Dysregulation of Ca_v1.4 channels disrupts the maturation of photoreceptor synaptic ribbons in congenital stationary night blindness type 2

Abstract

Mutations in the gene encoding Ca_v1.4, CACNA1F_, are associated with visual disorders including X-linked incomplete congenital stationary night blindness type 2 (CSNB2). In mice lacking Ca_v1.4 channels, there are defects in the development of “ribbon” synapses formed between photoreceptors (PRs) and second-order neurons. However, many CSNB2 mutations disrupt the function rather than expression of Ca_v1.4 channels. Whether defects in PR synapse development due to altered Ca_v1.4 function are common features contributing to the pathogenesis of CSNB2 is unknown. To resolve this issue, we profiled changes in the subcellular distribution of Ca_v1.4 channels and synapse morphology during development in wild-type (WT) mice and mouse models of CSNB2. Using Ca_v1.4-selective antibodies, we found that Ca_v1.4 channels associate with ribbon precursors early in development and are concentrated at both rod and cone PR synapses in the mature retina. In mouse models of CSNB2 in which the voltage-dependence of Ca_v1.4 activation is either enhanced (Ca_v1.4_I756T) or inhibited (CaBP4 KO), the initial stages of PR synaptic ribbon formation are largely unaffected. However, after postnatal day 13, many PR ribbons retain the immature morphology. This synaptic abnormality corresponds in severity to the defect in synaptic transmission in the adult mutant mice, suggesting that lack of sufficient mature synapses contributes to vision impairment in Ca_v1.4_I756T and CaBP4 KO mice. Our results demonstrate the importance of proper Ca_v1.4 function for efficient PR synapse maturation, and that dysregulation of Ca_v1.4 channels in CSNB2 may have synaptopathic consequences.

Introduction

In the retina, light-dependent changes in the photoreceptor (PR) membrane potential modulate the opening of presynaptic voltage-gated Ca_v1 (L-type) Ca²⁺ channels– a process that shapes Ca²⁺-dependent exocytosis of glutamate at the first synapse in the visual pathway. Of multiple classes of Ca_v1 channels (Ca_v1.1–1.4), Ca_v1.4 is thought to be the major Ca_v channel at PR synapses. More than 50 mutations in the CACNA1F gene encoding the pore-forming α₁ subunit of Ca_v1.4 are linked to multiple visual disorders including incomplete congenital stationary night blindness type 2 (CSNB2).¹ The visual phenotypes associated with CSNB2 are heterogeneous and may include abnormal visual acuity, night blindness, myopia, and/or nystagmus.²

Despite strongly reduced transmission by rod and/or cone PRs, vision impairment is relatively modest in CSNB2 patients.^3-5 It has been suggested that Ca_v1.4 may not predominate at cone PR synapses,⁶ which is at odds with the lack of cone and rod PR synaptic transmission in Ca_v1.4 knockout (KO) mice.⁷ A confounding factor is that in addition to their role in regulating PR synaptic transmission, Ca_v1.4 channels are required for the development of both rod and cone PR synapses in the mouse.^8,9 The mature “ribbon” structure that is specialized for high-throughput and tonic exocytosis¹⁰ is absent in Ca_v1.4 KO mice,^8,9 which may contribute to the complete absence of retinal output in these mice.⁷ It is not known if CSNB2-associated mutations that alter the function rather than the expression of Ca_v1.4 channels similarly affect PR synapses.

In the mouse retina, rod PR synaptogenesis occurs postnatally and involves formation of dyadic synapses between PRs and horizontal cell dendrites around postnatal day 8 (P8); invagination of postsynaptic bipolar dendrites after eye opening (~P13) completes maturation of the PR synaptic triad.¹¹ A hallmark of PR synapse development is the assembly of the ribbon, which originates as small round spheres that gradually take on the horseshoe-shaped morphology of the mature synapse.¹² Presynaptic abnormalities have been evaluated only after P10 in Ca_v1.4 KO mice,^8,9 so whether Ca_v1.4 channels are required for synapse formation and/or maintenance is unknown. Moreover, it is not clear if alterations in PR synapse development are common features of CSNB2 that can contribute to the pathogenic mechanisms.

To clarify these ambiguities, we defined the subcellular localization of Ca_v1.4 in the developing and mature retina, and compared the development of PR synaptic ribbons in WT mice and in mouse models of CSNB2 in which the voltage-dependence of activation of Ca_v1.4 is either enhanced or inhibited. Our findings indicate that Ca_v1.4 channels are positioned at the developing and mature active zone of PR synapses, where their presence is required for the initial stages of ribbon assembly. Moreover, the proper function of Ca_v1.4 channels is required for the efficient maturation of the synaptic ribbon, which is functionally disrupted in different mouse models of CSNB2.

Results

Polyclonal antibodies are selective for Ca_v1.4 and report the localization of the channel at rod and cone PR synapses in the retina

To characterize the subcellular localization of Ca_v1.4 during development in the mouse retina, we generated polyclonal antibodies against the cytoplasmic N-terminal domain of the Ca_v1.4 α₁ subunit. Since both Ca_v1.3 and Ca_v1.2 are thought to be expressed in cone PRs¹³ and bipolar neurons,¹⁴ respectively, we confirmed that the antibodies detected Ca_v1.4, but not Ca_v1.2 or Ca_v1.3, by western blot and immunofluorescence of transfected HEK293T cells (Fig. 1A and B). To determine the specificity of the antibodies for detecting Ca_v1.4 in the retina, we performed western blotting of retinal lysates from wild-type (WT) mice and mice lacking expression of functional Ca_v1.4 channels (Ca_v1.4 KO) due to excision of exons 14 to 17 of the mouse CACNA1F gene.¹⁵ In lysate from WT but not Ca_v1.4 KO retina (Fig. 1C), the antibodies detected a ~200–250 kDa protein that corresponded to the predicted molecular weight of the Ca_v1.4 α₁ subunit. A protein of higher molecular weight was also detected, which may reflect the reduced electrophoretic mobility of aggregated Ca_v1.4 protein upon denaturation, as such high-molecular bands were also detected in HEK293T cells transfected with Ca_v1.4 (Fig. 1A). Additional lower-molecular weight proteins were also detected, but were considered non-specific in that they were seen in both genotypes.

Figure 1. Rabbit polyclonal antibodies specifically recognize Ca_v1.4 in transfected cells and in mouse retina. (A and B) western blot (A) and immunofluorescence (B) of HEK293T cells that were untransfected (U) or transfected with Ca_v1.2, Ca_v1.3, or Ca_v1.4. In (Aand B), antibodies against Ca_v1.2 (top panels), Ca_v1.3 (middle panels), or Ca_v1.4 (bottom panels) were used. (C) Western blot of retinal lysate from WT and Ca_v1.4 KO mice (50 μg protein/lane) probed with Ca_v1.4 antibodies. Astericks indicate proteins present in WT but not Ca_v1.4 KO retina. (D) Immunofluorescent labeling with Ca_v1.4 and RIBEYE antibodies in retina from WT and Ca_v1.4 KO mice. Nuclei are stained with Hoechst. Lower panels show high magnification images of double-labeling in the outer plexiform layer (OPL). ONL, outer nuclear layer; INL, inner nuclear layer. Scale bars: 2 μm (B); 10 μm, (D, upper panels); 2 μm (D, lower panels). Results shown are representative of at least 3 independent experiments.

By immunofluorescence of WT mice, Ca_v1.4 antibodies labeled numerous horseshoe-shaped structures in the outer plexiform layer (OPL), which contains mostly rod PR terminals (spherules) in the mouse retina (Fig. 1D). These structures were also labeled with antibodies against RIBEYE, the major component of the ribbon,¹⁶ and resembled synapses formed between a single rod spherule and dendrites of bipolar neurons and horizontal cells.¹¹ As in previous reports that used different Ca_v1.4 antibodies,¹⁵ Ca_v1.4 labeling was slightly shifted compared with that for RIBEYE (Fig. 1D), consistent with the localization of Ca_v1.4 at the arciform density, a subsynaptic structure adjacent to the ribbon.^10,17 This pattern of labeling was absent in the OPL of Ca_v1.4 KO retina (Fig. 1D). Using fluorescently tagged peanut agglutinin (PNA) to label cones, we also observed Ca_v1.4 labeling in cone terminals (pedicles) in mouse retina (Fig. 2A). In macaque and human retinas, Ca_v1.4 antibody labeling closely aligned with that for PNA, and was associated with horseshoe-shaped structures resembling rod synapses throughout the OPL (Fig. 2B and C). Taken together, these results confirm the specificity of our antibodies, and show that Ca_v1.4 channels are localized to both rod and cone PR synapses in the mammalian retina.

Figure 2. Ca_v1.4 antibodies label cone PR synapses. Double-labeling with antibodies against Ca_v1.4 (green) and PNA (red) in retina from mouse (A), macaque (B), and human (C). In (A and B), arrows and arrowheads indicate localization of PNA and Ca_v1.4, respectively, at elongated structures resembling cone synapses. In (C), arrowheads indicate Ca_v1.4 labeling clustered at PNA-labeled cone synapses. Inset in (C) shows high magnification view of boxed region. Scale bars: 2 μm. Results in (A) are representative of at least 3 independent experiments. Results in (B and C) are from 1 experiment.

Mature rod PR synapses fail to form in Ca_v1.4 KO mice

Using these Ca_v1.4-selective antibodies, we profiled the distribution of Ca_v1.4 at developing PR synapses from P5-P15 in the mouse retina. From P5-P8, labeling for Ca_v1.4 and RIBEYE was largely colocalized in small puncta resembling ribbon precursor spheres^11,12 in the OPL (Fig. 3A). Between P11-P15, Ca_v1.4-labeled ribbons become more elongated, and depending on the plane of section, take on the horseshoe-shaped morphology of mature rod PR synapses (Fig. 3A).¹¹ The temporal and spatial coincidence of Ca_v1.4 and RIBEYE labeling in the OPL is consistent with a role for Ca_v1.4 channels in directing PR synapse assembly.

Figure 3. Ca_v1.4 is required for synapse formation. (A) Immunofluorescence for RIBEYE (green) and Ca_v1.4 (red) in retina from WT mice (P5-P15). (B) Immunofluorescence for RIBEYE (green) and GluR2 in retina from WT (left) and Ca_v1.4 KO (right) mice (P5-P15). Mature (arrows) and immature (arrowheads) ribbon morphologies are indicated. Scale bars: 2 μm. Results shown are representative of at least 3 independent experiments.

To determine if the synaptic defects in Ca_v1.4 KO mice^7,9,15 result from alterations in PR synapse formation or maintenance, we double-labeled with antibodies against RIBEYE and a glutamate receptor (GluR2) expressed in postsynaptic horizontal cell dendrites.¹⁸ In the OPL of WT retina, morphologically mature synapses are seen as early as P11 and predominate at P15 (Fig. 3B). In Ca_v1.4 KO retina, mature ribbons were not found; only RIBEYE-positive spheres were present, none of which were associated with GluR2-labeling at any age that was examined (Fig. 3B). The amount of RIBEYE labeling declines slightly beginning at P15 and is most apparent in adult Ca_v1.4 KO mice (compare Figs. 3B and 1D). These results demonstrate that Ca_v1.4 channels are required for the initial formation of PR synapses.

PR synapse maturation is impaired in mouse models of CSNB2

CaBP4 is a Ca²⁺ binding protein that interacts with, and enhances voltage-dependent activation of, Ca_v1.4 channels.^19,20 Electroretinograms (ERGs) of CaBP4 KO mice indicate a loss of both rod and cone PR synaptic transmission,^19,21 consistent with reduced function of Ca_v1.4 channels and the CSNB2-like phenotypes associated with human mutations in the Cabp4 gene.^5,22,23 PR synapses are morphologically abnormal in CaBP4 KO mice,¹⁹ which may be due to aberrant ribbon synapse assembly and/or maturation. To probe the underlying mechanism, we analyzed developmental changes in PR ribbon morphologies in WT and CaBP4 KO mice. Compared with age-matched WT control retina, ribbons were similar in CaBP4 KO retina through P13 (Fig. 4A and B). At this age, a number of elongated and horseshoe-shaped ribbons were seen in both genotypes, which indicated no deficits in the initial stages of synapse assembly. However at P15, when all ribbons take on elongated and horseshoe-shaped morphologies in WT retina, RIBEYE-positive structures that were small and round outnumbered those resembling mature ribbons in CaBP4 KO retina (Fig. 4A and B). Double-labeling with Ca_v1.4 antibodies revealed the association of Ca_v1.4 channels with both normal and abnormal ribbons, which indicated that alterations in ribbon structure did not result from failures in Ca_v1.4 expression or trafficking to synapses. These findings supported a requirement for CaBP4, perhaps via regulation of Ca_v1.4 channels, in the maturation of rod PR ribbons.

Figure 4. The maturation of synaptic ribbons is impaired in CaBP4 KO and Ca_v1.4_I756T mouse retina. (A–C) Immunofluorescence for RIBEYE (red) and Ca_v1.4 (green) in retina from WT (A), CaBP4 KO (B), and Ca_v1.4_I756T (C) mice (P5-P15). Mature (arrows) and immature (arrowheads) ribbon morphologies are indicated. Scale bars: 2 μm. (D) Quantitative analysis of ribbon morphologies in retinal sections that were single-labeled with RIBEYE antibodies. RIBEYE-positive profiles were categorized according to morphology as described in the Materials and Methods and presented as the mean ± SEM. The total number of profiles examined was 1003 (WT), 1073 (Ca_v1.4_I756T), 1140 (CaBP4 KO), and 1118 (Ca_v1.4 KO) in 5 animals per genotype; the retinas from each were processed in independent experiments. *, p < 0.001 compared with WT by ANOVA; #, p < 0.05 compared with CaBP4 KO by Student Newman-Keuls test.

To confirm the importance of proper Ca_v1.4 function for ribbon synapse development, we utilized knock-in mice harboring a mutation that causes CSNB2 in humans (Ca_v1.4_I756T). This mutation (I745T in the human CACNA1F gene) results in a large negative shift in the voltage-dependence of channel activation,²⁴ and so should potentiate Ca_v1.4 function, in contrast to the inhibition of Ca_v1.4 that would be expected in CaBP4 KO mice. As in CaBP4 KO mice, the pattern of RIBEYE/Ca_v1.4 labeling was similar in WT and Ca_v1.4_I756T mice through P13, but diverged at P15, with many ribbons exhibiting the immature morphology in the mutant but not the WT mice (Fig. 4A and C). Thus, both the loss- and gain-of function in Ca_v1.4 channel properties can impair PR ribbon development in the mouse retina.

To quantitate the severity of the synaptic defects in the various adult mutant mice, we compared the distribution of RIBEYE-labeled profiles as either circular, ellipsoid, or elongated. We assumed that circular profiles generally represent immature ribbon precursors, as nearly all (99.8 ± 0.1%) RIBEYE-positive structures were circular in Ca_v1.4 KO retina (Fig. 4D). However, some mature elongated ribbons could appear circular or ellipsoid, depending on the plane of section. Indeed, a minor fraction of RIBEYE labeling in WT mice was associated with circular (14.8 ± 0.94%) or ellipsoid (8.0 ± 2.9%) profiles. Unlike in WT mice, circular profiles accounted for most of the RIBEYE labeling in CaBP4 KO (97.0 ± 0.9%) and Ca_v1.4_I756T (84.9 ± 0.2%) (Fig. 4D). Notably, the proportion of elongated and ellipsoid profiles was significantly greater (p < 0.001 by ANOVA) in Ca_v1.4_I756T (2.6 ± 0.5% elongated, 12.6 ± 1.5% ellipsoid) than in CaBP4 KO (0.4 ± 0.3% elongated, 2.6 ± 0.7% ellipsoid; p < 0.001 by ANOVA) mice (Fig. 4D). Taken together, these results suggested a defect in PR ribbons of the following rank order in severity: Ca_v1.4 KO ≥ CaBP4 KO > Ca_v1.4_I756T (Fig. 4D).

Structural remodeling of horizontal and bipolar cell dendrites and the formation of ectopic synapses in the outer nuclear layer (ONL) are common features of mice with PR synapse defects.²⁵ If postsynaptic modeling results from altered PR ribbon maturation in CaBP4 KO and Ca_v1.4_I756T mice, it should be evident between P13 and P15, when presynaptic defects in these mice are first apparent (Fig. 4B and C). To test this, we analyzed retinas from P8 to adulthood that were double-labeled with RIBEYE and PKCα, a marker for rod bipolar cells. As shown previously,^7,19 bipolar dendrites extending into the ONL were found in adult retina of Ca_v1.4 KO and CaBP4 KO mice. RIBEYE-labeled puncta were associated with these processes, suggestive of ectopic ribbon synapses (Fig. 5). This pattern of labeling emerged earlier in Ca_v1.4 KO (~P11) than in CaBP4 KO and Ca_v1.4_I756T retinas (~P13, Fig. 5), consistent with the more severe presynaptic defects in the Ca_v1.4 KO mice (Figs. 3B and 4D).

Figure 5. Postsynaptic remodeling and ectopic ribbons in mice with altered Ca_v1.4 function. Double-label immunofluorescence of PKCα (green) and RIBEYE (red) in retina from WT, Ca_v1.4 KO, CaBP4 KO, or Ca_v1.4_I756T mice (P8-adult). In WT retina, bipolar dendrites labeled with PKCα antibodies ramify in the OPL where they form synapses with rod spherules. In Ca_v1.4 KO mice, bipolar dendrites (arrows) start extending into the outer nuclear layer (ONL) at P11, and ectopic ribbons labeled by RIBEYE antibodies are evident (arrowheads). Postsynaptic remodeling emerges later in CaBP4 KO and Ca_v1.4_I756T mice (~P13), and is less severe in adulthood (2-mo old) compared with WT mice. Scale bars: 2 μm. Results shown are representative of at least 3 independent experiments.

To determine if the impaired ribbon maturation in the mutant mice corresponded to deficits in PR synaptic transmission, we performed ERG analyses, which monitor light-induced changes in the electrical activity of presynaptic PRs (a-wave) and postsynaptic second-order neurons (b-waves) (Fig. 6A). Previous ERG analyses established strongly reduced b-waves in CaBP4 KO mice under dark-adapted conditions to measure rod PR transmission.¹⁹ Therefore, we restricted analyses to comparisons of Ca_v1.4_I756T mice with WT mice. Due to the presence of rod- and cone-driven responses in WT mice, plots of b-wave amplitudes against flash intensity were fit with a double-sigmoidal function (Fig. 6B). In contrast, the corresponding data in Ca_v1.4_I756T mice were described by a single sigmoidal fit (Fig. 6B). Since ERG recordings under light-adapted conditions indicated that Ca_v1.4_I756T mice lack cone-driven responses (Fig. 6A), we assumed that the b-wave responses from dark-adapted Ca_v1.4_I756T mice are largely due to rod transmission. Compared with rod-driven responses in WT mice, b-waves were significantly reduced, although not abolished, in Ca_v1.4_I756T mice (~66–74% compared with WT, at light intensities (I) < -1.0 log cd•s/m², p < 0.001 by ANOVA; Fig. 6B). In addition, the sensitivity of rod-rod bipolar synaptic transmission in Ca_v1.4_I756T mice (I_1/2) was significantly reduced (I_1/2 = -3.2 ± 0.2 log cd•s /m² for WT vs. -0.3 ± 0.6 log cd s /m² for I_756T; p < 0.001 by t-test). While the maximal a-wave amplitudes appeared to be reduced in Ca_v1.4_I756T, the difference with WT was not significant (385.3 ± 14.2 for WT vs. 301.3 ± 48.2 for Ca_v1.4_I756T; p = 0.1 by t-test, Fig. 6C). There was also no significant difference in the flash intensities evoking the half-maximal a-wave response (I_1/2 = -0.9 ± 0.1 for WT vs. -0.5 ± 0.4 log cd•s/m² for Ca_v1.4_I756T; p = 0.2 by t-test). These results argued against the possibility that deficient b-wave responses in dark-adapted Ca_v1.4_I756T mice were due to decreased sensitivity or changes in the numbers of functional rod PRs. Moreover, at light intensities at which the a-waves were nearly identical in dark-adapted WT and Ca_v1.4_I756T mice (e.g., I < -1 log cd•s/m²), the b-wave was still significantly smaller in the latter (p < 0.001 by ANOVA; Fig. 6B and C). Taken together, these results strongly supported a major defect in PR transmission in Ca_v1.4_I756T mice. The less severe rod-driven responses in Ca_v1.4_I756T mice compared with CaBP4 KO and Ca_v1.4 KO mice^7,19 are consistent with the more moderate abnormalities in ribbon maturation in the former compared with the latter 2 mice (Figs. 4 and 5). Our results support a requirement for normal Ca_v1.4 function in the maturation of rod PR synaptic ribbons, which may impact the efficacy of synaptic transmission in the adult mice.

Figure 6. ERGs from WT and Ca_v1.4_I756T mice. (A) Representative voltage traces from flash ERG recordings in 5–6-week old dark- adapted or light-adapted WT and Ca_v1.4_I756T mice. Arrowheads indicate time of flash, numbers indicate flash intensities (cd•s /m²). (B and C) b-wave amplitudes (B) and a-wave amplitudes (C) measured from recordings of dark-adapted mice obtained as in (A). Points represent the mean ± SEM (n = 8; left and right eyes from 4 mice). Smooth lines represent fits from double (WT in B) or single sigmoidal functions.

Discussion

Localization of Ca_v1.4 at rod and cone PR synapses

Due to the notorious non-specificity of Ca_v channel antibodies, we rigorously characterized the Ca_v1.4 antibodies that we generated and used in this study (Fig. 1). While other antibodies generated against Ca_v1.4 were shown to label rod PR synapses,^7,15 our study is the first to document the localization of Ca_v1.4 in cone PR synapses (Fig. 2). Cones comprise < 3% of the PRs in the rodent retina,^26,27 which may explain the difficulties in detecting Ca_v1.4 colocalization with markers of cone PR synapses in the rat.²⁸ Ca_v1.3 expression has been reported in cone PRs in mouse and tree shrew retinas,^6,29 and modest morphological changes in PR ribbon morphology have been reported in mice lacking Ca_v1.3.³⁰ However, we propose that Ca_v1.4 channels are functionally the dominant Ca_v channels in cone PRs for the following reasons. First, we found strong labeling of cone PR terminals in multiple species (Fig. 2). Second, mice lacking Ca_v1.4 but not Ca_v1.3 expression exhibit severe defects in cone-mediated visual responses,^7,31 and a loss-of function mutation in the CACNA1D gene encoding Ca_v1.3 channels, causes cardiac arrhythmia and deafness, but no major visual deficits.^30,32 In contrast, a number of mutations in CACNA1F_, including the human I745T mutation,³³ produce severe defects in cone and rod responses in humans.^34,35 Therefore, Ca_v1.4 channels likely play an analogous role in regulating exocytosis at cone and rod PRs.

A dual role for Ca_v1.4 in PR ribbon assembly

The stages involved in the molecular and morphological development of PR ribbon synapses have been elegantly described.^11,12,17 Prior to P10, ribbon precursor spheres complete their molecular assembly with the addition of ribbon-associated proteins (e.g., RIBEYE, bassoon, and piccolo) and proteins of the arciform density (e.g., RIM2, munc13, Ca_v1.4). Between P10 and P14, ribbons are largely mature at the molecular level, but make the final morphological transition and become anchored to the plasma membrane.¹² The observation that PR ribbons develop normally through P13 in CaBP4 KO and Ca_v1.4_I756T mice (Fig. 4), but not in Ca_v1.4 KO mice (Fig. 3B), indicates that the maturation but not the initial assembly of the ribbon depends on the proper activation properties of the Ca_v1.4 channel.

Initially, Ca_v1.4 channels may be required to scaffold key presynaptic proteins such as PSD-95 and the plasma membrane Ca²⁺ ATPase, both of which are absent in developing PR terminals of Ca_v1.4-deficient mice.^8,36 At this stage, Ca_v1.4 channels are clustered in RIBEYE-positive puncta that likely represent immature ribbon precursors (Fig. 3A),¹² and so may not be positioned at the presynaptic membrane where their Ca²⁺ conducting function may be required. Upon eye-opening at P13, light-modulated changes in Ca_v1.4 opening may shape presynaptic Ca²⁺ signals and exocytosis which support efficient PR synapse maturation. These Ca²⁺ signals would be strongly reduced in CaBP4 KO mice, since there would be limited activation of Ca_v1.4 channels at the PR membrane potential in darkness (~-40 mV) in the absence of CaBP4.¹⁹ In Ca_v1.4_I756T mice, presynaptic Ca²⁺ influx in darkness should be abnormally elevated due to a ~-30 mV shift in the voltage-dependence of activation, and slower voltage-dependent inactivation, caused by this mutation.²⁴ The more severe morphological and functional defects in PR synapses in CaBP4 KO compared with Ca_v1.4_I756T mice (Figs. 4 and 6)¹⁹ may reflect greater synaptic consequences of the loss-of function compared with the gain-of function of Ca_v1.4, although we cannot discount the possibility that CaBP4 may have effects independent of Ca_v1.4 on promoting ribbon maturation. While our immunocytochemical data indicated no major differences in the intensity of Ca_v1.4 labeling at PR synapses in CaBP4 KO and Ca_v1.4_I756T mice at P15 (Fig. 4), we cannot rule out that there may be a progressive reduction in levels of Ca_v1.4 channels, which could contribute to the visual impairment measured in these mice at later ages (Fig. 6 and in Haeseleer et al.¹⁹). Ca_v1 channels may play a general role in the development of ribbon synapses, since proper function of Ca_v1.3 channels has been shown to regulate ribbon size and maintenance at zebrafish hair cell ribbon synapses.³⁷ Determining the molecular mechanisms by which Ca_v1.4 channels promote PR synapse assembly and maturation remains an important challenge for future studies.

Synapthopathic origins of CSNB2

Our findings that most (75–90%) RIBEYE-positive structures exhibited the immature (round) morphology in adult CaBP4 KO and Ca_v1.4_I756T mice (Fig. 4B) are consistent with the strong reductions in rod PR transmission measured in these mice¹⁹ (Fig. 6). Smaller, spherical ribbons are unlikely to support the tethering of hundreds of synaptic vesicles primed for tonic glutamate release at mature ribbon synapses in darkness.³⁸ Based on findings that alterations in ribbon structure significantly inhibit rod PR transmission,^39,40 weakened exocytosis at these immature synapses likely contributes to the almost complete absence of rod PR synaptic transmission in Ca_v1.4 KO (Fig. 6B) and CaBP4 KO¹⁹ mice. The greater preservation of ribbons in Ca_v1.4_I756T mice compared with Ca_v1.4 KO or CaBP4 KO mice (Fig. 4D) may account for the partial sparing of rod PR synaptic transmission in these mice (Fig. 6B). Since rod PRs mediate vision in low-light conditions, our findings may help explain the night-blindness phenotype in some patients with loss-of function mutations in CACNA1F¹ and Cabp4.^5,22 On the other hand, we found that cone-mediated (light-adapted) b-wave responses are absent in Ca_v1.4_I756T mice (Fig. 6A). Interestingly, patients harboring the analogous I745T mutation³³ as well as some patients with Cabp4 mutations^23,41 exhibit a more severe loss of vision mediated by cones than by rods. Considering the localization of Ca_v1.4 in cone PRs (Fig. 2), and previous results that cone PR synaptogenesis is impaired in Ca_v1.4 KO mice,⁹ we expect that a disruption in cone PR synaptic ribbon maturation may contribute to these visual phenotypes in humans.

Alterations in PR ribbon morphologies and synapses have been reported in mice with reduced or absent expression of Ca_v1.4.^7,42 Considering that a number of CSNB2 mutations in CACNA1F are predicted to yield dysfunctional channels rather than cause the complete absence of channel expression,^43,44 our findings reveal that alterations in PR synapse development may be a common feature of CSNB2, and that therapeutic interventions for offsetting visual impairment may require early targeting of the synaptopathic consequences of the mutations.

Materials and Methods

Animals

CaBP4 KO¹⁹ and Ca_v1.4 KO¹⁵ mice were characterized previously. Ca_v1.4 KO (B6.Cg-Cacna1ftm1.1Sdie/J) and Ca_v1.4_I756T (B6(Cg)-Cacna1ftm1Sdie/J) were obtained from the Jackson Laboratory. The generation of the Ca_v1.4 KO and Ca_v1.4_I756T lines was described previously.¹⁵ Ca_v1.4 KO mice lack exons 14–17, while Ca_v1.4_I756T mice bear a threonine substitution for isoleucine 756, of the mouse CACNA1F gene. Male and female mice (5-d -3 mo old) were used. All animals were maintained on a 12-h light/dark cycle. Experimental procedures using animals were approved by the Institutional Animal Care and Use Committee at the University of Iowa and the University of Washington.

Antibody production and characterization

Polyclonal rabbit Ca_v1.4 antibodies were generated against a peptide corresponding to amino acids 1–20 (MSESEVGKDTTPEPSPANGTC) of mouse CACNA1F (NP_062528.2) by a commercial source (Covance). Antiserum was subject to affinity purification by standard protocols prior to use.

Immunofluorescence and western blotting of transfected HEK293T cells

Human embryonic kidney (HEK) 293T cells were transfected with cDNAs encoding the α₁ subunit for Ca_v1.2, Ca_v1.3, Ca_v1.4 along with β_2A and α₂δ using GenePORTER transfection reagent (# T201015, Genlantis). After 24 h, transfected cells were processed the next day for immunofluorescence essentially as described previously.⁴⁵ Rabbit polyclonal primary antibodies were used at 1:1000 dilution: anti-Ca_v1.2,⁴⁶ anti-Ca_v1.3,⁴⁵ anti- Ca_v1.4 (this study). Alexa488-conjugated and Alexa568-conjugated secondary antibodies (1:1000; #A11011, Life Technologies) were used. Image analysis was performed using a Fluoview 1000 confocal microscope (Olympus). For western blots, lysates from transfected cells or retina (from 2-mo old mice) were prepared and subjected to SDS-PAGE and western blotting as described.⁴⁵ Primary antibodies were the same as those used for immunofluorescence experiments. Secondary detection was with HRP-conjugated anti-rabbit IgG (1:1000; #RPN4301, GE Healthcare) and SuperSignal West Pico Chemiluminescent Substrate (# 34077, Thermo Scientific).

Immunofluorescence of retina

Immunofluorescence processing of the retina was generally as described previously.¹⁹ Macaque retina was obtained at the University of Washington Regional Primate Center and frozen unfixed human retina was obtained from the Iowa Lions Eye Bank (provided by Dr Robert Mullins) following full consent of the donors' next of kin. Mice younger than P10 were killed by decapitation, and mice at P10 or older were anesthetized with isoflurane first and killed by decapitation. Three mice (male or female) per developmental stage were used for each genotype. Mice were allocated to groups according to age or genotype prior to experimentation. Eyes were quickly removed and incubated in 4% paraformaldehyde in 0.1 M phosphate buffer for 10–60 min. The anterior segments were removed, and the posterior eye cups were infiltrated with 30% sucrose on ice and frozen in embedding media on dry ice. Frozen blocks were cut in vertical sections (12 μm) with a cryostat (Leica), collected on electrostatic slides, and stored at -80°C until use. Sections were blocked in blocking buffer (3–10% normal goat serum [NGS] and 0.1% Triton X-100 (TX-100) diluted in phosphate-buffered saline [PBS]) overnight at 4 °C. Sections were incubated in primary antibodies for 4 h at room temperature. The following primary antibodies and working dilutions were used: rabbit Ca_v1.4 (1:1000), RIBEYE (1: 500–1:1000 #612044, BD bioscience), PKC-α (1:500, #SC208, Santa Cruz, Biotechnology). After rinsing, sections were incubated for 1 h in darkness in secondary antibodies (Life Technologies) used at 1:400 or 1:1000 dilution: Alexa Fluor 488- goat anti-mouse (#R37120); Alexa 555- goat anti-rabbit (# A31629); Alexa 488- goat anti-rabbit (#A11008); Alexa568- goat anti-mouse (#A11004). All antibodies were diluted in 0.1% TX-100 in PBS and incubations were performed in humidified chambers. In some experiments, Hoechst stain (# H6024, Sigma; 1:1000) was applied for 10 min at room temperature. Between incubations, sections were washed 3 times for 5 min using 0.1% TX-100 in PBS. Sections were coverslipped, sealed with clear nail polish, and stored at 4 °C. Confocal microscopy was performed using an Olympus Fluoview 1000 confocal microscope with 60X or 100X oil-immersion objectives or a Zeiss LSM710 confocal microscope with a Plan-Neofluar 63x/1.4 oil-immersion objective. Previous experience with variabilities in histological techniques was used to determine the number of animals (at least 3 per developmental age per genotype) required to substantiate conclusions. Samples with poor preservation of tissue morphology, as indicated by transmitted light microscopy or Hoechst staining, were excluded from analysis. For developmental series, processing of tissue from a given mouse strain was done in a single experiment to reduce variability due to differences in experimental conditions. Results shown in all figures are representative of at least 3 independent experiments.

For quantitative analyses of RIBEYE-labeled structures in the adult mouse retina, confocal Z-stack images (5 μm thick) were taken about 1 mm from the optic nerve (2 images/mouse for each of 3 mice per genotype). For elongated structures the length was measured with three segments along the ribbon. Ratios of length to width were calculated and categorized accordingly: spherical (ratio < 2), ellipsoid (2 to 3), elongated (> 3). To reduce experimenter bias, image acquisition and quantitative analyses were done by different researchers in a blinded fashion.

Electroretinography (ERG)

ERG recordings were obtained using the Espion E³ system (Diagnosys LLC). After overnight dark adaptation, 5- to 6-weeks old mice were prepared for ERG recording under dim red light. Mice (5–6 week old, males and females) were anesthetized by intraperitoneal injection of a ketamine/xylazine mixture (100 and 10 mg/kg respectively). The pupils were dilated by applying a drop of 1% tropicamide, followed by a drop of 2.5% phenylephrine hydrochloride. ERGs were recorded simultaneously from the corneal surface of each eye using gold ring electrodes (Diagnosys), with a needle electrode placed on the back of the head serving as reference. Another needle electrode placed near the tail served as ground. A drop of Hypromellose 2.5% Ophthalmic Demulcent Solution (# 17478–0064–12, Akorn Gonak) was placed on the corneal surface to ensure electrical contact and to prevent eyes from drying and cataract formation. Body temperature of mice was maintained at 37°C using the system's heating pad. Mice were placed in a Ganzfeld stimulator chamber (ColorDome; Diagnosys) for delivery of stimuli, and the mouse head and electrode positioning were monitored on the camera attached to the system. ERG responses were evoked in mice by a series of flashes ranging from 0.0001 to 100 cd∙s/m². Responses to 6 sweeps were averaged for dim flashes up to 0.6 cd∙s/m², 2 sweeps were averaged for 4 cd∙s/m², and responses to brighter flashes were recorded without averaging. Intersweep intervals for flashes with increasing strength were increased from 10 to 60 s to allow full recovery from preceding flashes. To record photopic ERGs, mice were exposed to a background light (30 cd•s /m²) for 3 min before flash stimulation (3, 30, or 100 cd•s/m²). Six sweeps were averaged. The data were analyzed as described previously⁴⁷ using GraphPad Prism software (version 4).

10.4161/chan.26376