TRPM1 mutations are associated with the complete form of congenital stationary night blindness

Subjects

Four separate patients with complete CSNB (#204, #373, #437, and #484) in whom previous molecular examination revealed no mutation in either NYX or GRM6 among 11 separate patients were analyzed. They were all male, and their ages were 26, 9,19, and 27 years, respectively. They complained of night blindness from early childhood, and otherwise had no health problems. All individuals examined had been under observation at the Department of Ophthalmology of Nagoya University, Nagoya, Japan. The research protocol was designed in compliance with the Declaration of Helsinki and approved by the institutional review board. Written informed consent was obtained from the subjects after an explanation of the purpose of this study was provided. Briefly, the protocol entails molecular analysis of causative genes from blood samples of patients with complete CSNB and in vitro analysis including expression assay and cytological assay using the samples without any disadvantages for the patients.

DNA analysis of patients with complete CSNB

Genomic DNA was extracted from the peripheral blood leukocytes. Exons 2 through 27 of TRPM1 were amplified by polymerase chain reaction (PCR) using the DNA Thermal Cycler 9700 (Applied Biosystems, Foster City, CA). Primers were designed by and purchased from Invitrogen (Carlsbad, CA). The primer sets used in this study are shown in Table 1. For all exons, 100 ng of genomic DNA was amplified in a 50 µl reaction with 0.5 µM of each primer, 0.2 mM of each dNTP, and AmpliTaq Gold polymerase (Applied Biosystems). The PCR conditions were as follows: 5 min at 94 °C; 35 cycles at 94 °C for 30 s, followed by 30 s at 54 °C, and 45 s at 72 °C; and a final extension step at 72 °C for 7 min. The PCR products were directly sequenced using an ABI PRISM 3100 Genetic Analyzer (Applied Biosystems) and BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems). To search for polymorphisms, exons 3, 8, 16, 21, and 24 of TRPM1 from 200 alleles (50 men and 50 women) from unrelated, normal Japanese individuals were directly sequenced.

TRPM1 splicing minigene constructs

DNA fragments of TRPM1 exon 2 with its downstream splice donor region (nucleotide position in GenBank # NC_000015: 24738–25055; exon 2+SD) and exon 8 with its downstream splice donor region (39020–39392; exon 8+SD) were PCR-amplified using human genomic DNA as a PCR template; SacI and KpnI sites were appended to the 5′ and 3′ ends, respectively. A partial fragment of β-globin (HBB) intron 2 (614–1166 in GenBank # NC_000011) was PCR-amplified; KpnI and SalI sites were appended to the 5′ and 3′ ends, respectively. DNA fragments of TRPM1 exon 3 with its upstream splice acceptor region (31325–31691; SA+exon 3) and exon 9 with its splice acceptor region (39976–40280; SA+exon 9) were PCR-amplified using human genomic DNA as a PCR template; SalI and BamHI sites were appended to the 5′ and 3′ ends, respectively. The exon 2+SD, HBB intron, and SA+exon 3 fragments (exon 2+3 construct), or exon 8+SD, HBB intron, and SA+exon 9 (exon 8+9 construct) were ligated and subcloned into the SacI and BamHI sites of pBluescript KS+ (Stratagene), containing three repeats of the Flag-tag between the BamHI and NotI sites. The generated exon 2+3 construct and exon 8+9 construct were digested with SacI and NotI and subcloned into SacI and BamHI digested pEGFP-C3 and pEGFP-C2 (Clontech, Palo Alto, CA), respectively, each with a synthesized BglII-NotI linker.

Transfection and western blots analysis

HEK293T cells were cultured in D-MEM containing 10% fetal bovine serum (FBS; Nissui, Tokyo, Japan). These cells were grown under 5% carbon dioxide at 37 °C. Transfections were performed using the calcium phosphate method. Transfected cells were incubated at 37 °C for 48 h, and harvested for further analysis. Cell extract proteins were separated by SDS–PAGE, and transferred to a polyvinylidene difluoride membrane (ATTO, Tokyo Japan). The membrane was incubated with a mouse anti-Flag antibody (1:1,000; Sigma, St Louis, MO), a rabbit anti-β−gal antibody (1:10,000; Chemicon, Temecula, CA), or a mouse anti-β−actin antibody (1:2,000; Sigma), and incubated with a horseradish peroxidase-conjugated goat anti-mouse or -rabbit IgG (1:10,000; Zymed Laboratories, San Francisco, CA). The bands were visually developed using Chemi-Lumi One L (Nacalai Tesque, Kyoto, Japan).

In vivo electroporation and section immunohistochemistry

In vivo electroporation was performed at postnatal day 0 (P0) as described previously [23]. Approximately 0.3 μl of DNA solution (5 μg/μl) was injected into the subretinal space of P0 mouse pups, and square electric pulses (80 V; five 50 ms pulses with 950 ms intervals) were applied with electrodes. Wild-type (WT) and mutant forms of TRPM1 (R624C and F1075S) containing 3× Flag at the C-terminus were expressed from the mouse mGluR6 promoter (mGluR6-TRPM1–3×Flag). To generate this construct, the chicken β-actin gene (CAG) promoter region in CAG-TRPM1 was replaced with the 9.5 kb fragment of mGluR6 5′ upstream genomic sequence [24]. mGluR6-TRPM1–3×Flag (4 μg/μl) and CAG-GFP (1 μg/μl) were co-electroporated and harvested at P14. Immunohistochemistry to observe the subcellular localization of TRPM1–3×Flag expressed in the electroporated ON BCs was performed on sections as previously described [23]. Mouse anti-Flag antibody (1:500), rabbit anti-PKCα antibody (1:20,000; Sigma), and rat anti-GFP antibody (1:1,000; Nacalai Tesque) were used. The signal intensity of Flag immunostaining at the dendritic tips and somas of ON BCs was measured using AxioVision software (Carl Zeiss, Oberkochen, Germany). The ON BCs that showed more intensity in the dendritic tips than in the somas were counted. One hundred electroporated ON BCs were counted from three retinas independently prepared for each construct.

Clinical evaluations in patients associated with TRPM1 mutations

The ophthalmologic examination included best-corrected visual acuity, refraction, biomicroscopy, ophthalmoscopy, fundus photography, Goldmann kinetic perimetry, color vision testing, and electroretinography [25]. Standardized full-field ERGs were elicited by Ganzfeld stimuli after pupillary dilation with 0.5% tropicamide and 0.5% phenylephrine hydrochloride and 20 min of dark-adaptation. The rod (scotopic) ERGs were elicited by a blue stimulus with a luminance of 5.2×10⁻³ cd/sec•m². The mixed rod-cone, single flash (bright white) ERGs were elicited by a white stimulus of 44.2 cd/sec•m². The photopic (cones) single-flash ERGs and 30 Hz flicker ERGs were elicited by white stimuli at a luminance of 4 cd/sec•m² and 0.9 cd/sec•m², respectively, on a white background of 68 cd/m². The single flash bright white ERGs were recorded again after 17 h of dark adaptation.

ERG recordings in the trpm1 null mutant (trpm1^–/–) mice

We generated trpm1^–/– mice as described previously [11]. Our methods for recording the ERGs from the mice have been described in detail [26]. In brief, the mice were dark-adapted overnight, and then anesthetized with an intramuscular injection of 70 mg/kg ketamine and 14 mg/kg xylazine. ERGs were recorded with a gold-wire loop electrode placed on the anesthetized cornea. The mice were placed in a Ganzfeld bowl and stimulated with stroboscopic stimuli of 1.0 log cd-s/m² (photopic units) maximum intensity. Four levels of stimulus intensity ranging from −5.0 to 1.0 log cd-s/m² were used for the dark-adapted ERG recordings, and four levels of stimulus intensity ranging from −0.5 to 1.0 log cd-s/m² were used for the light-adapted ERGs. The light-adapted ERGs were recorded on a rod-suppressing white background of 1.3 log cd-s/m² (photopic units). The amplitude of the a-wave was measured from the baseline to a set time of 8 msec after the initial negative wave for the dark-adapted ERG, and 11 msec for the light-adapted ERG. The b-wave amplitude was measured from the negative trough of the a-wave to the maximal positive peak.

Animal care

All procedures conformed to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research, and these procedures were approved by the Institutional Safety Committee on Recombinant DNA Experiments and the Animal Research Committee of Osaka Bioscience Institute. The mice were housed in a temperature-controlled room with a 12h:12h light-dark cycle. Fresh water and rodent diet were available at all times.

DNA analysis of the complete CSNB patients

Five different mutations in TRPM1 were identified: IVS2–3C>G, IVS8+3_6delAAGT, R624C (c.1870C>T), S882X (c.2645C>A), and F1075S (c.3224T>C). Mutations were present in three unrelated patients with complete CSNB (#373, #437, and #484). All three patients were compound heterozygotes. Their pedigrees and the DNA sequence chromatogram for the patients harboring these mutations are presented in Figure 1A,B.

The two intron changes, IVS2–3C>G and IVS8+3_6delAAGT (Figure 1C), were interpreted to be pathogenic because no RNA splicing activities were detected when each of the two changes was present (see splicing assay section). They were also predicted to affect RNA splicing, as assessed by an online splice-site prediction tool. The donor site prediction score from the software for the exon 8 donor site in TRPM1 was 0.99 for the WT sequence, whereas it was under 0.01 for the corresponding mutated sequence with the IVS8+3_6delAAGT sequence change. The acceptor site prediction score for the exon 3 acceptor site in TRPM1 was 0.98 for the WT sequence, whereas it was 0.44 for the mutated sequence with the IVS2–3C>G sequence change. The two mutations were not found in an analysis of 100 Japanese normal controls.

The S882X mutation was considered to be a pathogenic null mutation since mRNA containing the predicted early stop codon would most likely be subject to nonsense mediated decay without being translated into a protein [27]. Even if the translated protein was produced, it would lack the C-terminus region of TRPM1, including transmembrane domains 4–6, which is very likely to lead to a loss of function as a cation channel (Figure 1D). This change was not found among the 100 healthy control individuals we examined. The two missense changes in the TRPM1 gene, R624C and F1075S, were located in the N-terminal region and in the C-terminal end of the transmembrane domain, respectively (Figure 1D). They were interpreted as likely pathogenic because the arginine at position 624 and the phenylalanine at 1075 are well conserved among the TRPM subfamily, suggesting the importance of these amino acid residues for TRPM function (Figure 1E). The two mutations were not found in an analysis of 100 Japanese normal controls. None of the five mutations mentioned above have been reported in a SNP database.

Analysis of protein expression from TRPM1 mutant minigenes

Our previous expression analysis of mouse TRPM1 showed that the long form (TRPM1-L), which functions as a cation channel, is predominantly expressed in the retina [11], suggesting that human patient samples for directly analyzing splicing are not readily available. To overcome this limitation, minigenes for expression in tissue culture cells were designed; these minigenes included the upstream exons (exon 2 or 8) fused with EGFP, downstream exons (exon 3 or 9) fused with 3×Flag-tags, and a HBB intron, which is often used in splicing assays (an exon 2+3 minigene and an exon 8+9 minigene; Figure 2A). Since the signals required for constitutive splicing are usually within ~100 nucleotides of splice junctions [11], we took ~200 nucleotide regions for both splicing donors and acceptors to ensure the splicing assay. We also generated mutated minigenes containing the IVS2–3C>G mutation in the exon 2+3 minigene or the IVS8+3_6delAAGT mutation in the exon 8+9 minigene. Minigene splicing was analyzed in HEK293T cells since they have been used extensively to study constitutive RNA splicing. HEK293T cells were transfected with minigene DNA and a lacZ expression internal control vector, and western blotting was performed on whole cell extracts isolated 48 h later. We then analyzed protein products of the EGFP-exon 2+3 fusion and the EGFP-exon 8+9 fusion proteins by western blot analysis of whole cell extracts transfected with the WT and mutated minigenes. We detected an expected 45 kDa or 39 kDa band with the anti-Flag antibody in HEK293T cell extracts transfected with the WT exon 2+3 minigene or the exon 8+9 minigene (Figure 2B,C). In contrast, we did not detect any bands in either of the cell extracts transfected with the mutated minigenes. We detected the β-gal bands of the transfection internal control, and the β-actin bands of the cell extract control in both the WT and mutant minigene transfected cells. These results showed that both the IVS2–3C>G and IVS8+3_6delAAGT mutations abrogate normal splicing and lead to abnormal protein production, suggesting that these trpm1 alleles are loss-of-function alleles.

In vivo analysis of intracellular localization of TRPM1 protein carrying missense mutations

The TRPM1 missense mutations R624C and F1075S are located in the N-terminal and the C-terminal end of the six-transmembrane domains (6TM), respectively (Figure 1D). Mutations in these regions often cause misfolding/mislocalization of the protein, and/or a defect in channel activity [28-32]. To determine the functional relevance of R624C and F1075S TRPM1 residues, we mutagenized the TRPM1 cDNA to introduce R624C and F1075S mutations. We first prepared expression vectors encoding R624C or F1075S mutant proteins attached with a Flag-tag at the C-terminus, and expressed them both, as well as WT vectors, in HEK293T cells. We did not observe a significant difference in the protein level between the mutants and WT by western blotting, suggesting that the protein stability of R624C and F1075S was not affected (data not shown). We then prepared expression vectors of R624C- and F1075S-TRPM1 cDNAs fused with the mouse mGluR6 9.5 kb promoter that drives ON BC-specific expression (Figure 3A). We electroporated each of them as well as the WT vector into the P0 mouse retinas. We examined the localization of the expressed TRPM1 mutant proteins in ON BCs at P14 when the endogenous mouse TRPM1 is preferentially localized to the dendritic tips [11,19]. The electroporated human TRPM1 was visualized by an anti-Flag tag antibody. TRPM1 WT was localized at the soma (arrow) and dendritic tips (arrowheads) of the ON BCs (Figure 3B,C). On the other hand, the immunostaining signals of both the R624C and the F1075S TRPM1 proteins were significantly weaker than those of the WT protein (Figure 3D-G). We quantified the number of electroporated ON BCs in which the WT TRPM1 protein was localized more in the dendritic tips than somas. About 70% of ON BCs that electroporated with WT TRPM1 showed a brighter TRPM1 signal at the dendritic tips, and the signals on the dendritic tips were dramatically reduced in both mutant proteins (Figure 3H). These results may imply that R624C and F1075S mutations lead to mislocalization of TRPM1 protein and are responsible for the protein trafficking of TRPM1 channel proteins. However, it should be noted that there is also a possibility that the transfected TRPM1 proteins are misfolded due to the addition of the 3xFlag tag, leading to the undetectability of Flag-tagged mutant proteins specifically in the dendrites. Future analysis is required to obtain a conclusive result on this point.

Clinical characteristics of the complete CSNB patients associated with TRPM1 mutations

The clinical characteristics of the three complete CSNB patients with TRPM1 mutations are summarized in Table 2. All the patients showed the typical clinical features of complete CSNB. They complained of night blindness from their early childhood, and the best-corrected visual acuities were normal in two patients (#373 and #437) and mildly reduced in one (#484). The refractive errors were highly or moderately myopic, and astigmatism was present in all of the patients (Table 2). Mild nystagmus and esotropia was observed in the patient with reduced visual acuity (#484). Slit lamp examination revealed no abnormality, and all fundus examinations revealed no abnormalities in the posterior pole other than myopic changes, including the tilted discs and temporal pallor of the optic discs (Figure 4). Goldmann kinetic visual fields and color vision tests with the Ishihara plates, the AO H-R-R pseudoisochromatic plates, and Farnsworth D-15 panel test were conducted in two patients (#437 and #484) with normal results. No family history was obtained from the patients (Figure 1B).

The full-field scotopic (rod) ERGs elicited by a blue light were non-recordable after 20 min of dark-adaptation (Figure 5B). The photopic ERG showed an a-wave with apparently normal implicit time and a b-wave with delayed implicit time (Figure 5C). The amplitudes of the 30 Hz flicker ERG were within normal range (Figure 5D). The single bright-flash, mixed rod-cone ERG elicited by a white stimulus had a “negative-type” configuration with a reduced normal a-wave and with a significantly reduced b-wave amplitude (Figure 5A). The oscillatory potentials were reduced.

ERG analysis of the trpm1 null mutant (trpm1^–/–) mice

To address a possible role of TRPM1 in ON BC function, we previously generated trpm1 null mutant (trpm1^–/–) mice by targeted gene disruption [11]. Next, to compare the physiologic retinal function of trpm1^–/– between humans and mice, we recorded mouse ERGs in 8-week-old mice (Figure 6). The dark-adapted (rod) ERGs elicited by different stimulus intensities from WT and trpm1^–/– mice are shown in Figure 6A. The amplitudes of the a-wave, originating from the photoreceptors, were approximately equal for both types of mice (Figure 6B). In contrast, the dark-adapted ERG b-wave, originating mainly from the ON BCs, in trpm1^–/– mice was not present at the lower intensities, and only a very small positive inflection was seen at the highest stimulus intensity of 1.0 log cd-s/m² (Figure 6C). The ERGs elicited by different stimulus intensities from light-adapted WT and trpm1^–/– mice are shown in Figure 6D. The b-wave was severely depressed or absent leaving only the a-wave in the trpm1^–/– mouse (Figure 6E,F). These ERG results are similar to those obtained by other groups [19,33], and suggested that the function of both rod and cone ON BCs, and the signal transmission from rod and cone photoreceptors to BCs were severely impaired in trpm1^–/– mice. In contrast, the presence of a normal a-wave indicated that the rod and cone photoreceptors were functioning normally in this mutant mouse.