PAX6 Mutational Status Determines Aniridia-Associated Keratopathy Phenotype

Congenital aniridia (Online Mendelian Inheritance in Man identifier, 106210) is a rare, severely visually impairing disease caused principally by heterozygous mutation in the paired box 6 (PAX6) gene that orchestrates normal ocular development. 1Gehring W. J. The master control gene for morphogenesis and evolution of the eye. Genes Cells. 1996; 1: 11-15Crossref PubMed Scopus (215) Google Scholar The disease results in underdevelopment or abnormal development of eye structures including the cornea, leading to a bilateral and progressive limbal stem cell insufficiency and conjunctivalization of the cornea called aniridia-associated keratopathy (AAK). However, clinical manifestation of AAK, rate of progression, and prognosis can vary widely across individuals, precluding the development of general guidelines for treatment. Congenital aniridia can result from any of more than 400 unique mutations in the PAX6 gene that may lead to a spectrum of clinical phenotypes. 2Lim H. T. Seo E. J. Kim G. H. et al. Comparison between aniridia with and without PAX6 mutations: clinical and molecular analysis in 14 Korean patients with aniridia. Ophthalmology. 2012; 119: 1258-1264Abstract Full Text Full Text PDF PubMed Scopus (26) Google Scholar Aniridia-associated keratopathy phenotype can vary from a fully transparent cornea to a thick, opaque, vascularized pannus at any stage of life. As clinical genetic analysis becomes more sophisticated and widespread, the clinical consequence of various PAX6 mutations requires more detailed attention. However, to date, genotype–phenotype studies in aniridia describe the entire eye, 3Hingorani M. Williamson K. A. Moore A. T. van Heyningen V. Detailed ophthalmologic evaluation of 43 individuals with PAX6 mutations. Invest Ophthalmol Vis Sci. 2009; 50: 2581-2590Crossref PubMed Scopus (131) Google Scholar, 4Lee H. J. Colby K. A. A review of the clinical and genetic aspects of aniridia. Semin Ophthalmol. 2013; 28: 306-312Crossref PubMed Scopus (50) Google Scholar providing only general assessment of corneal opacity. Accordingly, we performed detailed clinical characterization of AAK phenotype across a range of ages and in parallel documented PAX6 mutational status to determine how genotype influences the clinical phenotype of AAK. Adult and pediatric patients with clinically diagnosed congenital aniridia included in a patient registry maintained at the Saarland University Medical Center, Department of Ophthalmology, Homburg/Saar, Germany, were identified. Patients provided blood samples for analysis at centers in Germany specialized in clinical genetics analysis and counselling. Genetic datasets were harmonized to match with the entries of the Leiden Open Variation Database (https: //www. lovd. nl/) referenced to GenBank sequence number NM₀00280. 4 from the National Center for Biotechnology Information. Mutations were checked against ClinVAR, Leiden Open Variation Database, and EXAC databases and a Google search. If the mutation was not found, it was regarded as novel. Clinical examinations were conducted at the Department of Ophthalmology, Saarland University Medical Center, and consisted of slit-lamp biomicroscopy (Haag-Streit, Koeniz, Switzerland) with digital photography to perform detailed grading of AAK, Cochet-Bonnet esthesiometry (Luneau Technology, Pont-de-l’Arche, France), anterior segment swept-source OCT (Casia2; Tomey GmbH, Nürnberg, Germany) for corneal thickness measurement, and laser scanning in vivo confocal microscopy (Heidelberg Retina Tomograph 3 with Rostock Corneal Module; Heidelberg Engineering, Heidelberg, Germany) to determine central corneal epithelial phenotype and subbasal nerve density. A detailed clinical phenotypic assessment of AAK was performed for all eyes to assign an AAK grade from 0 to 4, as shown in Figure S1 (available at www. aaojournal. org). Statistical regression analysis was performed using IBM SPSS Statistics version 25 (IBM Corporation, Armonk, NY). The collection of clinical and genetic data for this study was approved by the ethics committee of the Medical Association of Saarland (protocol no. , 144/15). Written informed consent to participate was obtained from all aniridia patients (or from one or both parents of children younger than 18 years with aniridia) following the tenets of the Declaration of Helsinki. Forty-six patients in the cohort were examined bilaterally (92 eyes). The mean ± standard deviation cohort age was 23. 0±17. 9 years and included 23 children (50%) younger than 18 years. Demographic, genetic, and phenotypic data are presented in Table 1, including 9 novel PAX6 mutations not reported previously. Five patients (11. 1%) showed non-PAX6 aniridia, that is, without detectable mutation in coding regions of PAX6 or other genes, based on either whole exome sequencing (3 patients 4, 5, and 10 years of age) or by multiplex ligation-dependent probe amplification analysis (2 patients 26 and 34 years of age). In the 2 latter cases, no PAX6 coding mutation was evident, but a heterozygous deletion of flanking genes was found, with deletion of ELP4 and DCDC1 in the first patient and deletion of ELP4, DCDC1, DNAJC24, and IMMP1L in the second patient. Table 1Molecular Cytogenetic and Clinical AAK Phenotype in the Aniridia CohortMutation TypeSubject No. FamilyNo. Age (yrs) GenderFunctional ConsequenceExon/IntronDNA ChangeProtein ChangeAniridia-Associated Keratopathy Grade (Right Eye/Left Eye) Distance-Corrected Visual Acuity SnellenSensitivity (mm) CNFL (mm/mm2) Epithelial PhenotypeCentral Corneal Thickness (μm) LOVDPMID entriesNon-PAX6 coding14MUnknown, no WAGR———0/050/4045/40—/12. 7—/co591/600—25FUnknown, no WAGR———1/180/8030/35—/——/—578/582—310FUnknown, no WAGR———1/0100/8050/5020. 9/11. 1Co/co594/589—426FGene deletion—ELP4, DCDC1—1/150/5055/5522. 8/21. 1Co/co624/601—534MGene deletion—ELP4, DCDC1, DNAJC24, IMMP1L—1/067/6730/3011. 1/17. 1Co/co639/641—Missense64M36 amino acid substitutionIntron 6c. 357+1G>Ap. 84₁19del1/2167/16745/3012. 9/4. 2Co/mix613/6488710M36 amino acid substitutionIntron 6c. 357+2dupTp. 84₁19del1/2100/10035/3014. 9/20. 8Co/co654/6583820F1 amino acid substitutionExon 5c. 80A>Cp. Gln27Pro1/167/10055/608. 9/7. 4Co/co579/601This study936F1 amino acid substitutionExon 6c. 266A>Cp. Gln89Pro3/3200/20020/300/—Conj/mix599/610This study1038F1 amino acid substitutionExon 5c. 86T>Gp. Ile29Ser1/2240/24035/253. 6/5. 6Mix/mix694/572This studyPTC1111MNMD inducingExon 5c. 112delCp. Arg38Glyfs*16-/0160/160—/——/——/—518/53591219MNMD inducingExon 5c. 112delCp. Arg38Glyfs*161/180/6745/35—/——/—627/631—134MNMD inducingExon 7c. 365C>Ap. Ser122*2/2180/36040/40—/—Co/co591/5973144MNMD inducingExon 7c. 386₃87delACp. Asn129Thrfs*32/2330/40030/30—/——/—631/641This study156FNMD inducingIntron 10c. 916+1G>Cp. (Val256Phefs*59) 2/2125/12525/35—/——/—579/5712167FNMD inducingExon 7c. 392delCp. Ala131ValFs*162/2400/40030/2514. 7/7. 9Co/mix594/545This study178MNMD inducingExon 5c. 109delGp. Ala37Profs*171/1400/20025/3515. 8/15. 0Co/co613/6283188FNMD inducingExon 8c. 584₅85delAAinsCAGp. Glu195Alafs*52/15000/40020/15—/——/—548/625This study19211MNMD inducingExon 7c. 391₃92delGCp. Arg131*2/2100/16025/206. 2/8. 1Co/co572/574This study20212MNMD inducingExon 7c. 391₃92delGCp. Arg131*2/3250/25030/30—/—Co/mix573/509—21246FNMD inducingExon 7c. 391₃92delGCp. Arg131*Enucleated/3—/80—/25—/0—/conj-/517—22311MNMD inducingExon 5c. 140A>Gp. Gln47Arg2/21000/40010/100/0Conj/conj985/580223342FNMD inducingExon 5c. 140A>Gp. Gln47Arg3/Kpro1000/LP35/—0/-Conj/—670/-—2413MNMD inducingExon 8c. 607C>Tp. Arg203*2/2200/20030/3016. 1/19. 4Co/co627/-272520FNMD inducingExon 9c. 718C>Tp. Arg240*4/4400/100030/150/0Conj/conj806/-352628FNMD inducingExon 9c. 718C>Tp. Arg240*2/280/12545/355. 2/7. 6Co/co636/634352728FNMD inducingExon 5c. 112₁16delp. Arg38ValFs*163/3750/75020/100/0Conj/conj657/596128434MNMD inducingExon 10c. 829C>Tp. Gln277*3/31000/40025/200/0Conj/conj595/521229464FNMD inducingExon 10c. 829C>Tp. Gln277*3/31000/LP—/50/0Conj/conj873/1026—3036MNMD inducingExon 5c. 76delCp. Arg26Glyfs*42/380/24040/155. 9/0Mix/conj665/692This study3142FNMD inducingIntron 5c. 141+1G>Aunknown4/4400/200020/300/0Conj/conj821/89033246FNMD inducingExon 5c. 130C>Tp. Arg44*1/2800/100030/2017. 2/9. 0Mix/mix599/60223354MNMD inducingExon 8c. 551delGp. Gly184Glufs*232/2200/16025/200. 6/6. 2Mix/mix616/72703457MNMD inducingExon 9c. 764A>Gp. Gln255Arg3/2200/10015/200/10. 7Conj/mix681/-13516FMultiple exon deletionExon 11–15 + ELP4 exon 9—3/enucleated12500/—5/—0/—Conj/—737/-—3652FMultiple exon deletionExon 5–6 + 15-base insertion—Transp/transp330/33010/50/0Conj/conj817/741—CTE378MRun-on mutationExon 13c. 1268A>Tp. *423Leuext*152/2400/40055/50—/——/—686/69763818FRun-on mutationExon 13c. 1268₁269delinsGTp. *423Cysext*152/2100/10030/206. 5/5. 5Mix/co621/618This study3918MRun-on mutationExon 13c. 1268A>Tp. *423Leuext*152/2400/20030/559. 5/15. 6Co/co655/62964051FRun-on mutationExon 13c. 1268A>Tp. *423Leuext*153/3400/100025/100/0Conj/conj714/-6Chromosomal414FWAGR, gene deletionPAX6, ELP4, DCDC1, FSHB, RCN1, WT1, HIPK3, LMO2, EHF, CB442/2333/16730/303. 1/0Mix/mix635/620—4211MWAGR, gene deletionPAX6, ELP4, WT1, del (11) (p13p13) 4/42400/240025/25—/——/—1071/894—4311MPAX6, no WAGRPAX6, del (11) (p13p13) 4/4240/400020/20—/——/—1172/915—4414FWAGR, gene deletionPAX6, WT1, del (p13-ter) 4/31000/125005/100/0Conj/conj1134/651—4520FWAGR, gene deletionPAX6, WT1, del (11) (p11. 2p13) 3/2LP/40055/25—/0—/conj670/623—Unknown4655F————3/41400/50000/250/0Conj/conj1034/996—— = indicates no information available or not relevant; CCT = central corneal thickness; CNFL = corneal nerve fiber length density of subbasal nerves, epithelial phenotype; Co = corneal cells; Conj = conjunctival cells; CTE = c-terminal extension; DCVA = distance-corrected visual acuity measured in Snellen units of feet (20/) ; Kpro = keratoprosthesis; LOVD = Leiden Open Variation Database (number of unique entries with associated publications) ; LP = light perception; Mix = mixed corneal/conjunctival cells; NMD = nonsense-mediated decay; PAX 6 = paired box gene 6; PMID = PubMed identification; PTC = premature termination codon; transp = corneal transplant; WAGR = Wilms’ tumor, aniridia, genitourinary anomalies, mental retardation. Values separated by a forward slash (/) indicate values for each eye for the given parameter. DCVA was not be evaluated in cases of Kpro. “This study” indicates novel mutation found in a subject or family. Open table in a new tab — = indicates no information available or not relevant; CCT = central corneal thickness; CNFL = corneal nerve fiber length density of subbasal nerves, epithelial phenotype; Co = corneal cells; Conj = conjunctival cells; CTE = c-terminal extension; DCVA = distance-corrected visual acuity measured in Snellen units of feet (20/) ; Kpro = keratoprosthesis; LOVD = Leiden Open Variation Database (number of unique entries with associated publications) ; LP = light perception; Mix = mixed corneal/conjunctival cells; NMD = nonsense-mediated decay; PAX 6 = paired box gene 6; PMID = PubMed identification; PTC = premature termination codon; transp = corneal transplant; WAGR = Wilms’ tumor, aniridia, genitourinary anomalies, mental retardation. Values separated by a forward slash (/) indicate values for each eye for the given parameter. DCVA was not be evaluated in cases of Kpro. “This study” indicates novel mutation found in a subject or family. Notably, AAK grade and corneal phenotype worsened with increasing degree of PAX6 mutation (i. e. , no PAX6 coding mutation, amino acid substitution, single exon, multiple exon, and chromosomal PAX6 gene deletion). Linear regression analysis indicated that AAK grade was associated strongly with type of PAX6 mutation (P < 0. 001) when adjusted for age and gender. For the entire cohort, AAK was age dependent (β = 0. 02; P = 0. 001), but the age dependence was significant only for premature termination codon (PTC) and C-terminal extension mutations (β = 0. 02; P = 0. 004), with other PAX6 mutation types being age independent (nonprogressive). Relative to non-PAX6 coding mutations, missense mutation resulted in an AAK grade increase of 0. 97 (β = 0. 97; 95% confidence interval, 0. 08–1. 85; P = 0. 03), PTC and C-terminal extension mutations resulted in an AAK grade increase of 1. 59 (β = 1. 59; 95% confidence interval, 0. 90–2. 28; P < 0. 001), and chromosomal mutations resulted in an AAK grade increase of 2. 69 (β = 2. 69; 95% confidence interval, 1. 80–3. 57; P < 0. 001). In patients with non-PAX6 aniridia, AAK was mild with a transparent cornea, relatively preserved visual acuity, near normal subbasal nerve density, moderately reduced ocular surface sensitivity, and moderately increased central corneal thickness. Missense mutations resulting in amino acid substitution (5 patients 11. 1%) resulted in a generally milder form of AAK that was not progressive, with comparatively good vision, modestly reduced sensitivity and subbasal nerve density, and moderately increased central corneal thickness. Patients with PTC mutations in PAX6 inducing nonsense-mediated mRNA decay comprised most cases (26 patients 57. 7%) in the present cohort and exhibited a classical AAK that typically is mild in childhood and progresses to a central corneal fibrovascular pannus after 20 years of age, 5Lagali N. Wowra B. Dobrowolski D. et al. Stage-related central corneal epithelial transformation in congenital aniridia-associated keratopathy. Ocul Surf. 2018; 16: 163-172Crossref PubMed Scopus (15) Google Scholar with associated loss of corneal sensitivity, nerves, visual acuity, limbal niche function, and elevated central corneal thickness resulting from the pannus. Of the PTC cases, 2 patients showed PAX6 mutations spanning more than 1 exon, resulting in a more severe phenotype; 1 patient had AAK grade 4 and multiple (failed) corneal transplantations bilaterally, whereas the other had undergone an enucleation of one eye (because of severe pain and buphthalmos without light perception) and AAK grade 3 in the other eye. Those with C-terminal extension mutations (4 patients 8. 9%) showed a similar classical progressive phenotype as those with PTC mutations. Those with chromosomal deletion of the PAX6 gene with additional deletions of 1 or more flanking genes (such as WT1, FSHB, DCDC1, ELP4, RCN1, HIPK3, LM02, EHF, and CB44) showed a severe, early aggressive AAK phenotype (5 patients 11. 1%), with 80% of these patients exhibiting Wilms’ tumor, aniridia, genitourinary anomalies, mental retardation (WAGR) syndrome and all demonstrating poor vision, thick opaque pannus, severe nerve deficit, and poor corneal sensitivity at a relatively young age (4–20 years). Finally, the central corneal subbasal epithelial plexus of all patients in the cohort regardless of age or genetic mutation status contained a large population of dendritic cells of activated phenotype not normally present in the healthy, noninflamed cornea. 6Mastropasqua L. Nubile M. Lanzini M. et al. Epithelial dendritic cell distribution in normal and inflamed human cornea: in vivo confocal microscopy study. Am J Ophthalmol. 2006; 142: 736-744Abstract Full Text Full Text PDF PubMed Scopus (165) Google Scholar Mutational correlation to detailed AAK morphologic features is rare, and therefore, clinical genetic and imaging studies are recommended in aniridia cohorts. Additionally, longitudinal natural progression studies are warranted to document AAK progression in the same individuals over a time span of several decades. The present findings indicate that AAK is not homogeneous. Although all cases of congenital aniridia have a minimal keratopathy (reduced touch sensitivity, increased corneal thickness, elevated dendritic cell invasion), further clinical and genetic classification of AAK into noncoding, mild, classical progressive, and early aggressive subtypes is recommended, as summarized in Table S1 (available at www. aaojournal. org). From a clinical management perspective, the AAK subtypes could be considered as separate disease entities, thereby facilitating treatment decisions7Käsmann-Kellner B. Seitz B. Aniridia syndrome: clinical findings, problematic courses and suggestions for optimization of care (aniridia guide) in German. Ophthalmologe. 2014; 111: 1145-1156PubMed Google Scholar and patient stratification for future clinical studies and trials. Download. pdf (. 26 MB) Help with pdf files Figure S1 Download. pdf (. 09 MB) Help with pdf files Table S1