Molecular Vision 2019; 25:527-xxx
Received 23 June 2019 | Accepted 19 September 2019 | Published 21 September 2019
1Department of Ophthalmology, Flinders University, Bedford Park, South Australia, Australia; 2South Australian Health and Medical Research Institute, Adelaide, South Australia, Australia; 3SA Pathology, Flinders Medical Centre, Adelaide, Australia; 4Department of Ophthalmology, University of Melbourne, Melbourne, Victoria, Australia
Correspondence to: Owen M Siggs, Department of Ophthalmology, Flinders University, Bedford Park, South Australia, Australia; Phone: +61 8 8204 5062; FAX: +61 8 8277 0899; email: firstname.lastname@example.org
Purpose: Nanophthalmos is a rare subtype of microphthalmia associated with high hyperopia and an increased risk of angle-closure glaucoma. We investigated the genetic cause of nanophthalmos and high hyperopia in an autosomal dominant kindred.
Methods: A proband with short axial length, high hyperopia, and dextrocardia was subjected to exome sequencing. Human and rodent gene expression data sets were used to investigate the expression of relevant genes.
Results: We identified a segregating heterozygous frameshift variant at the 3′ end of the penultimate exon of MYRF. Using Myc-MYRF chromatin immunoprecipitation data from rat oligodendrocytes, MYRF was found to bind immediately upstream of the transcriptional start site of Tmem98, a gene that itself has been implicated in autosomal dominant nanophthalmos. MYRF and TMEM98 were found to be expressed in the human retina, with a similar pattern of expression across several dissected human eye tissues.
Conclusions: C-terminal variants in MYRF, which are expected to escape nonsense-mediated decay, represent a rare cause of autosomal dominant nanophthalmos with or without dextrocardia or congenital diaphragmatic hernia.
Refractive error is the leading cause of visual impairment, and the second leading cause of blindness worldwide . Precise developmental regulation of ocular axial length is essential to avoid refractive error, and the study of rare inherited refractive disorders has highlighted several critical genes and molecular pathways.
One of these rare disorders is nanophthalmos, characterized by high hyperopia associated with a reduction in posterior and anterior segment length, and a predisposition to primary angle-closure glaucoma [2,3]. Variants in at least four genes have been associated with nanophthalmos, with the majority due to recessive variants in membrane frizzled-related protein (MFRP; OMIM: 606227)  or protease, serine, 56 (PRSS56; OMIM: 613858) [5-7]. In rare cases, nanophthalmos may be inherited as a dominant trait, and we and others have reported two families segregating heterozygous missense variants in transmembrane protein 98 (TMEM98; OMIM: 615949) [8,9]. Common variants in PRSS56 and TMEM98 have also been implicated in multiple independent genome-wide association studies of myopia [10-12], highlighting the importance of studying nanophthalmos and other extremes of refractive error to understand the broader biology of common errors of refraction.
The latest gene to be implicated in nanophthalmos is myelin regulatory factor (MYRF; OMIM: 608329). The dominant NNO1 locus (OMIM: 600165), initially mapped to chromosome 11 in a large family in 1998 , was recently found to harbor a C-terminal essential splice variant in MYRF , with simultaneous independent reports of other variants in nanophthalmos and high hyperopia [15,16]. We describe an additional family with dominant high hyperopia and nanophthalmos, and reveal a segregating heterozygous frameshift variant in MYRF.
Patients and family members were recruited under the Australian and New Zealand Registry of Advanced Glaucoma . Written informed consent was provided under protocols approved by the Southern Adelaide Clinical Human Research Ethics Committee (305–08), and adhering to the tenets of the Declaration of Helsinki.
DNA was prepared from whole blood and subjected to exome capture (Agilent SureSelect v5, Santa Clara, CA) as described . DNA was prepared from venous blood samples, after temporary storage at -80°C, using the QIAGEN DNeasy Blood and Tissue Kit (Hilden, Germany), according to the manufacturer’s instructions. Paired-end libraries were generated and sequenced on an Illumina NovaSeq 6000 (San Diego, CA) instrument, with reads aligned to the GRCh37 human reference, and variants were joint called across samples according to GATK Best Practice workflows. Variant annotation was performed using Variant Effector Predictor (VEP) , and annotated Variant Call Format (VCF) files packaged into Gemini databases for downstream analysis . MYRF variant and exon coordinates refer to consensus transcript (NM_001127392.3) and protein (NP_001120864.1) sequences. Protein domains were as defined by UniProt (Q9Y2G1). Genomic Evolutionary Rate Profiling (GERP) constrained elements were defined by alignments across 35 mammalian species, and overlaid in Ensembl. MYRF variants were confirmed in a National Association of Testing Authorities (NATA)-accredited laboratory (SA Pathology, Flinders Medical Centre, Adelaide, Australia) using bidirectional capillary sequencing of the relevant PCR-amplified MYRF region. PCR products were sequenced and base called on an Applied Biosystems 3130xl Genetic Analyzer (ThermoFisher Scientific). PCR was performed using 100 ng of genomic DNA template in an AmpliTaq Gold reaction mix (Thermo Fisher Scientific, Waltham, MA). The following standard PCR conditions were used on a Veriti (Thermo Fisher Scientific) thermal cycler: Step 1, 95 °C for 10 min; Step 2, 95 °C for 30 s, 55 °C for 30 s, 72 °C for 1 min, repeated for 40 cycles; and Step 3, 72 °C for 7 min. PCR amplified products were prepared for sequencing with the ExoSAP method using 10 μl of each PCR reaction treated with 5 U of Exonuclease I (New England Biolabs, Ipswich, MA) and 1 U of Shrimp Alkaline Phosphatase (USB Corporation, Cleveland, OH) to remove residual primers and dNTPs. Bidirectional BigDye Terminator Cycle Sequencing (Thermo Fisher Scientific) reactions of the appropriate template and PCR primer were resolved and base called on an Applied Biosystems 3130xl Genetic Analyzer (Thermo Fisher Scientific). Detection of sequence variants was performed with the aid of Mutation Surveyor v4.0 (SoftGenetics LLC, State College, PA), with trace files assembled against the MYRF (NM_001127392.3) hg19 reference. The variant was deposited in ClinVar, with accession number VCV000635185.1.
Expression of human MYRF (204073_s_at) and TMEM98 (gnf1h00184_at) in selected tissues was retrieved from the GeneAtlas U133A data set . Human cadaveric eye tissue dissection and RNA sequencing (RNA-seq) were performed as described . Cadaveric human eyes with no known ocular disease were obtained from the Eye Bank of South Australia (Adelaide, Australia). Tissue dissection was performed under light microscope with a mean post-mortem time of 9.7 ± 5.3 h. Tissue from corneal epithelium, corneal stroma, corneal endothelium, trabecular meshwork (TM), pars plicata of the ciliary body, retina, optic nerve head, and optic nerve were collected and fixed in RNAlater solution (Thermo Fisher Scientific) for approximately 5 days prior to storage at -80 °C. A standard Trizol extraction protocol was used for RNA isolation (Thermo Fisher Scientific). RNA extracted from the pars plicata was passed through a Genomic-tip 20/G (QIAGEN, Hilden, Germany) as per the manufacturer’s instructions to remove excess melanin. RNA quality was assessed using the Agilent Bioanalyzer 2100 RNA 6000 Nano Assay (Santa Clara, CA; mean RIN = 6.5 ± 1.8). A Qubit 2.0 Fluorometer (Thermo Fisher Scientific) was used to quantify RNA.
We ascertained a 35-year-old female proband of European ancestry with high hyperopia (+13.00 D), and short axial lengths (17.53 to 17.72 mm), who had previously undergone bilateral peripheral iridotomies for angle closure (Figure 1A). This trait appeared to segregate in a fully penetrant autosomal dominant manner across four generations (Figure 1A). The daughter of the proband (IV:1) had shallow anterior chambers (1.98 to 2 mm) with no visible angle structures and associated ocular hypertension (28 to 30 mmHg), for which she was treated with bilateral peripheral iridotomies. The proband’s son (IV:2), had deep anterior chambers (3.67 to 3.7 mm) and no ocular hypertension. Examination of the proband’s father (II:1) was remarkable for disc crowding. None of the affected family members had glaucoma at their most recent follow-up. Microcornea, papillomacular folds, uveal effusion syndrome, or marked vascular tortuosity was absent. The medical history of the index case (III:2) and her brother (III:3) was remarkable for dextrocardia, and although there were no reported genitourinary abnormalities, a routine chest X-ray in her affected son (IV:2) revealed an incidental right congenital diaphragmatic hernia (Figure 1B). The spectrum of the clinical details for all examined family members is summarized in Table 1.
Exome sequencing was performed on DNA from the proband, with no evidence of rare variants (gnomAD maximum allele frequency <0.01) predicted to be deleterious (Phred-scaled CADD score >10) in canonical transcripts of the previously known nanophthalmos genes (TMEM98, PRSS56, or MFRP). Given the phenotypic similarities of this case with a recently described family segregating an essential splice variant in MYRF , we searched for rare variants in the same gene. This revealed a heterozygous 1 bp deletion in MYRF (NM_001127392.3 c.3361delC), absent from gnomAD (v2.1.1), and predicted to introduce a C-terminal frameshift at codon 1121, followed by a premature termination codon 36 amino acids later in the terminal exon (NP_001120864.1 p.(Arg1121Glyfs*36); Figure 1C,D). This variant was validated with capillary sequencing in the proband (III:2), and all available affected family members (II:1, III:3, IV:1, and IV:2).
Compared to previously reported MYRF variants associated with congenital diaphragmatic hernia, congenital heart disease, and/or genitourinary abnormalities [23-27] (Figure 1E,F), the variant described here was at the 3′ end of the penultimate exon of MYRF, and therefore, may escape nonsense-mediated decay due to the introduction of a premature termination codon in the terminal exon (Figure 1D) . The only described variants further downstream have been associated with isolated nanophthalmos and incompletely penetrant dextrocardia , or high hyperopia .
MYRF is translated as an endoplasmic reticulum (ER) membrane-bound protein, and following homotrimerization, is autoproteolytically cleaved to release an N-terminal transcriptional activator that translocates to the nucleus. One such transcriptional target is thought to be TMEM98, for which heterozygous variants have also been described to cause nanophthalmos [8,9]. Using a chromatin immunoprecipitation data set from rat oligodendrocytes probed with a Myc-MYRF construct , we identified a binding peak immediately upstream of and overlapping the Tmem98 transcriptional start site, which also overlaps a genetic element constrained across 35 mammalian species (Figure 2A). MYRF and TMEM98 were highly expressed in the human retina, with MYRF also highly expressed in the brain (Figure 2B). Within tissues of dissected cadaveric human eyes, MYRF and TMEM98 shared an expression pattern more similar to one another than to MFRP and PRSS56 (Figure 2C).
MYRF was initially described as a transcriptional regulator essential for oligodendrocyte differentiation and myelin gene expression . Missense, nonsense, frameshift, and essential splice variants in MYRF have since been implicated in syndromic presentations of congenital diaphragmatic hernia, congenital heart disease, and genitourinary abnormalities [23-27]. For those with available mortality data, 2/6 died within the first month of life, with a third case electively terminated at 19 weeks’ gestation . At least one syndromic case also was associated with extreme hyperopia and short axial length , although it is unclear if detailed ocular examinations have been performed on other syndromic cases. Similarly, it is unclear to what extent cardiac, diaphragmatic, and urogenital abnormalities have been investigated in individuals with apparently isolated nanophthalmos or hyperopia: Congenital diaphragmatic herniation was an incidental finding in at least one member of the kindred presented here.
The family presented here, along with others reported recently [14-16], represents a milder clinical presentation of MYRF-associated disease, with nanophthalmos or high hyperopia associated only occasionally with dextrocardia or congenital diaphragmatic hernia. There appear to be at least two important genetic distinctions between syndromic and non-syndromic MYRF variants. First, all syndromic MYRF variants reported thus far were found to be de novo, while at least three kindreds with nanophthalmos or high hyperopia show germline transmission of MYRF variants across multiple generations. Second, half (5/10) of the variants associated with nanophthalmos or high hyperopia occur in the final four exons of MYRF, or their associated splice sites. Three of these five affect either the terminal exon or the 3′ end of the penultimate exon, with a predicted termination codon introduced after read-through into the final intron. These three variants would, therefore, be expected to evade nonsense-mediated decay due to the 50 bp rule , and thus, be translated into protein with an altered C-terminus projecting into the ER lumen. The most C-terminal variant reported in a syndromic case to date is p.(Glu1081Glyfs*5), which lies in exon 25, and thus, expected to be degraded by nonsense-mediated decay . Therefore, it would appear that more severe syndromic presentations are usually associated with MYRF haploinsufficiency, while isolated nanophthalmos may sometimes be a consequence of hypomorphic, dominant negative, or potentially hypermorphic variants that affect MYRF homotrimerization, autoproteolysis, or transcriptional activity.
Conspicuous exceptions to this include recently described variants in nanophthalmos , such as two independent frameshift variants affecting codon 264 (c.789delC, c.789dupC). One of these variants (NM_001127392.3, c.789dupC, p.(S264Qfs*74)) was also reported in a syndromic case of extreme axial hyperopia with mitral valve prolapse, unilateral cryptorchidism, and micropenis . All three frameshift variants alter the length of an eight nucleotide cytosine mononucleotide repeat, and therefore, are a likely consequence of slipped-strand mispairing , but highlight that even N-terminal MYRF frameshift variants can lead to isolated nanophthalmos.
Variation in MYRF represents one of only two known genetic causes of autosomal dominant nanophthalmos. The second is TMEM98, first reported in a large European kindred . MYRF encodes a membrane-bound transcription factor, which releases an active transcription factor complex after homotrimerization and autoproteolytic cleavage [31-33]. One of the transcriptional targets of this complex is TMEM98, and the TMEM98 protein itself binds directly to MYRF in the ER, where it acts as an inhibitor of MYRF autoproteolysis and nuclear translocation .
Studies in mice have suggested that variants in TMEM98 associated with nanophthalmos represent gain-of-function alleles , which, in turn, would be predicted to increase inhibitory activity against MYRF. Although we cannot exclude the possibility of hypermorphic effects of certain MYRF alleles, reduced MYRF activity would be consistent with the predicted dominant-negative or loss-of-function effect of MYRF alleles suggested elsewhere [15,16], with similar support from the small eye phenotype of myrf knockdown zebrafish . This intimate connection of two genes associated with nanophthalmos and high hyperopia provides further support for their roles in ocular development.
Supported by the Australian National Health and Medical Research Council (NHMRC, Centres of Research Excellence Grant APP1116360 to JEC, Project Grant APP1107098 to JEC and OMS), and The Rebecca L Cooper Medical Research Foundation (Project Grant to OMS). JEC was an NHMRC Practitioner Fellow.