Phenotype Variations Caused by Mutations in the RP1L1 Gene in a Large Mainly German Cohort

Abstract

Purpose: Mutations in the retinitis pigmentosa-1-like-1 (RP1L1) gene are the major cause of autosomal dominant occult macular dystrophy (OCMD), while recessive mutations have been linked to autosomal recessive retinitis pigmentosa (arRP). We present the clinical phenotype of a large German OCMD cohort, as well as four RP patients.

Methods: A total of 42 OCMD patients (27 families) and 4 arRP patients (3 families) with genetically confirmed mutations in RP1L1 were included. Genomic DNA was analyzed by targeted analysis of the c.133C>T;p.R45W mutation for all RP or macular dystrophy-related genes. All patients underwent ophthalmologic examination including psychophysical tests, electrophysiology, fundus autofluorescence (FAF), and spectral domain optical coherence tomography (SD-OCT). Follow-up time was up to 12 years.

Results: In 25 OCMD index patients genomic testing revealed the heterozygous mutation c.133C>T;p.R45W in RP1L1; one patient was homozygous for the mutation. Two OCMD patients displayed the variants c.3599G>A;p.G1200D and c.2849G>A;p.R950H, respectively, in a heterozygous state. All OCMD patients showed characteristic clinical findings and typical microstructural photoreceptor changes. Two arRP patients displayed the novel homozygous mutations c.3022C>T;p.Q1008* and c.1107G>A;p.W369*, respectively, while two RP-siblings carried the two heterozygous mutations c.455G>A;p.R152Q and c.5959C>T;p.Q1987*, the first also being novel. All arRP cases were mild with disease onset ≈30 years and preserved ERG-responses.

Conclusions: OCMD phenotype showed consistent clinical findings including classical microstructural changes on SD-OCT. An important hallmark of RP1L1-related OCMD is the dominant family history with reduced penetrance. Furthermore, novel mutations in association with arRP were identified, outlining the complexity of the protein.

Mutations in RP1L1 (retinitis pigmentosa-1-like-1; OMIM *608581) were originally identified in patients with occult macular dystrophy (OCMD; OMIM #613587),^1–4 a phenotype first described as a hereditary macular dystrophy with progressive visual decline but without visible fundus abnormalities.^5–7 The diagnosis of OCMD is mainly based on reduced amplitudes on focal macular ERG or multifocal ERG, whereas full-field electroretinography (ERG) displays normal responses.^6–10 With the development and refinement of optical coherence tomography (OCT), eyes diagnosed with OCMD by electrophysiology were found to have structural abnormalities in spite of the normal appearance of the macula on ophthalmoscopy.^7,11–14 Spectral-domain OCT (SD-OCT) imaging has shown subtle morphologic changes including blurring or disruptions in the ellipsoid zone (EZ) of the photoreceptor layer, the absence of the normal “bowing effect” and the loss of cone outer segment tips.^11–14 

In 2010, Akahori et al.¹ successfully linked the disorder to chromosome 8 and the RP1L1 gene. So far, a limited number of dominant mutations and variants of uncertain significance have been identified: most important is p.R45W,^1,3,4,15,16 which is recurrently found in all populations irrespective of ethnic descent, and indicative for a mutation hotspot, but also other missense variants have been discussed.^1,2,17 However, in addition to autosomal dominant pedigrees, sporadic cases have also been reported,^9,18 indicating de novo mutations but also reduced penetrance. Recently, few patients with typical autosomal recessive retinitis pigmentosa (RP) caused by homozygous RP1L1 mutations have been described.^19–21 In addition, a patient with cone dystrophy phenotype most likely caused by a homozygous missense mutation, c.3628T>C;p.S1210P, has also been reported,²² as well a single case with atypical RP, hearing loss, ataxia, cerebellar atrophy, and a complex heterozygous protein-truncating mutation in RP1L1 [c.326_327insT, p.(Lys111Glnfs*27) and c.7117C>T, p.(Gln2373*)], both on the maternal allele and a heterozygous nonsense mutation in C2orf71, p.(Ser512*).²³ These findings imply an important and diverse role for RP1L1 in human retinal physiology and disease. 

Our study summarizes detailed clinical characterization and long-term follow-up results of a large mainly German cohort of OCMD patients with mutations in the RP1L1 gene. Additionally, four autosomal recessive retinitis pigmentosa (RP) patients with mutations in the RP1L1 gene, including three novel variants explaining the disease phenotype are described here for the first time. 

Methods

Clinical Assessment

Clinical data from patients with mutations in the RP1L1 gene were collected. Examinations were carried out after written informed consent and in accordance with the tenets of the Declaration of Helsinki. The study was approved by the ethics committee of the Medical Faculty of University of Tübingen. For follow-up observations, data were partly collected in a retrospective fashion. 

Patients underwent complete ophthalmologic examination including psychophysical (decimal best corrected visual acuity [BCVA] and perimetry) as well as electrophysiologic tests (full-field and multifocal ERG). Semi-automated kinetic perimetry using Goldmann stimulus III4e within the 90° visual field and automated static perimetry within the 30° visual field were performed using a perimeter (Octopus 900; Haag-Streit International, Wedel, Germany). Full-field and multifocal ERGs were recorded according to the standards of the International Society for Clinical Electrophysiology of Vision (ISCEV) with different setups.^24,25 Morphologic examinations including color fundus photography, FAF, and SD-OCT (Heidelberg Engineering GmbH, Heidelberg, Germany) were performed to characterize detailed morphologic changes. 

Genetic Analysis

DNA was isolated from peripheral blood according to standard procedures and banked at the Molecular Genetics Laboratory of the Institute for Ophthalmic Research, Tübingen, Germany. 

Genotypes from patients with mutations in the RP1L1 gene were collected over a period of 6 years. In a first screen, exon 1 and flanking exon-intron boundaries of RP1L1 was PCR amplified with primers 5′-CCA TAG TGG AGT GGA GCA CA-3′ and 5′-CCA CTT ACC GCA TGC TCT CT-3′, and assayed by restriction fragment length polymorphism (RFLP) with restriction enzyme NcoI (New England Biolabs, Ipswich, MA, USA) according to manufacturers' recommendations to detect the common RP1L1 mutation c.133C>T;p.R45W in 25 patients with a clinical diagnosis of (occult) macular dystrophy. Mutations were confirmed by Sanger sequencing, using the same primers and PCR conditions. PCR products were purified by ExoSAP-IT treatment (GE Healthcare, Freiburg, Germany) and sequenced with dye termination chemistry (BigDye Terminator ver. 1.1; Applied Biosystems [ABI], Darmstadt, Germany). Sequences were run on a capillary sequencer (ABI 3100; ABI) and analyzed with sequence analysis software (version 5.1; ABI) and sequence trace alignment software (SeqMan; DNASTAR, Madison, WI, USA). In a subsequent screen, exons 1, 2 (forward primer 5′-GAG GTG TTA TGG GGA GTG GA-3′, reverse primer 5′-GAG GCA AGA GAG ATC CCA AA-3′) and exon 3 (forward primer 5′-ATT TTG GAG GCA GTG GTG ATG CTG AG-3′, reverse primer 5′-CGC AGA GCT CTC TGA CAC TTC TGG ACT TA-3′) of RP1L1 were analyzed in selected cases via Sanger sequencing as described above. The repetitive region covering amino acids 1,312–2,160, were not analyzed. In addition, 16 cases were identified by means of panel-based next-generation sequencing in a research or diagnostic-genetic setup,^26,27 followed by validation by Sanger sequencing as described previously. The RP1L1 cDNA is numbered according to Genebank NM_178857 with +1 corresponding to the A of the ATG translation initiation codon in the reference sequence. Segregation analyses were performed by PCR/RFLP (p.R45W) or Sanger sequencing upon availability of DNA of family members (Fig. 1). 

Figure 1

 

Pedigrees of 27 families with OCMD and three families with arRP. The solid squares (men) and circles (women) represent the affected patients. Unaffected family members are represented by white icons. The slash symbol indicates deceased individuals. The original index patient of each pedigree that was analyzed in the initial genetic testing is marked by an arrow, clinically examined individuals are indicated by the patient ID (i.e., OCMD01) that is also given in Table 1. The genotypes are given underneath the patient ID for those individuals that were clinically examined, or beneath each symbol for those individuals of which only DNA was available for testing and segregation analysis. Healthy mutation or obligate carriers are marked by a black dot within their gender symbol. Additional genetic findings by comprehensive retinal panel sequencing are also given beneath each individual if applicable (i.e., OCMD09 and OCMD21).

Statistical Analysis

The statistical analysis of the data was conducted by using statistical software (JMP 11; SAS Institute, Cary, NC, USA). Normality of data was assessed by evaluation of the histogram plots, correlation parameters were calculated using Pearson or Spearman analysis, where appropriate. Correlations between the right and left eye were assessed, for further analysis the right eye was selected. 

Results

OCMD Phenotype

We included 42 OCMD patients from 27 families (23 female and 19 male, range: 10–74 years, median: 36 years) with genetically confirmed mutations in the RP1L1 gene were included in the clinical study. Three additional OCMD affected family members were shown to carry the c.133C>T;p.R45W mutation, but were not available for clinical assessment. Initial clinical diagnosis included OCMD, Stargardt disease, macular dystrophy, cone dystrophy, and optic atrophy. Genetic and clinical findings of patients are summarized in Tables 1 and 2, pedigrees of the families including the results of the segregation analysis as well as genotypes are presented in Figure 1. In 39 individuals, genetic testing revealed the heterozygous mutation c.133C>T;p.R45W in RP1L1 establishing and confirming a genetic diagnosis of OCMD. In addition, one patient (OCMD40) was homozygous for this very same mutation. Another patient with OCMD was heterozygous for the variant c.3599G>A;p.G1200D (OCMD41) and one patient for c.2849G>A;p.R950H (OCMD42). Both variants are already published as pathogenic variants, but the latter may be rather considered a variant of uncertain significance with respect to its frequency in the Genome Aggregation Database (gnomAD; available in the public domain, http://gnomad.broadinstitute.org/; minor allele frequency [MAF] 0.02685%; including four homozygous cases listed in 275,576 alleles), and especially compared to the far more common mutation c.133C>T;p.R45W with a ×10 lower allele frequency in the normal population (MAF 0.0018%; three heterozygous in 274,220 alleles). In contrast, the c.3599G>A;p.G1200D mutation has never been observed in healthy subjects. It should be noted that a huge number of variants has been annotated in the gnomAD browser in RP1L1, and variant classification is challenging. Yet most variants are extremely rare. 

Table 1

 

Clinical Characteristics and Genetic Findings of RP1L1-OCMD Patients*

 

Table 2

 

Clinical Characteristics of RP1L1-OCMD Patients*

 

Family segregation analysis identified nine RP1L1 c.133C>T;p.R45W unaffected mutation carriers or obligate carriers in five families (Fig. 1: #2, #8, #11, #14, #18, #25), indicating and confirming that reduced penetrance is a hallmark of RP1L1-OCMD.¹ Within our cohort and families, we therefore calculate a penetrance of 82.7%. 

The age at which a patient recognized a decrease of vision varied widely from as young as 6 to 74 years of age (median age at disease onset 22.5 years). This may also explain the observed reduced penetrance, indicating that variable expressivity may also be a factor contributing to this, although it may be important to note that most of the anamnestically unaffected mutation carriers were not clinically investigated and may show subclinical signs of OCMD. Disease duration varied between 0 and 36 years (mean duration 11.5 ± 9.1 years). In two affected but unrelated children, the clinical diagnosis was made shortly after the first symptoms occurred (OCMD04, OCMD23) and confirmed, and two siblings of an affected parent were clinically diagnosed with OCMD presymptomatically in a routine ophthalmologic investigation before a decrease of vision had subjectively been noticed (OCMD25, OCMD26). Their genetic testing in adult age confirmed the diagnosis retrospectively. Eight patients were clinically examined once, the other 34 patients underwent routine clinical follow-up with a maximal examination time of 12 years (mean follow-up time 4.1 ± 3.5 years, age at first visit 37.8 ± 17.3, at last visit 41.9 ± 16.4 years). 

Two of the patients that had been analyzed by comprehensive diagnostic-genetic retinal or ocular panel sequencing displayed additional genetic findings. Patient OCMD09 was shown to carry an already published additional heterozygous mutation in ABCA4 c.5714+5G > A.²⁸ And more important, patient OCMD21 was shown to carry the two compound heterozygous mutations [c.1255C > T];[c.1327G > A] [p.R419W];[p.V443I] in OCA2, indicating that in addition to OCMD she is also affected by oculocutaneous albinism. Both mutations have already been reported in the literature.^29,30 The clinical findings of oculocutaneous albinism in this case included reduced pigmentation of the iris and fundus, as well as foveal hypoplasia on SD-OCT images. 

Figure 2

 

Characteristic findings of a representative OCMD patient. Functional and morphological results of patient OCMD25 are shown. Static perimetry reveals small relative central defects and reduced DLS (A), while multifocal ERG recording shows reduced responses in the central three rings (B). The fundus is essentially normal on funduscopy examination (C) and FAF imaging (D), however, SD-OCT reveals classical microstructural changes in the foveal area (E).

Figure 3

 

SD-OCT findings of OCMD patients. (A) SD-OCT findings of representative OCMD cases harboring the heterozygous mutation c.133C>T;p.R45W in the RP1L1 gene. Most of the patients presented with classical microstructural changes, the two youngest were categorized as nonclassical (OCMD04 and OCMD23). Interestingly, one patient presented with atypical foveal hypoplasia (OCMD21) and somewhat flattened foveal dip due to the additionally diagnosed oculocutaneous albinism. Patient ID is indicated in the upper right corner of each image. (B) Some progression in the microstructural defects could be observed over time, this is shown in two representative cases (OCMD11 and OCMD15). Patient ID and follow up time are indicated in the upper right corner of each image, black arrows point to the area, where structural changes of the photoreceptors could be observed at follow-up examinations. (C) Three patients with different genotypes are presented. One patient (OCMD40) was homozygous for the mutation c.133C>T;p.R45W, while the patient with OCMD (OCMD41) displayed the heterozygous variant c.3599G>A;p.G1200D, which was reported recently to be pathogenic. SD-OCT images in both cases show classical morphological changes including blurring of the EZ line and absence of the IZ. The last OCMD patient (OCMD42) displaying the variant of uncertain significance c.2849G>A;p.R950H, presented with pronounced loss of photoreceptor outer segments in the fovea. Patient ID is indicated in the upper right corner, genotype in upper left corner of each image. Additionally, an SD-OCT image of a normal subject is included for comparison.

All of the patients presented with characteristic functional and morphologic findings for OCMD on both eyes (Table 2, Fig. 2). Decimal BCVA at the first visit was 0.38 ± 0.33 (range between 0.05 and 1.5), while at the last visit BCVA was 0.27 ± 0.21 (range between 1.0 and 0.05), resulting in a decrease of visual acuity of only 0.12 ± 0.24 (median: 0.0), with most patients staying stable over time. The results of the right and left eye showed a strong correlation (r = 0.84 for BCVA at first and last visit as well). 

Kinetic perimetry was performed at least once in every patient, showing normal outer boundaries for the target III4e in every case. Static perimetry within the 30° visual field revealed relative central defects, the central differential luminance sensitivity (DLS) was moderately reduced to 23.3 ± 8.2 dB, where normal sensitivity level was at least 30 dB (static perimetry was not performed at every visit). During disease progression, a further slight decrease of DLS could be observed. BCVA and DLS results showed only a moderate correlation (r = 0.5), which could probably be explained with fixation problems during perimetry in low vision patients. 

Full-field ERG was performed in every patient at least once to rule out other possible diagnoses, which could cause a decrease of vision. In patients, where full-field ERGs were recorded in follow-up, no marked changes could be observed. The full-field ERGs were normal for both the rod and cone components in most patients, however, seven patients showed mild reduction of the cone ERGs, which did not change significantly over time. The central amplitudes of the mfERGs were reduced in all cases; changes of the central responses could be detected even in patients without vision loss. Unfortunately, estimation of yearly progression (i.e., loss of amplitudes) for patients could not be reliably calculated, since over the time of follow-up, various electrophysiologic setups were used and a proper comparison of all results was not possible. Therefore, mfERG results were calculated only for visits, where the same device (Espion System; Diagnosys LLC, Cambridge, England) was used for all recordings. A correlation between the central response amplitude and BCVA or DLS could not be observed (r = 0.28 and r = 0.16, respectively). 

On the morphologic level, funduscopy revealed no pathologic changes in the majority of the patients; however, in 14 cases subtle granulation of pigmentation in the macular area was observed. All of these patients were above 30 years of age and most of the patients above 50 years showed these changes, suggesting that these morphologic changes could also be caused by aging processes. FAF images were also essentially normal in the entire posterior pole; however, in five cases, increased AF signals and in two cases reduced AF signal were observed in the macular area. 

We further performed OCT measurements in most patients and—where possible—follow-up recordings were carried out. Four patients were not available for FAF and SD-OCT images. All investigated patients were classified into one of the two groups based on the microstructural changes of the photoreceptors according to Fujinami et al.¹⁴: One group with classical SD-OCT findings including blurring of the ellipsoid zone (EZ) and absence of the interdigitation zone (IZ), and a second nonclassical group in which at least one of the two classical features was lacking. While most of the patients presented with classical microstructural changes, the two youngest patients heterozygous for c.133C>T;p.R45W in RP1L1 were categorized as nonclassical (OCMD04 and OCMD23). Foveal thinning could be observed in all cases, mean central retinal thickness (CRT) was 187.3 ± 28.5 μm. In one patient (OCMD24), CRT values were higher due to epiretinal membrane (308 and 260 μm), these results were excluded from further analysis. Interestingly, one patient presented with atypical foveal hypoplasia (OCMD21) and somewhat flattened foveal dip (Fig. 3A). While CRT values did not change markedly over time, some progression in the microstructural defects could be observed (Fig. 3B). CRT values did not correlate with BCVA (r = 0.1), but a moderate correlation was found with the central response amplitudes of the mfERG (r = 0.5) and with DLS in static perimetry (r = 0.46). Patient OCMD21 presented with additional foveal hypoplasia due to oculocutaneous albinism. SD-OCT images of representative patients are presented in Figure 3. 

Interestingly, functional and morphologic values seemed to be independent from age generally or from the age at disease onset; however, BCVA showed a stronger correlation (r = −0.6) with disease duration. A high degree of right to left eye symmetry was observed for BCVA (r = 0.84) and CRT values (r = 0.87). 

Three OCMD patients need to be further discussed, since their genotypes differed from the majority of patients. One female patient (OCMD40) was homozygous for the mutation c.133C > T;p.R45W and was followed clinically for 10 years. First problems with decrease of vision and photophobia were noticed at the age of 13 years, the diagnosis of OCMD was confirmed at the age of 18 years in our clinic. BCVA was 0.2 on both eyes at first visit and remained unchanged over 10 years. Visual field examinations showed central relative defects and normal outer boundaries, characteristic for OCMD. Full-field ERG responses were within normal limits, while central responses of the mfERG were reduced, as expected. SD-OCT images showed classical morphological changes including blurring of the EZ line and absence of the IZ (Fig. 3C), while funduscopy and FAF images were inconspicuous. In conclusion, although the patient was homozygous for the mutation c.133C > T;p.R45W, no extraordinary clinical findings were observed. Yet it needs to be mentioned that segregation analysis confirmed that both parents are heterozygous for c.133C > T;p.R45W but do not report any visual deficits, in line with the observation of reduced penetrance in some families.¹ In a large family (family #11) carrying the heterozygous mutation c.133C > T;p.R45W mutation we also identified individuals with variable clinical symptoms, later onset and milder decrease of vision, which further suggest a reduced penetrance and variable expressivity associated with this mutation (Tables 1, 2). 

One male patient with OCMD (OCMD41) displayed the heterozygous variant c.3599G > A/p.G1200D, which was reported recently to be pathogenic.¹⁴ This male patient was seen in our clinic first at age 53 years, although first visual problems had been noticed 15 years before. BCVA was 0.25 and 0.2 on the right and left eye, respectively, visual field results were characteristic for OCMD showing central defects and normal outer boundaries. While rod responses were normal, a slight reduction of cone responses was observed in full-field ERG, and mfERG responses were markedly reduced in the central two rings. Morphologically, fine granulation of pigmentation in the macular area was observed funduscopically and also in FAF images as an inhomogenously reduced AF signal. SD-OCT findings, however, were also categorized as classical microstructural changes (Fig. 3C). 

The third OCMD patient (OCMD42, male) was analyzed by panel sequencing of all known macular dystrophy associated genes in a diagnostic-genetic setup. The analysis was extended to all known genes in inherited retinal dystrophy, but the only variant identified was the heterozygous variant c.2849G >A;p.R950H in RP1L1. This variant has already been reported by Davidson et al.¹⁹ to be associated with OCMD. Yet it may rather be considered a variant of uncertain significance as the MAF is too high for such rare disease like OCMD (0.02685%) and prediction programs rate this variant as benign (i.e., mutationtaster, SIFT, PolyPhen). His unaffected mother did not display this variant, and other family members were unfortunately not available. He presented at the age of 54 years, a slow decrease of visual acuity had been noticed several years before the diagnosis. At examination, BCVA was 0.5 on both eyes, visual fields showed the typical findings for OCMD. In full-field ERG recordings, cone responses were at the lower limit of normal values, while central mfERG responses were slightly reduced. Interestingly, the macular area showed fine pigmentary changes, which were observed as diffuse increased AF signal and pronounced loss of photoreceptor outer segments in the fovea (Fig. 3C). It also needs to be considered that the patient was classified as nonclassical OCMD, which rather supports that this genetic variant may not be pathogenic. 

RP Phenotype

While heterozygous, probably dominant-negative acting missense mutations are associated with autosomal dominant or sporadic OCMD, few arRP cases caused by biallelic, autosomal recessive mutations in RP1L1 have also been reported.^19–21 In our patient cohort, four such cases have been identified. 

The first patient (arRP1, male) was diagnosed with RP at the age of 35 years. First symptoms including night blindness had first been noticed 3 years before. BCVA was slightly reduced (0.63 on both eyes); however, visual fields for the target III4e were only slightly narrowed. Dark adaptation thresholds were only mildly increased, full-field ERGs still showed remaining cone responses, and preserved macular function in the mfERG recordings could be detected. Morphologically, attenuated vessels, pale optic disks, intact macular structure and subtle granulation of the RPE in the periphery without pronounced pigmentary changes were seen. SD-OCT images revealed intact retinal layers in the fovea and thinning of the photoreceptor layer toward the periphery. The patient displayed the novel homozygous nonsense mutation c.3022C > T;p.Q1008*, which had never been observed in any database and is predicted to be disease-causing. 

The second patient (arRP2, female, 45 years) noticed night blindness at the age of 25 years and a slow progression of visual problems (i.e., concentric narrowing of the visual field, slight decrease of central vision) over the years. At examination, BCVA was 0.8 and 0.63 on the right and left eye, respectively, visual fields for the target III4e were narrowed to approximately 15° without remaining peripheral islands. Dark adaptation thresholds were slightly elevated, full-field ERGs still showed remaining cone responses and mfERG recordings detected well- preserved macular function, similar to patient arRP1. Morphologic results as well as SD-OCT findings were also similar to patient arRP1 with slightly more pigmentary changes and RPE atrophy in the periphery. This patient is heterozygous both for the novel missense variant c.455G > A;p.R152Q and the already published mutation c.5959C > T;p.Q1987*, as is her affected brother (arRP3, male, 53 years). Unfortunately, the brother (the third patient, arRP3) could not be clinically examined and parents or further family members were not available for further segregation analysis. Consequently, we cannot formally prove that both mutations are biallelic and inherited independently in trans. But these mutations were detected by comprehensive panel sequencing of all known genes associated with inherited retinal dystrophies in a diagnostic-genetic setup, and these were the only two convincing variants observed, apart from single heterozygous variants in C2orf71 (c.3581C > T;p.A1194V) and in RPE65 (c.245+5A > G). The novel missense mutation c.455G > A;p.R152Q affects an evolutionary conserved amino acid residue located in the doublecortin domain of RP1L1 and is predicted to be disease-causing (i.e., mutationtaster, PolyPhen, SIFT). It is a rare variant (gnomAD browser MAF 0.03513%). The nonsense mutation c.5959C > T;p.Q1987* results in a premature termination codon and will result in a considerably truncated polypeptide if translated. Although this variant is predicted to be disease-causing, it has to be noted that it was observed in the gnomAD browser with a MAF of 0.1543%, and also four homozygous individuals are listed therein. 

The fourth patient (arRP4, male, 40 years) had been suffering from night blindness and slowly progressing visual field loss for 10 years. At examination, BCVA was 0.25 on both eyes, kinetic perimetry showed a concentric narrowing to 12° without remaining peripheral islands. Dark adaptation thresholds were markedly elevated. However, ERG findings were similar to the other three RP patients' results reported here. Morphologically, besides the typical changes in RP a bull's eye maculopathy with marked RPE atrophy could be seen, although SD-OCT still revealed good photoreceptor structure in the macular area. The patient displayed the novel homozygous nonsense mutation c.1107G > A;p.W369*. Other family members were not available for analysis but reported to be unaffected. The mutation is found in gnomAD browser with a low MAF of 0.00285% and is predicted to be disease-causing, and results in a premature termination codon and severely truncated polypeptide, if not undergoing nonsense-mediated decay. 

Again, the large number of variants observed in RP1L1 might suggest that such variants rather need to be considered benign as the protein has accumulated so many, yet it is important to know that although a large number of nonsense mutations are reported in the normal population, very few of these have been reported homozygously, and all patients reported here have been analyzed by comprehensive panel sequencing for all RP associated genes, minimizing the chance that the true molecular defect lies in other RP genes. 

In summary, all patients presenting with the RP phenotype showed rather mild symptoms with a disease onset around 30 years. Patients' results are presented in Table 3 and Figure 4. 

Table 3

 

Clinical Characteristics and Genetic Findings of RP1L1-arRP Patients*

 

Figure 4

 

Functional and morphologic findings of patients with RP1L1-associated arRP. From left to right: Kinetic perimetry results, color fundus images, FAF and OCT images are presented for the three patients suffering from RP1L1-associated RP. Patient arRP1 (top) displays the novel homozygous nonsense mutation c.3022C>T;p.Q1008* and shows only mild visual field defects, well preserved macular structure and subtle granulation of RPE in the periphery without pronounced pigmentary changes. Patient arRP2 (middle) is heterozygous both for the novel variant c.455G>A;p.R152Q and the already published mutation c.5959C>T;p.Q1987* and reveals concentric narrowing of the visual field, but intact retinal structure in the macular area. Likewise, patient arRP4 (bottom), displaying the novel homozygous nonsense mutation c.1107G>A;p.W369* also presented with narrowed visual fields, but due to the observed bull's eye maculopathy, BCVA was markedly reduced (0.25 on both eyes, respectively).

Discussion

Inherited retinal dystrophies are a heterogeneous group of rare diseases affecting the posterior segment of the eye, including photoreceptors and RPE. The diseases can be classified based on whether they predominantly affect the rods (i.e., RP) or the cones (i.e., cone dystrophy), causing a rather localized defect (i.e., macular dystrophy) or a more generalized retinal degeneration. A particular hallmark of retinal dystrophies is the impressive genetic heterogeneity. Some allelic genetic mutations lead to different clinical phenotypes (i.e., mutations in the ABCA4 gene can cause Stargardt disease, fundus flavimaculatus, cone dystrophy or RP), on the other hand, certain clinically indistinguishable forms of retinal degenerations (i.e., RP) can be associated with mutations in various genes and different genotypes.³¹ 

The RP1L1 gene was first described as a candidate gene for retinal degenerations by Conte et al.³² It encodes a 2480 amino-acid photoreceptor specific protein, the RP1L1 protein, which shows some homology with the retinitis pigmentosa 1 (RP1) protein: the N-terminal conserved regions of 350 amino acids in both proteins consist of two doubelcortin (DCX) tandem repeat domains followed by a 34 amino acid RP1 domain; while the C-terminal region differs in RP1 and RP1L1.^19,32,33 Both genes, RP1 and RP1L1 are localized on chromosome 8 and their expression is strictly restricted to the retina. The RP1 and RP1L1 proteins have been proposed to function together and colocalize to the axoneme of rod photoreceptor outer segments and connecting cilia in mice.^19,32,33 While Rp1^−/− mice display severe retinal phenotypes including outer segment misalignment and dysplasia,^34–37 studies of an Rp1l1^−/− mouse model have revealed a milder and later onset retina-specific phenotype compared to Rp1^−/− mice with scattered photoreceptor outer segment disorganization and progressive photoreceptor degeneration.³² In human, mutations in RP1 are associated with both dominant and recessive forms of RP^38–40 accounting for approximately 5.5% and 1% of dominant and recessive RP, respectively (Human Gene Mutation Database or HGMD).^41,42 

In 2010, Akahori et al.¹ successfully linked the disorder OCMD to chromosome 8 and the RP1L1 gene. Since the discovery of causative RP1L1 mutations in patients with OCMD, a number of cases have been reported.^{1,4,14–17,43,44} It seems that RP1L1 is the sole gene associated with autosomal dominant OCMD. The most common mutation is c.133C > T;p.R45W in exon 2,^1,14,16 but very recently, a second hotspot located between amino acid numbers 1196 and 1201—which is downstream of the doublecortin domain (DCX)—has been identified.¹⁴ According to the recommendation of Fujinami et al.,¹⁴ patients presenting with clinical symptoms of OCMD caused by mutations in the RP1L1 gene, belong to the subgroup of occult macular dysfunction syndromes, also called Miyake disease.¹⁴ According to this classification, occult macular dysfunction syndrome can be divided into three subcategories: RP1L1-associated OCMD (Miyake disease), other hereditary OCMD caused by other gene abnormalities, and nonhereditary occult macular dystrophy-like syndrome (progressive occult maculopathy). This nomenclature seemed necessary, since a significant association between the morphologic phenotypes and genotypes could be detected, indicating distinct pathophysiological processes underlying the occult macular dysfunction syndrome. 

In this study, our focus was set on a large, mainly German patient cohort harboring mutations in the RP1L1 gene. While most of the patients' genomic testing revealed the most common heterozygous mutation c.133C > T;p.R45W, similar to previously reported Japanese and Asian study populations, one further patient was homozygous for the same mutation and two patients with OCMD displayed the heterozygous variants c.3599G > A;p.G1200D and c.2849G > A;p.R950H, the latter being classified as variant of uncertain significance. Clinically, all patients presented with the classical findings, including reduced BCVA, central visual field defects, pathologic multifocal ERG (mfERG), normal fundus appearance, but typical changes in the microstructure of the photoreceptors detected by SD-OCT. The age of onset, rate of progression, degree of reduced central retinal function, as well as electrophysiologic and morphologic findings were all in line with previous study results.^7,13,44,45 Interestingly, the two youngest patients harboring the heterozygous mutation c.133C > T;p.R45W and the patient displaying c.2849G > A;p.R950H showed an atypical photoreceptor microstructural phenotype.²¹ Possible explanation for these findings could be the young age of these patients on one hand, and—in the latter case—limitation of mutation prediction as discussed above. Therefore, the latter patient might rather be diagnosed as nonhereditary occult macular dystrophy-like syndrome (progressive occult maculopathy). 

Previous studies also suggested a pathogenic role for RP1L1 in other retinal dystrophies. Two typical RP cases caused by the homozygous RP1L1 mutations c.601delG;p.K203Rfs*28, c.1637G > C;p.S546T and c.1972C > T;p.R658*, as well as a patient with a cone dystrophy phenotype caused by the homozygous missense mutation c.3628T > C;p.S1210P have been reported in the literature.^19–22 In our patient cohort, two patients with autosomal recessive RP were identified displaying the novel homozygous mutations c.3022C > T/p.Q1008* and c.1107G > A/p.W369*, respectively, while one RP patient and her brother were heterozygous for the two mutations c.455G > A;p.R152Q and c.5959C > T;p.Q1987*, the first also being novel. All RP cases were clinically mild with disease onset around 25 years and still preserved ERG responses. These findings support previous in vivo findings that the Rp1l1^−/− mouse has an adult onset degenerative retinal-specific phenotype.^19,33 Furthermore, the novel mutations reported here further underline the hypothesis that homozygous or compound-heterozygous loss of function variants in RP1L1 can cause arRP and may represent the null RP1L1 phenotype in humans.^14,19,33 

In summary, this study reports on the largest mainly German cohort of patients with mutations in the RP1L1 gene. Patients presenting with the OCMD phenotype did not seem to differ from the Asian population, therefore, the recently recommended classification from Fujinami et al.¹⁴ could also be helpful in our population as well. Characteristic clinical findings including classical microstructural changes in SD-OCT images and autosomal dominant family history with reduced penetrance and variable expressivity are important hallmarks of occult macular dysfunction syndromes linked to RP1L1. Furthermore, the fact that different mutations in RP1L1 are correlated with quite different morphologic and functional characteristics outlines the complexity of the protein. Identifying new mutations, and comparing the different phenotypes extends the mutation and phenotypic spectrum, and may help to better understand the function of the protein and the consequences in pathologic changes that involve cone and rod photoreceptors. 

Acknowledgments

Supported by BMBF Grant 01GM1108A (SK), TÜFF-Grant 2247-0-0 (DZ), and a grant of the Kerstan Foundation (GZ, DZ). 

Disclosure: D. Zobor, None; G. Zobor, None; S. Hipp, None; B. Baumann, None; N. Weisschuh, None; S. Biskup, None; I. Sliesoraityte, None; E. Zrenner, None; S. Kohl, None