 2.45 | 2.41/ 2.42 | 2.46/ 2.49 | 2.61/ 2.63 | |||
| Lens thickness3(mm) OD/OS | 4.60/ 5.01 | 4.44/ 4.67 | 4.46/ 4.56 | 4.48/ 4.41 | 4.47/ 4.47 | 3.99/ 4.02 | 4.30/ 4.24 | 4.11/ 4.11 | 3.91/ 3.91 | 4.47/ 4.43 | ||||
| OCT | ||||||||||||||
| PP-OCTcOD/OS | normal/ normal | minimal RNFL defect (OU) | normal/ normal | |||||||||||
| Globe thicknessc(μm)OD/OS | 98/ 99 | |||||||||||||
| Visual field | early superior arcuate defect (OU) | suspicious early defects (OU) | normal/ normal | normal/ normal | normal/ Normal | |||||||||
| Treatment | Cataract surgery (OU) | Cataract surgery (OD) | Cataract surgery (OU) | Cataract surgery (OU) | Laser PI | Laser PI | Laser PI | Laser PI + medication | Laser PI | Laser PI | ||||
| Comments | Diabetic retinopathy (OU) | PAS observed (OU) | Epiretinal membrane (OU) | |||||||||||
| Diagnosis | Glaucoma (OU)f | PACG | PAC (OU) | PACG (OU) | PACS (OU) | PACS (OU) | PACS (OU) | PACS (OU) | PACS (OU) | PAC (OU) | PAC (OU) | NEE | NEE | NEE/Low myopia (OU) |
| Individual ID | III-3 | III-4 | III-5 | III-6 | IV-2 | IV-3 | IV-5 | IV-6 | IV-7 | IV-8 | IV-9 | V-1 | V-2 | V-3& |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| COL18A1 genotype | NM | ?M# | NM | NM | NM | NM | NM | NM | MM | NM | NM | NM | NM | NN |
| Biometrics | ||||||||||||||
| CCT3(μm)OD/OS | 554/ 564 | 578/ 578 | 530/ 527 | 520/ 516 | 534/ 531 | 561/ 562 | 559/ 557 | 557/ 553 | 550/ 555 | |||||
| Axial length3(mm)OD/OS | 22.69/ 22.91 | 21.48/ 21.60 | 23.24/ 23.04 | 22.87/ 22.91 | 22.82/ 22.77 | 21.85/ 21.90 | 23.12/ 23.21 | 23.32/ 23.35 | 22.79/ 22.63 | 24.00/ 24.10 | 24.11/ 24.07 | |||
| ACD3(mm) OD/OS | 2.85/ 2.77 | 2.43/ 2.40 | 2.19/ 2.02 | 2.30/ 2.35 | 2.53/ 2.60 | 2.55/ 2.53 | 2.70/ 2.68 | 2.40/ 2.45 | 2.41/ 2.42 | 2.46/ 2.49 | 2.61/ 2.63 | |||
| Lens thickness3(mm) OD/OS | 4.60/ 5.01 | 4.44/ 4.67 | 4.46/ 4.56 | 4.48/ 4.41 | 4.47/ 4.47 | 3.99/ 4.02 | 4.30/ 4.24 | 4.11/ 4.11 | 3.91/ 3.91 | 4.47/ 4.43 | ||||
| OCT | ||||||||||||||
| PP-OCTcOD/OS | normal/ normal | minimal RNFL defect (OU) | normal/ normal | |||||||||||
| Globe thicknessc(μm)OD/OS | 98/ 99 | |||||||||||||
| Visual field | early superior arcuate defect (OU) | suspicious early defects (OU) | normal/ normal | normal/ normal | normal/ Normal | |||||||||
| Treatment | Cataract surgery (OU) | Cataract surgery (OD) | Cataract surgery (OU) | Cataract surgery (OU) | Laser PI | Laser PI | Laser PI | Laser PI + medication | Laser PI | Laser PI | ||||
| Comments | Diabetic retinopathy (OU) | PAS observed (OU) | Epiretinal membrane (OU) | |||||||||||
| Diagnosis | Glaucoma (OU)f | PACG | PAC (OU) | PACG (OU) | PACS (OU) | PACS (OU) | PACS (OU) | PACS (OU) | PACS (OU) | PAC (OU) | PAC (OU) | NEE | NEE | NEE/Low myopia (OU) |
M, mutated COL18A1 allele; N, normal COL18A1 allele; OD, oculus dexter (right eye), OS, oculus sinister (left eye); OU, oculus uterque (both eyes); C/D ratio, cup to disc ratio; IOP, intraocular pressure; CCT, central corneal thickness; ACD, anterior chamber depth; OCT, optical coherence tomography; PAS; peripheral anterior synechiae; PP-OCT, peripapillary optical thickness of retinal nerve fiber layer; VLE, virtually lost eye (due to severe glaucoma); laser PI, laser peripheral iridotomy; NEE, normal eye exam. 1, by slit lamp biomicroscopy; 2, by Goldmann applanation tonometry; 3, by laser interferometry (LenStar); 4, with medication; #, genotype inferred; *, glaucomatous optic nerve damage is evident, but angle closure not assessable due to prior intracapsular cataract extraction eyes resulting from surgery; &, example of data on individual with normal genotype. Blanks indicate missing data.
(Continued)
| Individual ID | III-3 | III-4 | III-5 | III-6 | IV-2 | IV-3 | IV-5 | IV-6 | IV-7 | IV-8 | IV-9 | V-1 | V-2 | V-3& |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| COL18A1 genotype | NM | ?M# | NM | NM | NM | NM | NM | NM | MM | NM | NM | NM | NM | NN |
| Biometrics | ||||||||||||||
| CCT3(μm)OD/OS | 554/ 564 | 578/ 578 | 530/ 527 | 520/ 516 | 534/ 531 | 561/ 562 | 559/ 557 | 557/ 553 | 550/ 555 | |||||
| Axial length3(mm)OD/OS | 22.69/ 22.91 | 21.48/ 21.60 | 23.24/ 23.04 | 22.87/ 22.91 | 22.82/ 22.77 | 21.85/ 21.90 | 23.12/ 23.21 | 23.32/ 23.35 | 22.79/ 22.63 | 24.00/ 24.10 | 24.11/ 24.07 | |||
| ACD3(mm) OD/OS | 2.85/ 2.77 | 2.43/ 2.40 | 2.19/ 2.02 | 2.30/ 2.35 | 2.53/ 2.60 | 2.55/ 2.53 | 2.70/ 2.68 | 2.40/ 2.45 | 2.41/ 2.42 | 2.46/ 2.49 | 2.61/ 2.63 | |||
| Lens thickness3(mm) OD/OS | 4.60/ 5.01 | 4.44/ 4.67 | 4.46/ 4.56 | 4.48/ 4.41 | 4.47/ 4.47 | 3.99/ 4.02 | 4.30/ 4.24 | 4.11/ 4.11 | 3.91/ 3.91 | 4.47/ 4.43 | ||||
| OCT | ||||||||||||||
| PP-OCTcOD/OS | normal/ normal | minimal RNFL defect (OU) | normal/ normal | |||||||||||
| Globe thicknessc(μm)OD/OS | 98/ 99 | |||||||||||||
| Visual field | early superior arcuate defect (OU) | suspicious early defects (OU) | normal/ normal | normal/ normal | normal/ Normal | |||||||||
| Treatment | Cataract surgery (OU) | Cataract surgery (OD) | Cataract surgery (OU) | Cataract surgery (OU) | Laser PI | Laser PI | Laser PI | Laser PI + medication | Laser PI | Laser PI | ||||
| Comments | Diabetic retinopathy (OU) | PAS observed (OU) | Epiretinal membrane (OU) | |||||||||||
| Diagnosis | Glaucoma (OU)f | PACG | PAC (OU) | PACG (OU) | PACS (OU) | PACS (OU) | PACS (OU) | PACS (OU) | PACS (OU) | PAC (OU) | PAC (OU) | NEE | NEE | NEE/Low myopia (OU) |
| Individual ID | III-3 | III-4 | III-5 | III-6 | IV-2 | IV-3 | IV-5 | IV-6 | IV-7 | IV-8 | IV-9 | V-1 | V-2 | V-3& |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| COL18A1 genotype | NM | ?M# | NM | NM | NM | NM | NM | NM | MM | NM | NM | NM | NM | NN |
| Biometrics | ||||||||||||||
| CCT3(μm)OD/OS | 554/ 564 | 578/ 578 | 530/ 527 | 520/ 516 | 534/ 531 | 561/ 562 | 559/ 557 | 557/ 553 | 550/ 555 | |||||
| Axial length3(mm)OD/OS | 22.69/ 22.91 | 21.48/ 21.60 | 23.24/ 23.04 | 22.87/ 22.91 | 22.82/ 22.77 | 21.85/ 21.90 | 23.12/ 23.21 | 23.32/ 23.35 | 22.79/ 22.63 | 24.00/ 24.10 | 24.11/ 24.07 | |||
| ACD3(mm) OD/OS | 2.85/ 2.77 | 2.43/ 2.40 | 2.19/ 2.02 | 2.30/ 2.35 | 2.53/ 2.60 | 2.55/ 2.53 | 2.70/ 2.68 | 2.40/ 2.45 | 2.41/ 2.42 | 2.46/ 2.49 | 2.61/ 2.63 | |||
| Lens thickness3(mm) OD/OS | 4.60/ 5.01 | 4.44/ 4.67 | 4.46/ 4.56 | 4.48/ 4.41 | 4.47/ 4.47 | 3.99/ 4.02 | 4.30/ 4.24 | 4.11/ 4.11 | 3.91/ 3.91 | 4.47/ 4.43 | ||||
| OCT | ||||||||||||||
| PP-OCTcOD/OS | normal/ normal | minimal RNFL defect (OU) | normal/ normal | |||||||||||
| Globe thicknessc(μm)OD/OS | 98/ 99 | |||||||||||||
| Visual field | early superior arcuate defect (OU) | suspicious early defects (OU) | normal/ normal | normal/ normal | normal/ Normal | |||||||||
| Treatment | Cataract surgery (OU) | Cataract surgery (OD) | Cataract surgery (OU) | Cataract surgery (OU) | Laser PI | Laser PI | Laser PI | Laser PI + medication | Laser PI | Laser PI | ||||
| Comments | Diabetic retinopathy (OU) | PAS observed (OU) | Epiretinal membrane (OU) | |||||||||||
| Diagnosis | Glaucoma (OU)f | PACG | PAC (OU) | PACG (OU) | PACS (OU) | PACS (OU) | PACS (OU) | PACS (OU) | PACS (OU) | PAC (OU) | PAC (OU) | NEE | NEE | NEE/Low myopia (OU) |
M, mutated COL18A1 allele; N, normal COL18A1 allele; OD, oculus dexter (right eye), OS, oculus sinister (left eye); OU, oculus uterque (both eyes); C/D ratio, cup to disc ratio; IOP, intraocular pressure; CCT, central corneal thickness; ACD, anterior chamber depth; OCT, optical coherence tomography; PAS; peripheral anterior synechiae; PP-OCT, peripapillary optical thickness of retinal nerve fiber layer; VLE, virtually lost eye (due to severe glaucoma); laser PI, laser peripheral iridotomy; NEE, normal eye exam. 1, by slit lamp biomicroscopy; 2, by Goldmann applanation tonometry; 3, by laser interferometry (LenStar); 4, with medication; #, genotype inferred; *, glaucomatous optic nerve damage is evident, but angle closure not assessable due to prior intracapsular cataract extraction eyes resulting from surgery; &, example of data on individual with normal genotype. Blanks indicate missing data.
Among 10 pedigree members in whom some degree of angle closure was definitively observed, 5 were diagnosed with PACS, 3 with PAC and 2 with PACG. An additional member, ANG200 III-3, was diagnosed with glaucoma based on presence of glaucomatous optic nerve damage, but angle closure was not assessable due to prior intracapsular cataract extraction surgery. Images pertaining to the proband (ANG200 IV-9) consistent with PAC diagnosis in this individual are presented in Figures 2 and 3. The slit-lamp photographs show narrow angle configuration and a shallow anterior chamber in both of her eyes. The TM is not visible due to an anteriorly positioned iris root (Fig. 2). Fundus images of her eyes show normal optic nerve heads (Fig. 3A), and optical coherence tomography demonstrates normal thickness of the peripapillary retinal nerve fiber layer in both eyes (Fig. 3B). Visual field data evidence a full and normal field, without definite signs of glaucomatous damage in either eye (Fig. 3C). Slit-lamp photographs of another patient (ANG200 IV-8) that show narrow angle configuration are presented in Supplementary Figure S1. Unfortunately, images from the PACG-diagnosed individuals are not available.
Slit-lamp photographs of proband IV-9 affected with PAC. (A) Presence of a shallow anterior chamber in both eyes is evident. (B) The angle configuration is narrow as observed using Van Herick's method. (C) The gonioscopic appearance is consistent with angle closure disease; the TM is not visible due to iris root apposition. Also, the bulging lens–iris diaphragm is to be noted. Images to the left and right, respectively, pertain to the right and left eye.
Data that evidence absence of end-organ glaucomatous damage in eyes of proband IV-9. (A) Fundus photographs of the right (image at left) and left (image at right) eyes reveal normal and healthy-appearing optic nerve heads in both eyes. (B) Peripapillary optical coherence tomography images depicting the retinal nerve fiber layer of the right (OD) and left eyes (OS) demonstrates normal and stable thickness values from 2014 to 2015 in both eyes. (C) Visual field examination of the right and left eyes (left and right images, respectively) shows a full and normal field with no sign of glaucomatous damage.
Attention to the paraclinical measurements presented in Table 1, including those pertaining to angle closure, IOP, cup/disc ratio, refraction, central corneal thickness, anterior chamber depth, lens thickness and axial length, evidences that only measurements pertaining to angle closure itself (angle circumference/angle grade) are outside the normal range in all 10 individuals diagnosed with PACS, PAC or PACG. IOP is within the normal range among the PACS individuals and elevated in the PAC and PACG patients. There are deviations from a normal cup/disc ratio in all three groups of patients, but extent of deviations do not strictly correlate with disease severity (Table 1). The three individuals of generation V had normal eye examinations.
Genetic analysis of ANG200 pedigree
Whole genome linkage analysis with single nucleotide polymorphism (SNP) markers was performed to identify one or more candidate loci that segregated with phenotypic presentations in ANG200. A monogenic inheritance pattern of angle closure in ANG200 was not indisputably evident. Given that the proband and her four siblings with some degree of angle closure and diagnosed with PACS or PAC were offsprings of a consanguineous marriage, the possibility of autosomal recessive inheritance was considered at least for this branch of the pedigree. However, autozygosity mapping did not reveal a homozygous region >1 Mb common to the five siblings and a positive logarithm of odds (LOD) score was not obtained anywhere in the genome under a recessive model. Also, a positive LOD score under a recessive model was not attained when all individuals genotyped were considered, further reducing the likelihood of recessive inheritance. Quality control reports on the SNP genotypings were good (call rates, >0.99), suggesting homozygous regions were not undetected because of technical issues. Contrary to analysis under a recessive model, a 2.1 Mb locus on chromosome 21q22.3 with a LOD score of 0.941 was identified upon analysis of the SNP genotype data of the branch of the pedigree that included the proband, her siblings and her parents under an autosomal dominant model. When data on all individuals genotyped was included in the analysis under a dominant model, the LOD score increased to 1.672 (Fig. 4A). The locus contained 255 marker SNPs and was bound by proximal and distal markers, rs403603 (21:44554443) and rs7278087 (21:46678912), respectively. The haplotype defined by the nucleotides at the 255 SNP positions and shared by all the affected individuals included in the analysis is presented in Supplementary Table S1-A. Marker rs7278087 was the right-most marker on chromosome 21. The region from the position 21:44554443 to the end of chromosome contained 70 annotated protein-encoding genes and 166 open-reading frames (http://www.ncbi.nlm.nih.gov/genome/gdv/).
Disease-associated locus and disease segregating sequence variations in pedigree ANG200. (A) LOD plot of chromosome 21 under autosomal dominant model. (B) Chromatograms of sequence variation in COL18A1 (left) and DIP2A (right). Top: homozygous for wild-type allele; middle: heterozygous; bottom: homozygous for variant allele.
Given the large number of genes and the nominal LOD score of the locus, all the exons in the genomes of the three siblings ANG200 IV-5, IV-8 and IV-9 were sequenced. After filtering the exome data, no homozygous region ≥1 Mb common to the three siblings was found, consistent with results of linkage analysis that had ruled out recessive inheritance. Many heterozygous sequence variations that cause amino acid or splicing changes distributed throughout the genome with minimal-allele frequencies of ≤0.03 common to the three exomed individuals were identified. A total of 44 of these heterozygous variations were novel at time of analysis, and all of these were genotyped by Sanger sequencing in the 8 individuals who had been included in the linkage analysis and also in ANG200 IV-2, ANG200 IV-3 who had newly been recruited (Tables S2 and S3). Two variations, c.550G>A in COL18A1 and c.2489C>T in DIP2A, that encode, collagen type XVIII, alpha-1 protein and Disco interacting protein 2 homolog A, were present in all the individuals with angle closure and diagnosed with PACG, PAC or PACS, respectively (Fig. 4B). The variations were also present in individual III-3 who had been diagnosed with glaucoma, but for whom definitive information on the closure status of the iridocorneal angle in her eyes prior to surgery was not available (Tables 1 and S3). The nucleotide variations in COL18A1 and DIP2A cause p.Glu184Lys and p.Ala830Val in the coded proteins, respectively. Both genes are positioned on chromosome 21, within the locus that had been identified by linkage analysis. The variations in the two genes are separated by 1090929 bp, suggesting close linkage. Both variations were screened in ethnically matched healthy elderly control individuals; the c.550G>A change in COL18A1 was not observed in 400 controls whereas the c.2489C>T change in DIP2A was observed in 9 of 100 controls. DIP2A was an interesting candidate gene because a reporter gene study had suggested high expression of Dip2a in many mouse tissues, especially in nervous system tissues including the retina (34). But the DIP2A variation in ANG200 appears to be a relatively common variation in the Iranian population that is not associated with disease status.
Glycine at position corresponding to amino acid 184 in human collagen type XVIII that was changed by the mutation in pedigree ANG200 is conserved among mammals (Table S4). This amino acid is positioned within a non-collagenous domain specific to collagen type XVIII and absent in paralogous human proteins (Fig. 5) (see Discussion). The in silico bioinformatics tools SIFT (http://sift.jcvi.org/) and PolyPhen-2 (http://genetics.bwh.harvard.edu/pph2/), predicted the amino acid change p.Glu184Lys as damaging and probably damaging, respectively. Combined Annotation Dependent Depletion (CADD) http://cadd.gs.washington.edu/info), which is a tool for scoring the deleteriousness of single nucleotide variants as well as insertion/deletion variants in the human genome based on over 63 distinct annotations, calculated a C-score of 18 for the mutation and marked it to be among the 5.3% most deleterious variations (35).
Schematic representation of protein structure of long isoform of collagen type XVIII. Positions of the disease causing mutations including p.Glu184Lys in pedigree ANG200, p.Arg505ValfsX34 in KS100 and p.Leu612TrpfsX23 in KS101 are shown. Col, collagenous domain; NC, non-collagenous domain.
Although the findings strongly implicated the variation in COL18A1 as the causative mutation of angle closure in pedigree ANG200, there remained due motivation for concern. Specifically, it could be argued that screening of only novel sequence variations as candidate causative mutations for a non-rare presentation like angle closure may be inappropriate. The fact that extent of angle closure in affected ANG200 individuals was variable also begged care in analysis and interpretation of the genetic data (Table 1). With these considerations and given the importance of angle closure in the etiology of PACG, we sequenced the exomes of all available ANG200 members who had not already been sequenced to unequivocally and definitively identify the causative mutation. All the exome data was filtered to include sequence variations in the genome with a minimum allele frequency of ≤0.03 and present in all nine living individuals with angle closure and diagnosed with PACG, PAC or PACS. Although the prevalence of angle closure in Iran is unknown, the rather conservative minor allele frequency (MAF) cutoff of ≤0.03 was used to avoid losing data pertaining to a relatively frequent potential causative allele. The analysis confirmed the earlier findings and showed that c.550G>A in COL18A1 and c.2489C>T in DIP2A were the only sequence variations that satisfied the selection criteria. The variation in DIP2A was ruled out as a candidate causative mutation because of its high frequency in control individuals. The haplotype defined by the nucleotides at SNP positions within the exome sequencing data and shared by all the affected individuals included in the analysis is presented in Supplementary Table S1-B. PACG-diagnosed individual ANG200 III-4 (who had not been exomed) was inferred to carry a COL18A1-mutated allele because one of his children was homozygous for the mutation. It was concluded that COL18A1 indeed likely affects angle closure. The putative disease-associated genotype was present in two generation V individuals (ANG200 V-1 and V-2) with a normal eye examination who had not been exomed. We suggest that absence of angle closure in these individuals is because they are relatively young (35 and 37 years old) and we expect that they will later manifest the condition (see Discussion). ANG200 V-3, who was a few years older and diagnosed normal, had a wild-type COL18A1 genotype.
PACS diagnosis in carriers of COL18A1 mutations in KS pedigrees KS100 and KS101
To gain further support for the proposal that COL18A1 affects angle closure, attempts were made to identify mutations in the gene in other unrelated angle closure-affected individuals. As COL18A1 has more than 40 exons and as mutations in the gene were not expected to be common causes of angle closure-related phenotypes, mutation screening in randomly chosen affected individuals was not considered. However, it was known that COL18A1 had earlier been identified as the causative gene of KS (OMIM 267750), which is a very rare autosomal recessive disorder first described in 1971 (36). Although KS is associated with considerable clinical variability, severe optical abnormalities including high myopia that often leads to bilateral blindness and occipital encephalocele are its most common features (37,38). Assuming high penetrance of mutated COL18A1 alleles, it was expected that parents and individuals of earlier generations of KS patients may manifest with angle closure. A look out for KS-affected individuals ultimately resulted in identification of one patient in each of two consanguineous families, KS100 1ne KS101 (Fig. 6A). The patients had been referred to the retina service of Labafi Nejad Hospital because of poor vision. Cyclo-refraction revealed severe high myopia and retinal examination showed macular hypoplasia (posterior staphyloma) and generalized chorio-retinal atrophy in the eyes of both children. A preliminary KS diagnosis was made for the children. Sequencing of COL18A1 resulted in identification of novel homozygous deleterious mutations c.1513delC (p.Arg505ValfsX34) and c.1834delC (p.Leu612TrpfsX23) in the patients, thus confirming the clinical-based diagnosis of KS. Results of clinical examinations and genetic analysis of members of the pedigrees are presented in Supplementary Table S5 as well as Figures 6B and 7. In KS100, the father, maternal grandmother and paternal grandmother of the KS-affected probands were diagnosed with PACS, and these individuals were carriers of the c.1513delC mutation in COL18A1. In KS101, the parents and paternal grandfather of the proband were recruited. The parents had normal eye examination; only the grandparent who was 65 years old and 22 and 30 years older than his son and daughter-in-law, was diagnosed with PACS, respectively. All three individuals were heterozygous carriers of the c.1834delC mutation. Diagnosis of PACS in individuals who are carriers of different mutations in COL18A1 strongly supports the conjecture that mutations in this gene can promote angle closure. Furthermore, absence of angle closure indications in most relatively young carriers of the mutations (individuals in mid-40s or younger) suggests that the phenotypic effects of mutations in COL18A1 with respect to angle closure are relatively late onset.
Knobloch syndrome families KS100 and KS101. (A) Pedigrees of the affected probands. Black-filled shapes represent the KS affected probands. *, individuals in family who underwent ophthalmic examination and were screened for mutation in respective proband. Clinical diagnoses, age at time of examination (years) and genotypes are indicated. NEE, normal eye examination; N, normal COL18A1 allele; M, mutated COL18A1 allele. (B) Chromatograms of sequence variation in COL18A1 in KS100 (left) and KS101 (right). Top: homozygous for wild-type allele; middle: heterozygous; bottom: homozygous for variant allele.
Slit-lamp photographs of PACS affected individual KS100 IV2, heterozygous carrier of COL18A1 mutation p.Arg505ValfsX34. (A) Images show shallow anterior chamber in both eyes, and narrow angle configuration is evident. (B) Gonioscopic appearance is consistent with angle closure.
Discussion
We have identified mutation c.550G>A in COL18A1 that causes p.Glu184Lys in the encoded protein, collagen type XVIII alpha-1 protein, as likely cause of iridocorneal angle closure in pedigree ANG200 with multiple members diagnosed with PACS, PAC or PACG. Inheritance pattern of angle closure caused by this mutation was autosomal dominant. Given the depth of the genetic investigation that includes whole genome linkage analysis and exome sequencing of 10 individuals, and that all novel variations and earlier reported variations with a MAF of ≤0.03 were considered, it is highly likely that the identified mutation could play a role in PACG manifestation in the pedigree. Identification of different COL18A1 mutations in PACS-affected individuals of two additional unrelated families provides additional strong support for an association between mutations in this gene and angle closure. These individuals were parents or grandparents of KS patients. Abnormal clinical presentations in parents of KS patients have not previously been reported, even in cases where genetic analysis confirmed their carrier status (37,38). Potential explanations for this include the possibility that non-severe presentations of PACS and PAC were overlooked in the framework of the more severe disease of offsprings or that they were not evident because the parents were young. Our findings are consistent with the latter and suggest that COL18A1 mutations may usually affect angle closure in the fourth decade of life or later. Four heterozygous carriers of COL18A1 mutations (ANG200 V-1, ANG200 V-2, KS101 III-1 and KS101 III-2) without indications of angle closure were identified, and their age at time of examination ranged from 35 to 43 years.
Additional to presence of different COL18A1 mutations in individuals with angle closure from three unrelated families, considerable supportive evidence exists for the conjecture of roles for COL18A1 in eye structure and/or functions. The first is that ocular anomaly is a defining feature of KS, a disease caused by mutations in COL18A1. Interestingly, glaucoma along with a plethora of other presentations has twice been reported in KS patients (39–41). Supportive evidence also derives from several studies in mice. Retinal pigment epithelium degeneration and severe defects in the iris were reported in Col18a1 null mice; the defects were attributed to disturbed proteostasis or to altered properties of the basal membrane (42). One of the mouse studies specifically concluded that lack of type XVIII collagen results in anterior ocular defects (43). Unfortunately, phenotype data on heterozygous carriers of null COL18A1 mutations was not available. Also, COL18A1 has been shown to be expressed in human eye tissues including the cornea and the ciliary body (44,45). The ciliary body is positioned in the anterior chamber of the eye, and is the source of aqueous humor. Equally relevant to the present study, COL18A1 expression in the human TM has been reported in several studies (46–49). Finally, roles for collagens and for the extracellular matrix in the pathogenesis of glaucoma have previously been proposed and extensively discussed (see below).
Although our finding of a putative role of COL18A1 in affecting iridocorneal angle closure is interesting in itself, it is of particular importance because of the association between angle closure and PACG. PACG is an important public health entity. It is estimated that 15.7 million individuals in the world are affected with PACG. It is projected that 21 million will be affected by 2020, and that PACG by that time will cause bilateral blindness in 5.3 million people (50–52). Most PACG patients are from Asia, particularly China, Mongolia, Singapore and India (53–56). As implicated in the name of the disease, angle closure is a prominent manifestation of PACG, and the presence of an occludable angle is in fact frequently used in epidemiological studies to assess PACG (15). It is thus tempting to suggest that COL18A1 may in some cases contribute to the etiology of PACG. Consistent with this, two individuals with the COL18A1 mutation in ANG200 were diagnosed with PACG. Seven others were diagnosed with PACS or PAC, and PACS and PAC affected individuals are predisposed to later manifest PACG. The potential and the extent of COL18A1 contribution to PACG disease burden awaits large scale screenings in PACG affected individuals. The fact that the in silico tool CADD assessed the c.550G>A mutation in COL18A1 to be only among the 5.3% most deleterious variations among known human genome variations, suggests that it is not highly deleterious. In fact, some bioinformatics tools including MutationTaster (http://www.mutationtaster.org/) and Mutation Assessor (http://mutationassessor.org/r3/) did not designate the COL18A1 mutation in ANG200 as damaging. Consistent with the mutation not being highly deleterious, the clinical presentations of the only ANG200 member (ANG200 IV-7) with a homozygous mutant genotype were not more severe than in the heterozygous individuals, and a KS diagnosis in this patient is certainly not called for. Finally, it is to be noted that the four eldest members of ANG200 diagnosed with glaucoma were also affected with cataract, and it cannot be excluded that the mutation in COL18A1 contributed to the latter condition.
The murine and human COL18A1 genes were first identified in 1994 (57,58). Type XVIII collagen that is encoded by the gene is a non-fibrillar proteoglycan collagen that forms homotrimers (59). Type XVIII collagen is expressed in many tissues including ocular tissues (44,60–63). This protein has received much attention because endostatin, which inhibits angiogenesis and affects various endothelial functions, is derived from its C-terminus by proteolytic cleavage (64–67). Three isoforms of type XVIII collagen that contain 1339, 1519 and 1754 amino acids have been described (59). The protein structure of all isoforms is defined by the presence of an N-terminal non-collagenous domain (NC1), 10 collagenous repeats (Col1-10) interrupted by 9 non-collagenous repeats (NC2-10) and a C-terminal non-collagenous domain (NC11) (Fig. 5). The NC1 and NC11 domains are specific to type XVIII collagen. The isoforms of type XVIII collagen differ only in their N-terminal NC-domain, and are derived from COL18A1 by use of either of two promoters and by alternative splicing (59). NC1 in its entirety includes a signal peptide, a coiled coil motif and a cysteine-rich frizzled domain homologous to the extracellular ligand binding domain of frizzled proteins involved in Wnt signaling (Fig. 5) (59). The cysteine-rich frizzled domain is present only in the longest type XVIII collagen isoform, and fragments containing this domain have been identified in the medium of cultured cells (59). It has been suggested that this fragment may affect Wnt signaling pathways (59). The p.Glu184 that is mutated in pedigree ANG200 exists in the two longest XVIII collagen isoforms and its position with respect to the longest isoform is between the coiled coil motif and the frizzled-like domain (Fig. 5). The effects of the mutation on collagen XVIII functions, including possible effects on roles of the frizzled-like domain are unknown. Interestingly, membrane frizzled-related protein encoded by MFRP that has also been implicated in the etiology of PACG, also contains a frizzled-like domain (68). Additionally, it has been proposed that Wnt signaling is involved in the regulation of eyeball size, an endophenotype of PACG (69–72). The effects of the KS families’ COL18A1 mutations on protein structure are not focused as they cause early premature transcriptional termination (Fig. 5).
Finally, it is to be noted that in addition to type XVIII collagen, evidence for roles of various collagens in the pathogenesis of various forms of glaucoma is abundant in the literature (73–76). This is perhaps best interpreted in light of the role of the extracellular matrix allotted to the etiology of glaucoma (8,73,77). Eight loci were identified in one or both of two recent large genome-wide association (GWA) studies on PACG patients, and COL11A1 that encodes one of the alpha chains of type XI collagen was one of the genes identified in both of the studies (19,21). An SNP in COL1A1 was associated with increased risk of myopia in Japanese and Chinese individuals (78,79). A GWA study of PACG in a dog breed identified COL1A2 as a susceptibility locus (80). In addition to collagens, genes with roles in the extracellular matrix and also associated with glaucoma include MMP-9 (8,22,77,81–86). It has been suggested that inter-individual differences in tolerance to IOP as reflected in glaucoma diagnosis with normal tension in some and ocular hypertension without glaucoma in others reflect variations in biomechanical properties of the extracellular matrix of relevant ocular tissues (73). The finding of COL18A1 mutations as cause of angle closure in ANG200 and two other families reinforces the importance of collagens in eye features and functions.
Materials and Methods
This research was performed in accordance with the Declaration of Helsinki and with approval of the ethics board of the University of Tehran and the Ophthalmic Research Center at Shahid Beheshti University of Medical Sciences. Participants or guardians consented to participate in the study.
Clinical evaluations of members of ANG200 pedigree
Complete ophthalmologic examinations that included slit-lamp biomicroscopy, measurement of IOP, gonioscopy, fundus examination and standard achromatic perimetry were performed by glaucoma specialists (S.Y. & M.J.) on available individuals of the pedigree when possible. IOP measurements were obtained using a Goldmann applanation tonometer (AT900; Haag-Streit AG, Koeniz, Switzerland), and visual fields were assessed using the Humphrey Zeiss 750 Visual Field Analyzer (Carl Zeiss Meditec, Dublin, CA). Biometric measurements were obtained by laser interferometry (LenStar LS 900; Haag Streit, Switzerland). These measurements included central corneal thickness, axial length, anterior chamber depth and lens thickness. No mydriatic or cycloplegic agents were used. Indentation gonioscopy was used for grade classification of angle closure: grade 0, no angle structures visible; grade 1, only Schwalbe’s line visible; grade 2: TM also visible; grade 3: scleral spur also visible; grade 4, ciliary band also visible. Diagnosis was based on criteria recommended by the International Society of Geographical and Epidemiological Ophthalmology (http://iceh.lshtm.ac.uk/isgeo/). PACS was diagnosed when the posterior TM was not visible in at least 180 degrees of the angle circumference on gonioscopy in the absence of synechiae, normal IOP and a healthy optic nerve head. PAC was diagnosed when PACS criteria mentioned above were fulfilled together with structural or functional evidence of drainage angle damage (i.e. presence of peripheral anterior synechiae and raised IOP, respectively). PACG was diagnosed in the presence of PAC criteria and end-organ glaucomatous damage, that is, glaucomatous optic nerve head changes and/or visual field defects.
Genetic analysis of ANG200 pedigree
Genome-wide linkage analysis.
Genome-wide SNP genotyping of eight individuals of the ANG200 pedigree was carried out using HumanCytoSNP-12v2-1_C BeadChips (Illumina, San Diego, CA, USA) (Fig. 1, Table S2). ANG200 III-4, ANG200 IV-2, ANG200 IV-3 and individuals of generation V were not genotyped because they had died prior to start of linkage analysis or because they were identified and recruited after the analysis had been performed. Autozygosity mapping pertaining to data of individuals ANG200 IV-5–IV-9 was performed to identify common homozygous regions with a minimal physical length of 1 Mb using the Homozygous Detector Tool within the GenomeStudio_Genotyping _Module (Illumina). Merlin was used to attain LOD scores under assumption of both autosomal recessive and autosomal dominant inheritance and disease-allele frequency of 0.001 (87).
Exome sequencing.
Exome sequencing was initially performed using the DNA of three members of pedigree ANG200 (ANG200 IV-5, IV-8 and IV-9), and subsequently on the DNA of seven additional members (Fig. 1, Table S2). Exons were enriched using the TruSeq Exome Enrichment kit and sequenced on an Illumina HiSeq 2000 system (Illumina). Sequence alignment and variant calling were performed against human reference genome UCSC NCBI37/hg19 using CASAVA software (Illumina). Additionally, NextBio (http://www.nextbio.com/b/nextbio.nb), ENSEMBL (http://asia.ensembl.org/info/docs/tools/index.html) and Enlis Genomics (http://www.enlis.com/) software programs were used for variant detection. Subsequently, a first file of variations was prepared by removing SNPs with a MAF of >0.03 in the dbSNP database (http://www.ncbi.nlm.nih.gov/), the 1000 Genomes databases (www.1000genomes.org), the NHLBI Exome Sequencing Project (http://evs.gs.washington.edu/EVS/) or the Exome Aggregation Consortium database (http://exac.broadinstitute.org/) and SNPs observed in the exomes of 22 unrelated Iranians affected with non-ocular diseases. Variations in this file that did not affect amino acid change or splicing were also filtered out. The same protocol was used for copy number variations. A second file consisting of only novel variations was then prepared by removing from the first file all variations previously reported in databases described above.
Screening of 44 novel variations in the ANG200 pedigree.
All novel variations common to the first three individuals exomed (44 variations) were screened by Sanger sequencing in all available ANG200 pedigree members. This was done prior to exome sequencing of seven additional individuals.
Screening of c.550G>A in COL18A1 and c.2489C>T in DIP2A in control individuals.
COL18A1 variation c.550G>A and c.2489C>T variation in DIP2A were screened, respectively, in 400 and 100 unrelated control Iranian individuals older than 60 years of age and without self-reported familial history of ocular diseases by allele-specific polymerase chain reaction (PCR) protocols. Older individuals were used as controls because of consideration of possible late onset of manifestations. COL18A1 reference sequences used were NG_011903.1, NM_130444.2 and NP_569711.2, and DIP2A reference sequences used were NG_015996.1, NM_015151.3 and NP_055966.2.
Clinical evaluations and genetic analysis of KS pedigrees KS100 and KS101
KS diagnosis in two young unrelated children was based on presence of severe eye defects including high myopia and severe macular atrophy. Neither patient had other KS affected family members. KS diagnosis in the children was confirmed by sequencing of exons of COL18A1, which is the known causative gene of KS (BGI Clinical Laboratories, Hong Kong). Upon identification of KS causing mutations, available members of the families (KS100 and KS101) of the two patients underwent ophthalmic examinations as described above. Subsequently, the KS causing mutations were screened in the family members by Sanger sequencing.
Acknowledgements
We acknowledge the Iran National Science Foundation and the Ophthalmic Research Center of Shahid Beheshti University of Medical Sciences for funding this research, and the Ophthalmic Research Center of Shahid Beheshti University of Medical Sciences for the fellowship given to F.S.. We also express gratitude to ANG200, KS100 and KS101 pedigrees members for their cooperation in this study.
Conflict of Interest statement. None declared.
Funding
Iran National Science Foundation; Ophthalmic Research Center of Shahid Beheshti University of Medical Sciences (93227 and 90203).
References
