Male C57BL/6J mice at 2–3 months old were used in this study. The Cav1.3^−/− mice (C57BL/6J background) were originally developed by Dr. Jörg Striessnig (University of Innsbruck, Innrain, Innsbruck, Austria (Platzer et al., 2000). The Cav1.3^+/− (heterozygous) breeding pair for generating Cav1.3^−/− (homozygous knockout) were obtained from Dr. Amy Lee (University of Iowa, Iowa City, IA, USA). The Cav1.3^−/−, Cav1.3^+/− and Cav1.3^+/+ wild type (WT) littermates used in this study were produced at Texas A&M University (College Station, TX, USA). All animal experiments were approved by the Institutional Animal Care and Use Committee of Texas A&M University. Mice were housed under temperature and humidity controlled conditions with 12:12 h light-dark cycles.

Fresh porcine eyes were obtained from a local meat processing plant (K&C Meat Processing, Navasota, TX, USA). Retinas were collected and homogenized in 1 ml lysis buffer (1% NP-40). Samples were rotated at 4°C for 3 h to solubilize membrane proteins. Samples were then centrifuged at 14,000 g for 30 min at 4°C to remove cell debris, and a small portion of the supernatant was taken for protein (loading control, total ERK) analysis. The rest of the supernatant was pre-cleared with Protein A agarose (GBiosciences, Maryland Heights, MO, USA). The beads were removed and 5 μl of anti-RS1 antibody (Santa Cruz Biotechnology, Dallas, TX, USA) was added and incubated for 3 h. A kit (Pierce/Thermo Fisher Scientific, Waltham, MA, USA) was used to remove heavy and light chain interference. No antibody was added to the control. After antibody incubation, 20 μl Protein A agarose were added to each tube and incubated for another 1.5 h. The beads were collected and processed for Western blot analysis of Cav1.3 (antibody from Alomone, Jerusalem, Israel) and Cav1.4 (antibody generated in Amy Lee’s laboratory, University of Iowa, Iowa City, IA, USA). Western blots were visualized by appropriate secondary antibodies (Cell Signaling, Danvers, MA, USA) and electrochemiluminescence kits (Pierce/Thermo Fisher Scientific, Waltham, MA, USA). A commercially available kit (Bio-Rad, Hercules, CA, USA) following the Bradford method was also used to determine total protein content of the samples. All co-immunoprecipitations (co-IPs) were repeated three times.

For the mammalian two-hybrid assay, human full length Cav1.4 (Gene ID: 778; obtained from Amy Lee’s laboratory, University of Iowa, Iowa City, IA, USA) and N-terminus (1-1481bp from ATG) of human Cav1.4 were amplified by PCR (Platinum PCR SuperMix High Fidelity, Thermo Fisher Scientific). The primers for the N-terminal were 5′-aaagtcgactgtcggaatctgaaggcgggaaag-3′ (forward) and 5′-gcatctagaggttttcatgatcttgtttaggca-3′ (reverse). The PCR products were purified (Qiaquick, Qiagen, Germantown, MD, USA) and subcloned into the pGEM-T-easy vector (Promega, Madison, WI, USA) to confirm the sequence. The human full length gene encoding RS1 (Gene ID: 6247; 675 bp from ATG to TGA) was also amplified by PCR from the pCDNA-RS1 plasmid (Shi et al., 2009). The RS1 and Cav1.4 N-terminal were inserted into mammalian expression vectors pBind and pACT, respectively (Promega). The transfection reporter assay was carried out by a luciferase assay system (Promega). Briefly, three constructs, pBind or pBind-RS1, pACT or pACT-Cav1.4-N terminal, and pG5Luc (100 ng each) were cotransfected into Cos1 cells (TransIT^®-COS transfection kit, Mirus, Madison, WI, USA). After cells were harvested, 10 μl of the supernatant was mixed with luciferase substrate, and the relative luciferase activity was determined by luminosity (Perkin-Elmer, Waltham, MA, USA). All luciferase assays were repeated six times.

pCDNA-RS1 (Human), pCDNA-RS1(R141G), pCDNA-RS1(W92C) were generous gifts from Dr. Dorothy Trump, University of Manchester, Manchester, UK (Shi et al., 2009). pCAGIG-RS1s (WT, R141G, W92C) were constructed by inserting the RS1 encoding fragment from the pCDNA3.1 vector into pCAGIG (EcoRI). The calcium channel α2δ1 subunit (rat) expression vector was a generous gift from Dr. Terrance P. Snutch (University of British Columbia, Vancouver, BC, Canada). The pCDNA-Cav1.3 α1 subunit (mouse) and pCDNA-Cav1.4 α1 subunit (mouse) were generated in Amy Lee’s laboratory (University of Iowa, Iowa City, IA, USA). The pCMV-Sport-β2 subunit (mouse) was purchased from the MGC cDNA clones collection (Dharmacon, GE, Lafayette, CO, USA). The empty plasmid vector phrGFP containing green fluorescent protein (EGFP) was obtained from Agilent Technologies (Santa Clara, CA, USA). The plasmids were amplified in E. coli and purified with a kit (Qiagen). All plasmid sequences were confirmed by DNA sequencing (Gene Technologies Lab, Texas A&M University, College Station, TX, USA).

Human HEK 293 cells (American Type Culture Collection, ATCC, Manassas, VA, USA) were maintained in Dulbecco’s modified Eagle medium (Lonza, Portsmouth, NH, USA) containing 10% fetal bovine serum (Thermo Fisher Scientific, Waltham, MA, USA) and 100 μg/ml penicillin/100 μg/ml streptomycin (Life Technologies, Grand Island, NY, USA), 1mM sodium pyruvate (Life Technologies) and 1× non-essential amino acids (Life Technologies) at 37°C and 5% CO₂. The cells were plated and cultured on glass coverslips (12 mm diameter) at 70%–80% confluence 24 h before the transfection. Transfections were performed using a lipofectamine 2000 transfection reagent (Life Technologies) according to the manufacturer’s protocol. Up to 500 ng of DNA (100 ng for each plasmid) was transfected into the cells. All cells were co-transfected with EGFP. The culture media were replaced 12 h after transfections. The cells were recorded 48–60 h after transfections.

Whole-cell voltage-clamp recordings of LTCC currents were carried out on transfected cells that expressed GFP. The external solution for barium (Ba²⁺) carried LTCC currents contained the following (in mM): Tris 140, BaCl₂ 10, MgCl₂ 1 and glucose 5.6, pH 7.4 adjusted with HCl. The external solution for calcium (Ca²⁺) carried LTCC currents contained the following (in mM): Tris 140, CaCl₂ 10, MgCl₂ 1 and glucose 5.6, pH 7.4 adjusted with HCl. The pipette solution was (in mM): Cs acetate 135, CsCl 10, MgCl₂ 2, EGTA 1.1 and HEPES 10, pH 7.4 adjusted with CsOH. The transfected cells were visualized under a fluorescence microscope (IX71, Olympus America, Center Valley, PA, USA). Cells were recorded using a 200 ms step command with holding potential at −65 mV and steps from −80 mV to 60 mV at 10 mV increments. Currents were recorded at room temperature using a patch-clamp amplifier (Model 2400, A-M Systems, Carlsborg, WA, USA). Signals were low pass-filtered at 2 kHz and digitized at 5 kHz with a Digidata 1550A interface and pCLAMP 10.5 software (Axon Instruments/Molecular Devices, Union City, CA, USA). After gigaohm seals were formed, the electrode capacitance was compensated. The membrane capacitance, series resistance and input resistance of the recorded cells were measured by applying a 5 mV (100 ms) depolarizing voltage step from a holding potential of −65 mV. The membrane capacitance reading was used as the value for whole cell capacitance. The current density (pA/pF) was obtained by dividing the current amplitude (pA) by the membrane capacitance (pF). Currents were leak-subtracted after data acquisition. The conductance-membrane potential relationships were analyzed by fitting the Boltzmann equation: G/Gmax = 1/(1 + exp[Vmid − V/Ka]); G: conductance, V: membrane voltage, Vmid: the membrane potential that elicits half of the maximal activation (current), and Ka: the activation slope factor. The protocol used to determine calcium-induced inactivation (CDI) of LTCCs was based on Peterson et al. (1999). The r₃₀ is the ratio of remaining currents at the end of 30 ms voltage steps (I₃₀) against the peak maximal current (I_max) and is used to quantify the level of inactivation. The strength of CDI was further quantified by the parameter f, defined as the difference between r₃₀ values in Ba²⁺ vs. Ca²⁺ taken at −10 mV (Peterson et al., 1999; Mori et al., 2004; Tan et al., 2012). An f value of 0 indicates that there is no CDI, whereas the maximal f value of 1 indicates a complete CDI.

The frozen tissue blocks for sectioning containing Cav1.4 null and WT littermate mouse eyes that were fixed and cryo-protected (as described previously; Liu et al., 2013) were obtained from Amy Lee’s laboratory and sectioned at 12 μm thickness. The frozen RS1 null and control WT mouse retina sections (10 μm) were provided by Dr. Paul Sieving’s laboratory (National Eye Institute, Bethesda, MD, USA), and the processing of the retinal sections was described previously (Takada et al., 2004). The eyes excised from the Cav1.3^−/− and WT littermates were fixed in Zamboni fixative (American Matertech Scientific Inc, Lodi, CA, USA) and then cryo-protected in a 30% sucrose-phosphate-buffered saline (PBS) solution. Cav1.3^−/− and WT eyes were embedded side by side in Tissue-Tek O.C.T. Compound (Sakura Finetek Inc, Torrance, CA, USA) and stored at −80°C. The frozen eye sections (10 μm) were cut using a cryostat (Leica Biosystem, Buffalo Grove, IL, USA) and mounted on glass slides. After washes with 0.1 M sodium PBS (pH 7.4), the sections were incubated with a blocking solution containing 10% fetal bovine serum in PBS for 2 h at room temperature then incubated with the primary antibody at 4°C overnight. The next day, sections were washed three times with PBS containing 0.1% Triton 100 (PBST), incubated with a secondary antibody at room temperature for 2 h in a dark chamber, then washed with PBST, and mounted with ProLong Gold antifade reagents with 4′,6-diamidino-2-phenylindole (DAPI; Life Technologies). The primary antibodies used were rabbit anti-Cav1.3 (1:100; Alomone), mouse anti-Ribeye (1:100; EMD Millipore, Billerica, MA, USA), rabbit anti-Cav1.4 (1:1000; generated in Amy Lee’s laboratory) and rabbit anti-RS1 (1:100; Santa Cruz Biotechnology, Dallas, TX, USA). The secondary antibodies used were Alexa fluor 488 goat anti-rabbit IgG (1:200; Life Technologies), Cy5 goat anti-mouse IgG (1:200; Abcam, Cambridge, MA, USA) and Texas red donkey anti-goat IgG (1:200; Life Technologies). The images were captured with a Zeiss LSM 780 NLO Multiphoton Microscope (Carl Zeiss AG, Oberkochen, Germany), and the conditions (magnification and exposure time) of images taken from WT or mutant retinal sections for each antibody were identical.

All data are presented as mean ± standard error of the mean (SEM). Student’s t-test and one-way ANOVA followed by Tukey’s post hoc test for unbalanced n were used for statistical analyses. Throughout, p < 0.05 was regarded as significant.

Using porcine retinas, we took advantage of tissue abundance and determined the physical interaction between RS1 and Cav1.3 or Cav1.4 by co-IP. An antibody against RS1 (RS1 Ab) was able to pull down Cav1.3, as well as Cav1.4 (Figures 1A,B), thus providing the first evidence that RS1 interacts with mammalian Cav1.3 and Cav1.4. We next used mammalian two-hybrid (luciferase reporter) assays to determine the interaction between RS1 and Cav1.4. We previously showed that RS1 interacts with the first 500 amino acids from the N-terminal of chicken Cav1.3 containing the first (I) of four (I–IV) homologous motifs that is highly conserved with the human Cav1.3 (Shi et al., 2009). Using a similar strategy to determine the interaction between RS1 and Cav1.4, a full-length human rs1 cDNA (hRS1) was inserted into a pBIND vector that encoded a recombinant protein with GAL4 DNA binding domain as the bait. The human Cav1.4-N terminal fragment (hCav1.4-N; 500 amino acids) included the short N-terminus, the first motif (I), and a partial junction sequence between the first (I) and second (II) motifs. The hCav1.4-N was inserted into a pACT vector that encoded a protein containing a VP16 activation domain. The pG5Luc contained five tandem GAL4 binding sequences upstream of a luciferase coding region that was used to report protein interactions. The relative luciferase activity was at least four times higher in cells co-transfected with hRS1 and hCav1.4-N compared to the controls (Figure 1D), which supported our co-IP data. Thus, RS1 was able to interact with the first motif of Cav1.4 (Figure 1), as well as Cav1.3 shown previously (Shi et al., 2009). We next investigated the functional interaction between RS1 and Cav1.3 or Cav1.4.

To determine whether RS1 was able to regulate the Cav1.3-LTCC channel gating behavior, we co-transfected HEK-293T cells with both, and systematically examined how RS1 might modulate Cav1.3-LTCC gating behaviors in the absence or presence of its auxiliary subunits. We first investigated whether RS1 alone was able to elicit calcium influx through Cav1.3 without the auxiliary (β2 and α2δ1) subunits. In the absence of β2 subunit, Cav1.3-mediated currents were not measurable (Figure 2A, Cav1.3 + EGFP), and RS1 was not able to stimulate calcium influx via Cav1.3 (Figure 2A, Cav1.3 + RS1). To further decipher how RS1 might interact with Cav1.3 in the presence of various auxiliary subunits, cells were co-transfected with Cav1.3 and β2 (Cav1.3 + β2) or all auxiliary subunits (Ca1.3 + β2 + α2δ1) concurrently with or without RS1. While α2δ1 does not have any effect on the Cav1 voltage-dependence, this auxiliary subunit enhances the expression of Cav1 + β subunits (Shistik et al., 1995; Bangalore et al., 1996). As expected, α2δ1 mildly increased Cav 1.3-LTCC currents in the presence of the β2 subunit (Ca1.3 + β2 + α2δ1) compared to Cav1.3-LTCCs without α2δ1 (Ca1.3 + β2; Figures 2B,C). We found that co-transfection with RS1 significantly enhanced Cav1.3-LTCC currents in the presence of β2 (Cav1.3 + β2 + RS1; Figures 2B,C), but α2δ1 did not further increase Cav1.3-LTCC currents when RS1 was present (Ca1.3 + β2 + α2δ1 + RS1; Figures 2B,C). Hence, RS1 was able to enhance the Cav1.3-LTCC channel activation that required the presence of β2. The interaction between RS1 and Cav1.3-LTCCs was mainly limited on the α1 subunit (Cav1.3) without interacting with the other extracellular auxiliary subunits (α2δ1).

We next examined whether RS1 was able to enhance the voltage-dependent activation of LTCCs. Co-transfection with RS1 and functional Cav1.3-LTCCs (Cav1.3 + β2 + α2δ1 + RS1 or Cav1.3 + β2 + RS1) significantly facilitated the voltage-dependent activation compared to the Cav1.3-LTCCs without RS1 (Figures 2D,E). RS1 shifted the current-membrane voltage relationship of Cav1.3-LTCCs by −10 mV (Figures 2D,E). One biophysical characteristics of Cav1.3-LTCC is its Ca^{2 +}-dependent inactivation (CDI; Peterson et al., 1999; Mori et al., 2004; Tan et al., 2012). When the Cav1.3-LTCC currents are carried by Ca^{2 +}, but not Ba^{2 +}, Cav1.3-LTCCs display CDI (Figure 2F) with the “f” parameter calculated to quantify the strength of CDI (Peterson et al., 1999; Mori et al., 2004; Tan et al., 2012). We found that RS1 did not affect the CDI of Cav1.3-LTCCs in cells co-transfected with RS1, since there was no statistical difference of the f values between the Cav1.3-LTCCs with or without RS1 (Cav1.3 + β2 + α2δ1: 0.385 ± 0.076; Cav1.3 + β2 + α2δ1 + RS1: 0.402 ± 0.084). Since the CDI of Cav1.3-LTCCs depends on the binding of calcium-calmodulin at the C-terminal of the Cav1.3 (Peterson et al., 1999), these data indicate that RS1 did not interfere with the calcium-dependent structural changes. Through interacting with the first motif, RS1 was able to facilitate the voltage-dependent activation and augment the channel conductance of Cav1.3-LTCCs.

Both W92C and R141G are missense mutations of RS1 detected in XLRS patients (Wang et al., 2002, 2006). The W92C mutation causes cysteine-triggered intermolecular bonding that result in intracellular retention of the mutant RS1 in the endoplasmic reticulum (ER; Wang et al., 2006). The R141G mutation does not interfere with secretion but affects a surface residue within the loop region causing RS1 to lose its ability to bind other molecules (Wang et al., 2006). Neither W92C nor R141G had the ability to augment Cav1.3-LTCCs compared to the WT RS1 (Figures 2G–I). These observations provide evidence that RS1 interacts with Cav1.3-LTCCs and facilitates membrane retention of Cav1.3-LTCCs (Shi et al., 2009), and it further augments the channel voltage-dependent activation without interfering with its calcium-dependent inactivation.

Since the RS1 antibody was able to co-immunoprecipitate Cav1.4 from porcine retinas, and RS1 interacted with the first 500 amino acids of the Cav1.4 N-terminal sequence containing the first motif (Figure 1), we next examined the functional interaction between RS1 and Cav1.4. We applied a similar strategy to systematically examine the interactions between RS1, Cav1.4, and the auxiliary subunits of Cav1.4. Co-transfection with RS1 and Cav1.4 without the β2 subunit did not elicit measurable LTCC currents (Figure 3A). When co-transfected with the full length Cav1.4 and the axillary subunits (Cav1.4 + β2 + α2δ1), RS1 significantly enhanced the Cav1.4-LTCC current density (Figure 3) and the voltage-dependent activation (Figure 3B). Unlike Cav1.2- or Cav1.3-LTCCs, full length Cav1.4 without a deletion of exon 47 shows no discernable CDI (Baumann et al., 2004) and displays unusually slow voltage-dependent inactivation (McRory et al., 2004; Haeseleer et al., 2016). Co-transfection with RS1 and functional Cav1.4-LTCC did not alter the CDI property of Cav1.4-LTCCs (Figure 3C). Likewise, R141G or W92C, missense mutations of RS1, had no impact on functional Cav1.4-LTCCs (Figures 3D–F). Hence, RS1 was able to interact with Cav1.4-LTCCs, enhance its voltage-dependent activation, and augment the channel currents similar to its interaction with Cav1.3.

Mice lacking RS1 have degenerated retinas that resemble the human XLRS phenotype, which includes schisis cavities in the inner nuclear layer (INL), disorganized outer nuclear layer (ONL) and progressive photoreceptor degeneration (Weber et al., 2002). Immunostaining with an antibody against Ribeye, a marker for ribbon synapses (Schmitz et al., 2000), showed that the outer plexiform layer (OPL) was disorganized in the RS1^−/− retina (Figure 4A) indicating that RS1 deletion disrupted synaptic structure. The expressions of Cav1.3 and Cav1.4 were markedly decreased in the RS1^−/− retina (Figure 4A). Focusing on changes in photoreceptors, Cav1.3 is normally present in the IS, as well as the ONL and OPL (Figure 4B). In RS1^−/− retinas, Cav1.3 was decreased in all three areas (Figure 4B), which verified that degeneration of photoreceptors had taken place. Cav1.4 is mostly located in the photoreceptor synaptic terminals and is responsible for sustained neurotransmitter release from photoreceptors (Morgans, 2001; Liu et al., 2013; Lee et al., 2015). While Cav1.4 existed mainly at the OPL in the WT mouse retina (Figure 4B), it was decreased in the OPL of the RS1^−/− retina (Figure 4B). Therefore, deletion of RS1 caused losses of LTCCs (both Cav1.3 and Cav1.4), which could be the consequence of photoreceptor degeneration.

Patients with mutations in the gene encoding Cav1.4 have incomplete CSNB2 (Bech-Hansen et al., 1998; Strom et al., 1998; Zito et al., 2003; Michalakis et al., 2014). Since the retinal light responses recorded by electroretinogram (ERG) of patients with CSNB2 are similar to that of patients with XLRS (Bradshaw et al., 2004), and there was a physical interaction between RS1 and Cav1.4 (Figure 1), we next examined the distribution of RS1 in Cav1.4 null mutant (Cav1.4^−/−) mouse retinas. No Cav1.4 was detected in the OPL of Cav1.4^−/− mouse retinas (Figure 5A), and Ribeye was decreased in the OPL of Cav1.4^−/− compared to the WT (Figure 5A). These data confirmed a previous report that mice lacking Cav1.4 have defects in the development of photoreceptor ribbon synapses (Liu et al., 2013; Zabouri and Haverkamp, 2013). While RS1 is normally present in the IS, ONL, OPL and INL (Figure 5B), its expression was reduced in all retinal layers in the Cav1.4^−/− retina (Figure 5B). Since we previously reported that the secretion of RS1 depends on LTCCs in the chicken retina (Ko et al., 2008), these morphological results suggest that the reduction of calcium influx through LTCCs might dampen the expression or secretion of RS1 from photoreceptors. Alternatively, since Cav1.4 is necessary for ribbon synapse formation, the deletion of Cav1.4 may cause structural defects and decreased numbers of ribbon synapses, which might diminish the amount of RS1, a binding partner of Cav1.4.

The retina of Cav1.3-null mutant (Cav1.3^−/−) mice displays mild morphological changes in the OPL and slightly dampened light responses (Busquet et al., 2010), but these animals still have vision. However, it is not clear whether deletion of Cav1.3 might affect RS1 distribution in the retina, so we next examined the distribution of RS1 in Cav1.3^−/− retinas. Cav1.3 and RS1 were both present in all major layers of the WT retina (Figure 6A). Deletion of Cav1.3 (Cav1.3^−/−) dampened the RS1 signal in the IS of photoreceptors and OPL (Figures 6A,B), suggesting that reduction of calcium influx due to a lack of Cav1.3-LTCCs decreased the expression and/or secretion of RS1 from photoreceptors and bipolar cells. An alternative explanation is that normal interaction between Cav1.3 and RS1 is necessary for the retention of both on the plasma membrane.