Introduction

Choroideremia (CHM, OMIM: 303100) is a chorioretinal dystrophy inherited in an X-linked recessive manner with an incidence between 1:50,0001 and 1:100,0002. It is characterized by progressive degeneration of the retinal pigment epithelium (RPE), photoreceptors, and choroid3. Individuals with choroideremia usually present with a distinctive fundus appearance, featuring a scalloped choroid due to atrophy of the choroidal vessels4.

Choroideremia is caused by mutations in the CHM gene (OMIM: 300390), which is located at Xq21.2 and comprises 15 exons5 encoding Rab escort protein 1 (REP-1). Currently, 293 disease-causing variants in the CHM gene are listed in the Human Genome Mutation Database6.

Due to the monogenic nature and distinctive phenotype of this disease, direct sequencing of the CHM gene, with follow-up deletion/duplication analysis where required, has been highly effective for genetic confirmation of clinically diagnosed individuals7. However, next-generation sequencing (NGS) has sometimes unexpectedly identified CHM mutations in male and female individuals with an alternative clinical diagnosis, such as retinitis pigmentosa (RP)8,9,10. Thus, choroideremia may have a variable phenotype, leading to underreporting11.

Clinical trials for therapeutic gene replacement of the CHM gene are at an advanced stage. Following the first of these trials (NCT01461213)12, further phase 1/2 trials are complete or underway. A phase 3 trial (NCT03496012) is underway for 140 participants across clinical sites in the United States, Canada, Europe, and the United Kingdom13. No CHM gene therapy trials are underway in Australia.

The Australian Inherited Retinal Disease Registry and DNA Bank (AIRDR) previously genetically confirmed choroideremia in individuals from 11 Australian families7. The primary aim of the present study was to genetically characterize recently recruited pedigrees to update the spectrum and prevalence of CHM mutations in Australian families and to further consolidate potential candidates for future gene-specific clinical trials or treatments. Here, we identified 16 additional males with genetically confirmed choroideremia, two of whom are currently asymptomatic, and 14 asymptomatic carrier females sourced from 14 families. We also identified four female carriers with a vision-threatening phenotype. Such female patients are also relevant to gene-specific clinical trials or treatments but tend to be overlooked.

In the cohort described in our previous study, we genetically confirmed 16 affected males and 12 asymptomatic carrier females sourced from 11 families. Here, we present the combined mutation spectrum, which includes five novel CHM mutations, for all 32 genetically confirmed males from 25 Australian families.

Methods

Research participants

Participants were identified from the AIRDR14. Interrogation of the registry identified nine pedigrees not previously reported with at least one individual clinically diagnosed with choroideremia. Three additional pedigrees, each containing one participant clinically diagnosed with RP, were included in this study where previous analyses revealed potentially disease-causing CHM variants. Two families were added after testing negative for the X-linked RP genes RP2 and RPGR, resulting in a suspected diagnosis of choroideremia. In all, 34 participants from 14 families were included in this present study. They comprised 14 symptomatic males, two asymptomatic males, four affected females of varying clinical severity and 14 unaffected, suspected carrier females. For 10 out of 14 families, the proband was a male with clinical features consistent with choroideremia. For two families, there were no consenting affected male participants in the registry. These families are identified here by the numbers 12–25 to distinguish them from families 1 to 11 in our previously published study7.

Genetic analyses

DNA samples were collected, processed, and stored, as previously described14,15.

Proband DNA was analyzed using various methods (Table 1).

Table 1 Genetic analysis methods used for 34 study participants.
Full size table

Proband DNA was analyzed by Sanger sequencing of all 15 exons and flanking intronic regions of the CHM gene (Molecular Vision Laboratory (MVL), Oregon, or Australian Genome Research Facility (AGRF), Perth). Where a candidate disease variant was not detected, the possibility of a large deletion/duplication was investigated by TaqMan quantitative PCR (qPCR) (MVL or Casey Eye Institute (CEI), Oregon). Targeted Sanger sequencing was used to verify detected variants, where required, and for testing familial variants in family members.

RefSeq Accession NM_000390.2/3 was used in genetic analyses. Nucleotide 1 corresponds to the A of the ATG translation initiation codon. Sequence variant nomenclature is reported in accordance with the recommendations of the Human Genome Variation Society16.

Classification of variant pathogenicity

The pathogenicity of detected CHM variants was ascertained by interrogation of the scientific literature and disease- and locus-specific databases and by in silico analysis, as detailed previously15. Variant pathogenicity was classified in accordance with recommendations of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology (ACMG)17. Splice site variants at consensus dinucleotides (± 1/2) were automatically assigned a pathogenic status18,19.

Results

Phenotypic data

As the AIRDR is a national registry containing information from participants throughout Australia, phenotypic data such as imaging or electrophysiology results contained in the registry are often incomplete or self-reported. Nevertheless, we report here those phenotypic data that were available.

Male participants

At the time of this study, the age range of the 14 affected males was 8–60 years. Their reported ages of onset ranged from 3 to 28 years. All symptomatic males reported night blindness as a presenting feature, with six reporting constricted fields at this time. One male also reported photophobia as a presenting symptom (Table 2).

Table 2 Phenotypic information for male and symptomatic female participants in this study.
Full size table

Self-reported presenting symptoms were similar among the four families containing more than one symptomatic male (Families 13, 14, 15, and 20), as was age of onset within three families. One exception was Family 14, where self-reported ages of onset differed by nine years (Table 2). This disparity may relate to the generational gap between the proband and his maternal uncle, with the uncle’s existing diagnosis possibly alerting the family to the possibility of disease in the proband.

For each of the two families, family testing revealed the presence of the familial CHM mutation in an asymptomatic male with retinal features consistent with choroideremia.

Female participants

Fourteen of 18 females in this study were asymptomatic. Four females had reported symptoms of varying severity (Table 2). The ages of onset were in the third (n = 1) or sixth (n = 3) decade, with night blindness reported as the presenting symptom in all cases, with or without photophobia and/or visual field constriction. Since onset, worsening symptoms have significantly affected their quality of life.

Genetic results

A pathogenic or likely pathogenic CHM variant was identified in all 14 families analyzed (Table 3). All males (14 affected and two asymptomatic) and 16 out of 18 females were hemizygous and heterozygous, respectively, for the detected familial CHM variant.

Table 3 Disease-causing CHM sequence variants identified in the Australian cohort in the present study. Segregation was complete for all families, apart from Families 17 and 22, in which the variants established in the probands are presumed de novo.
Full size table

The mothers of two affected males did not possess the familial CHM variant, suggesting de novo events. The apparent de novo variants identified in unrelated, isolated males (17-1 and 22-1) are expected to result in premature termination codons (PTCs): the nonsense variant c.715 C > T, p.(Arg239*) has frequently been described in the literature, while the frameshifting duplication c.589dup, p.(Ser197Lysfs*2) is one of 5 novel variants identified in this study.

The frameshift variants c.589dup, p.(Ser197Lysfs*2), c.767_768del, p.(Glu256Valfs*2), c.999_1000insT, p.(Gln334Serfs*84) and c.1010_1015delinsCA, p.(Val337Alafs*6) are predicted to result in protein truncation and nonsense-mediated decay (NMD), with abolition of the protein. Accordingly, these variants have been classified as pathogenic.

The novel splice variant c.820-1 G > A is expected to be pathogenic because it occurs within a canonical splice site. This variant was heterozygous in an affected female proband (16-1) with no known family history of choroideremia. Other pathogenic nucleotide substitutions at this splice acceptor site have been described20,21,22. Notably, a similar CHM mutation (c.820-1 G > C) was reported in a female carrier who displayed a highly abnormal RPE without atrophy, with severe loss of visual acuity secondary to a presumed neovascular membrane20. In view of the absence of an affected male in Family 16 and this variant not previously described in the literature, we conservatively assessed c.820-1 G > C as likely pathogenic.

Overall, four nonsense, five frameshift and two canonical splice site mutations were detected in 11 families. Gross deletions were identified in the remaining three families, including an entire gene deletion in one family and deletion of exon 15 in two families. Although breakpoints were not identified for the exon 15 deletions, for the purpose of this paper, we have classified them as a single variant. Thus, 13 different causative variants were identified among the 14 families included in this present study.

A genetic diagnosis of choroideremia was therefore confirmed for all nine families with a clinical diagnosis of choroideremia, as well as for the five families with a clinical diagnosis of RP at recruitment, for which no other candidate variants had been detected by previous genetic testing. A clinical re-evaluation of the diagnosis has been made for three of these RP families and is being sought for the other two.

Discussion

Present study

In this study, 13 different CHM variants classified as pathogenic or likely pathogenic were identified in 14 Australian families. Five variants were novel, and two were de novo, including one novel variant.

As in other studies21,23, we identified a predominance of causative point mutations. Two gross deletions and one entire gene deletion were detected. Once considered rare, gross deletions now reportedly comprise approximately 20% of disease-causing CHM variants6. With the identification of an entire gene deletion or deletions of exon 15 among three families, gross deletions now comprise 12% of our combined Australian cohort. Although an exon 15 deletion has been previously reported24, the two cases presented in this study are the first reported in an Australian cohort. It is not known whether these families carry the same or distinct nucleotide deletions, as breakpoints were not determined. Similarly, owing to the absence of breakpoint data, it is not known whether the entire CHM gene deletion identified in this study is the same as those reported previously.

Entire gene deletions have been reported involving the CHM gene alone21,25,26 or in various combinations with other genes, which can result in complex syndromic choroideremia phenotypes11,24,27. In the absence of breakpoint data for these gross deletions, we cannot establish if they encompass only the CHM region or regions and/or regulatory elements of other genes. As associated medical conditions were self-reported as absent in all cases, it is likely that these deletions do not affect the function of other genes.

The clinical features and reported symptoms of carrier females in this study are consistent with the view that females are typically unaffected. Nevertheless, four female participants did show symptoms of varying severity. An underrepresentation of the contribution of CHM mutations to disease in affected carrier females may contribute to a diagnosis of RP with autosomal dominant transmission. It is important that severely affected female carriers of X-linked disease be included in considerations regarding inclusion in gene-specific clinical trials or treatments.

Notably, over one-third of the choroideremia-affected pedigrees in the present study were not initially clinically diagnosed with choroideremia. This finding supports the view that choroideremia is underdiagnosed and sometimes misdiagnosed as RP owing to the overlapping clinical features and presenting symptoms of these related conditions and, in some cases, also owing to atypical fundus features or severe phenotype in a female29. This highlights the value of nonhypothesis genetic diagnostic testing for suspected RP-affected individuals11,29,30.

Combined studies

The age distribution of the participants in the combined studies, classified by gender and affectation status, is shown in Table 4.

Table 4 Demographic information for CHM mutation carriers in the combined studies.
Full size table

Self-reported phenotypic data for individuals with disease-causing CHM variants in the combined studies for males and symptomatic females are shown in Supplementary Table 1. One affected male (7-2) was added to our previous study.

The sequence variants established in this study combined with our previous study are detailed in Supplementary Table 2.

In this combined study, we identified a predominance of causative point mutations, including frameshift, nonsense and canonical splice site mutations, with an absence of missense mutations, which is well documented for this gene21,22,23,28,31. Gross deletions were also reported in this study. The preponderance of such mutations suggests that most of the familial variants identified within this updated Australian cohort are likely to be null mutations, as found in other studies22,28,31.

Only one mutation, c.1584_1587del (p. Val529Hisfs*7), was detected in both studies. This frameshift variant is thought to occur at a mutation hotspot frequently reported in apparently unrelated pedigrees9,22,26,28,32,33,34.

Of interest, all six nonsense mutations detected across the combined choroideremia cohort of 25 pedigrees are C > T transitions (24% of pedigrees). Five of these are recurrent disease-causing CHM variants located at CpG dinucleotides, known mutational hotspots35,36. Two were detected in the present study (de novo c.715 C > T; c.799 C > T), and three were detected in our previous study (c.757 C > T; c.808 C > T; c.877 C > T). These results reflect the hypermutability for C > T transitions at CpG dinucleotides that occur at these five arginine residues (CGA), resulting in their conversion to a stop codon (TGA), as reported by others11,22,28,37,38.

These results indicate that Sanger sequencing of the CHM gene in probands with a clinical diagnosis of choroideremia remains an efficient tool in the molecular diagnostic pipeline. In the three families in which a mutation was not detected by Sanger sequencing, follow-up qPCR analysis identified gross deletions in CHM. In addition to these reported Australian families, other undiagnosed, untested or unborn male family members may prove to be candidates for future gene therapy.

This Australian cohort now consists of 25 genetically confirmed choroideremia-affected families, with a total of 23 different CHM mutations identified.