CG is a heritable disease, with the majority of patients showing a genetic pattern of autosomal recessive inheritance; however, a number of patients show a sporadic pattern. Several genes (CYP1B1, FOXC1, LTBP2, etc.) have been found to be involved in its etiology, and CYP1B1 gene mutations explain some Japanese CG cases in Japan.11–13 In our subjects, heterozygous, compound heterozygous, and homozygous p.W57X, p.A202T, p.S215I, p.D274E, c.972_973insAT:p.I324fs, p.Q340X, p.V364M, p.V420G and p.D430E mutations were found in the CYP1B1 gene. Several studies have suggested that CG is related to polygenic inheritance involving more than one gene and etiological factors, but most of the CG gene spectrum remains to be elucidated. Thus, we screened CYP1B1 and FOXC1 mutations and performed WES in Japanese subjects to focus on the genetic background of CG in the Japanese population.
The prevalence of mutations in CYP1B1 was lower in this study than that reported in Saudi Arabians,3 but at almost the same as the rate reported in a Caucasian cohort2,31 and previous Japanese studies.12,13 According to our 14K Japanese Whole-Genome Reference Panel,24 24 individuals (1.1%) were carriers of a probable single pathogenic CYP1B1 mutation, which was a slightly higher rate than the expected carrier frequency (0.89%) in the United States.32 Our expected carrier frequency data could be consistent with the expected CYP1B1 mutation carrier frequency, and genetic testing for CYP1B1 mutations is a useful and accurate risk assessment method for some types of CG. In our study, the clinical evaluation of subjects with CYP1B1 mutations revealed that almost all of the subjects were bilaterally affected, with an early onset of the disease (less than 3 months of life) and a rather severe disease course.1
It has been reported that the CYP1B1 gene may interact with the MYOC gene as a modifier.33 Individuals with juvenile or early-onset glaucoma with mutations in both the MYOC and CYP1B1 genes may develop glaucoma at an earlier age and with more severe symptoms than those with only an MYOC mutation. In our study, there were no subjects with MYOC mutations, so the prevalence of MYOC mutations in CG could be low in the Japanese population. Moreover, we performed WES-based screening of the adult-onset open-angle glaucoma genes MYOC, OPTN, ASB10, TBK1, and WDR36, and we did not identify pathogenic variants. Furthermore, the identified p.A202T, p.D274E, p.Q340X, and p.V420G CYP1B1 gene mutations were novel, and it is possible that there are ethnic differences in the spectrum of genetics. As a number of CG-related mutations in the CYP1B1 gene have been reported, the evidence of a correlation between genotypes and disease phenotypes continues to accumulate. CYP1B1 mutations have also been reported in association with other clinical conditions, including secondary CG conditions such as Peters' anomaly and anterior segment dysgenesis.25 These associations are supported by the fact that both primary CG and Peters’ anomaly arise from defective neural crest cell migration during fetal development in the period of anterior segment formation. In our study, one subject (fam 129: G1199) showed Peters' anomaly with corectopia accompanied by high iris insertion.
FOXC1 is a member of the forkhead/winged-helix family of transcription factors, which contain a monomeric, 110-amino acid DNA binding domain (FHD). Various mutations in the FOXC1 gene have been implicated in the pathogenesis of a spectrum of ocular disorders,15,16,34,35 and FOXC1 gene dosage can cause anterior-chamber defects in the eye.36 We identified two missense mutations, p.Q70R and p.E163X, within the FHD, which contains regions at the N- and C-terminal boundaries critical for the proper nuclear localization of FOXA2 and FOXF237,38, and these mutations may cause changes in the expression pattern of a number of genes. Compared with the p.Q23X amino acid substitution39 found in Axenfeld-Rieger syndrome (ARS), which is associated with a variable degree of iris hypoplasia, it is noteworthy that our patients with the c.67delC frame-shift mutation, resulting in a truncated 43-amino acid protein, did not show apparent anterior segment dysgenesis. In our subjects, clinical ocular features of ARS, such as iris hypoplasia, corectopia, and a prominent, anteriorly displaced Schwalbe line (posterior embryotoxon), were either not seen or were subtle, and the reinvestigation of clinical phenotypes revealed only aortic regurgitation in the patient with the p.E163X mutation. It can be speculated that these mutations may affect the migration and/or differentiation of the mesenchymal cells that contribute to the anterior segment of the eye.40 FOXC1 mutations could affect normal development in the anterior segment of the eye and cause secondary CG. Regarding the prevalence of FOXC1 gene mutations, Chakrabarti et al. screened the FOXC1 gene in CYP1B1-negative CG and found a FOXC1 mutation rate of 2.38%,17 and Siggs et al. obtained a rate of 4.8% for FOXC1 variants in a cohort including CYP1B1-positive CG patients.18 Our screen showed FOXC1 variants in 3 out of 29 families (10%), representing a higher frequency of FOXC1 variants in CG patients than indicated by previous screening results. This study indicates that the FOXC1 gene may not play a major role in CG pathogenesis but that the FOXC1 gene does contribute to CG in the Japanese population.
We showed that the additional possible genotype–phenotype correlations found between the CYP1B1 and FOXC1 genes and CG may aid in predicting the disease prognosis and understanding the mechanisms associated with various types of anterior segment dysgenesis. Hollander et al41 first reported that CYP1B1 mutation caused severe-to-moderate angle abnormalities by identifying patients with angle dysgenesis who presented no cleavage and no development of Schlemm’s canal and showed mild dysgenesis, with the sole abnormality of mucopolysaccharide material deposition in the juxta-canalicular tissue. Garcia-Antón et al42 further studied the relationship between CYP1B1 genotype activity and angle abnormalities in CG and found that severe-to-moderate angle abnormalities ranged from the absence of CYP1B1 enzymatic activity to 60% CYP1B1 enzymatic activity. On the other hand, Nishimura DY et al36 reported that various types of FOXC1 mutations, including FOXC1 dosage abnormalities, cause various types of anterior chamber dysgenesis in ARA patients with or without systemic features. Unlike glaucoma associated with MYOC mutations, which are thought to be normal in outflow routes at birth,43 these gene mutations may cause obvious severe angle abnormalities resulting in the perinatal onset of glaucoma. Further study is needed to determine the relationship between the extent of angle abnormalities and the variety of CYPB1 or FOXC1 mutations.
It has been reported that PXDN mutations identified via homozygosity mapping in linkage analysis44 and WES19 cause corneal opacity and developmental glaucoma. Additionally, mutations in the TEK gene, which encodes a receptor tyrosine kinase that regulates vascular homeostasis, have been reported to likely underlie disease in an autosomal dominant pattern in patients with CG.8 In this study, we screened the CG-related genes LTBP2, TEK, ANGPT1, PITX2, PXDN and CPAMD8 through WES. We focused on 9 candidate genes related to signaling components of the ANGPT-TEK pathway (ANGPT2, TIE1, ANGPT4, and PTPRB in addition to TEK and ANGPT1) and genes expressed during SC development (PROX1, FLT4/VEGFR3, and VEGFC),45 but we did not identify possible pathogenic variants in those genes. According to our WES results, the causes of disease in most patients (17 families out of 29 families) remained unknown, and many of these patients did not share possible disease-causing mutations, revealing the heterogeneity of the genetic background of CG. This could mean that a common founder CG‑causing mutation does not appear in the Japanese population other than CYP1B1 and FOXC1 gene mutations, while certain CG‑causing mutations (e.g., CYP1B1, V364M) are more common in the Japanese population. This would be in accord with the involvement of a number of genes in the etiology of CG. Application of WES to additional cases may be needed to identify novel genes causing CG in patients who show no mutations in known CG disease-causing genes. Moreover, whole-genome sequencing would be useful for discovering novel genomic variants (e.g., structural variants, SNVs in splicing regions and noncoding regions of enhancers/promoters, insertion‒deletion variants, copy number variants).
In conclusion, in this study, we screened the CYP1B1 and FOXC1 genes and found that CYP1B1 and FOXC1 are two of the major causative genes of CG in Japanese individuals. This screening of the CYP1B1 and FOXC1 genes will be useful to ensure the proper diagnosis and adequate treatment of CG and for genetic counseling.