Rab Interacting Molecules 2 and 3 Directly Interact with the Pore-Forming CaV1.3 Ca2+ Channel Subunit and Promote Its Membrane Expression

Abstract

Rab interacting molecules (RIMs) are multi-domain proteins that positively regulate the number of Ca²⁺ channels at the presynaptic active zone (AZ). Several molecular mechanisms have been demonstrated for RIM-binding to components of the presynaptic Ca²⁺ channel complex, the key signaling element at the AZ. Here, we report an interaction of the C₂B domain of RIM2α and RIM3γ with the C-terminus of the pore-forming α-subunit of Ca_V1.3 channels (Ca_V1.3α1), which mediate stimulus-secretion coupling at the ribbon synapses of cochlear inner hair cells (IHCs). Co-expressing full-length RIM2α with a Ca²⁺ channel complex closely resembling that of IHCs (Ca_V1.3α1-Ca_Vß2a) in HEK293 cells doubled the Ca²⁺-current and shifted the voltage-dependence of Ca²⁺ channel activation by approximately +3 mV. Co-expression of the short RIM isoform RIM3γ increased the Ca_V1.3α1-Ca_Vß2a-mediated Ca²⁺-influx in HEK293 cells, but disruption of RIM3γ in mice left Ca²⁺-influx in IHCs and hearing intact. In conclusion, we propose that RIM2α and RIM3γ directly interact with the C-terminus of the pore-forming subunit of Ca_V1.3 Ca²⁺ channels and positively regulate their plasma membrane expression in HEK293 cells.

Keywords: active zone; channel clustering; exocytosis; hair cell; hearing; ribbon synapse.

Figure 1

Rab interacting molecules 2α (RIM2α) interacts with Ca_V1.3 via C₂-domain binding to the Ca_V1.3α C-terminus. (A) Schematic representation of RIM2α and HA-tagged Ca_V1.3 C-terminus (top). Immunoblot (IB) of an exemplary co-immunoprecipitation assay from co-transfected HEK293T cell lysates shows that full length RIM2α co-immunoprecipitated with the C-terminal region of Ca_V1.3 (bottom, input 3%). (A_i) Quantifications of co-immunoprecipitated RIM2α with the HA-tagged C-terminal region of Ca_V1.3 (N = 3). (B) Schematic representation of RIM3γ and HA-tagged Ca_V1.3 C-terminus (top). IB of an exemplary co-immunoprecipitation assay from co-transfected HEK293T cell lysates, showing that the C₂B domain of RIM3γ suffices to co-immunoprecipitate with the C-terminal region of Ca_V1.3 (bottom, input 3%). (B_i) Quantifications of co-immunoprecipitated RIM3γ with the HA-tagged C-terminal region of Ca_V1.3 (N = 2). (C) Schematic representation of fusion proteins of RIM2α subdomains, RIM3γ and Ca_V1.3 C-terminus as used for the binding assays (top). Immunoblot (IB) of an exemplary co-immunoprecipitation assay from co-transfected HEK293T cell lysates, showing that the peptide containing the RIM2α C₂-domains, but not the RIM2α ZF-PDZ peptide co-immunoprecipitated with HA-tagged C-terminal region of Ca_V1.3α (bottom, input 3%). (C_i) Quantifications of co-immunoprecipitated N-(ZF-PDZ, N = 3) or C-terminal (C₂A-C₂B, N = 3) domains of RIM2α and RIM3γ with the C-terminal region of Ca_V1.3. (D) Schematic representation of fusion proteins used for the GST pull-down assay (Left). IB of an exemplary GST pull-down assay of HA-tagged Ca_V1.3 (1509–2203aa) overexpressing HEK293T cell lysates, showing that the C₂B-domain of RIM2α (GST-RIM2α C₂B), but not the C₂A or PDZ domain of RIM2α (GST-RIM2α C₂A and GST-RIM2α PDZ) pulled down Ca_V1.3 and were detected by an anti-HA antibody (right, input 3%). (D_i) Quantification of GST-bound fraction of HA-tagged Ca_V1.3 pulled down by respective RIM2α and RIM3γ domains (N = 4). Note that the RIM2α-C₂B and RIM3γ pulled down Ca_V1.3 while the RIM2α-C₂A and -PDZ domains did not.

Figure 2

RIM2α positively regulates functional expression of Ca_V1.3 in HEK 293/SK3-1 cells. (A) Confocal maximum projection of a representative HEK293/SK3-1 cell co-transfected with Ca_V1.3 (green) and RIM2α (magenta; Scale bar: 10 μm). (A_i) 2D STED image of (A) showing overlap RIM2α and Ca_V1.3 immunofluorescence indicative of co-localization (Identical scale in all enlarged images; Scale bar: 1 μm). (B) The co-expression of RIM2α increases the Ca²⁺-current density amplitude in transiently transfected HEK293/SK3-1 cells: average I-V traces, depicted as current densities of HEK293/SK3-1 transfected with either Ca_V1.3 alone (con. for control, gray, n = 9) or cells co-transfected with RIM2α (black, n = 11; left). Summary plot of maximum current densities shown as box plot (10, 25, 50, 75 and 90% percentiles) overlaid with individual data points (right). Note the two-fold increase in maximum current-density amplitude in the presence of RIM2α (***p < 0.005, Wilcoxon rank-sum test). (C) Voltage-dependence of activation curve derived from (B) in the presence or absence of RIM2α. The voltage-dependence of activation curve is mildly shifted towards more positive potentials in presence of RIM2α ( p < 0.05, Student’s t-test). (D,E) Ca²⁺-current inactivation is not affected by co-expression of RIM2α: average I_Ca traces recorded in the presence (n = 11) or absence (n = 9) of RIM2α after step depolarization to the voltage of maximal Ca²⁺-currents ( V_Imax) for 5 s. For a better comparison traces were normalized to the maximum current (I_peak). Residual Ca²⁺-currents (I_res) were indistinguishable between recording conditions ( p > 0.05, One-way ANOVA with post hoc Holm-Sidak correction).

Figure 3

RIM3γ positively regulates functional expression of Ca_V1.3 in HEK 293/SK3-1 cells. (A) Confocal maximum projection of a representative HEK293/SK3-1 cell co-transfected with Ca_V1.3 (green) and RIM3γ (magenta; Scale bar: 5 μm). (A_i) 2D STED image of (A) showing overlap of RIM3γ and Ca_V1.3 immunofluorescence indicative of co-localization (Scale bar of enlarged images: 0.5 μm). (B) Elevated Ca²⁺-current amplitudes in the presence of RIM3γ in transfected HEK293/SK3-1 cells: average I-V traces recorded in HEK293/SK3-1 in the presence (dark gray, n = 10) or absence (con. for control, gray, n = 10) of RIM3γ (left). Summary plot of maximum current densities shown as box plot (10, 25, 50, 75 and 90% percentiles) overlaid with individual data points (right). Note the increase in maximum current density amplitude in the presence of RIM3γ (***p < 0.005, Wilcoxon rank-sum test). (C) Voltage-dependence of activation derived from B is not shifted in the presence of RIM3γ ( p > 0.05, Student’s t-test). (D,E) Ca²⁺-current inactivation is not affected by co-expression of RIM3γ: average Ca²⁺-current traces recorded in the presence (n = 8) or absence (n = 10) of RIM3γ. Residual Ca²⁺-currents inactivating were indistinguishable between recording conditions ( p > 0.05; One-way ANOVA with post hoc Holm-Sidak correction).

Figure 4

Analysis of RIM3γ knock-out mice shows that RIM3γ is not required for the function of inner hair cell (IHC) ribbon synapses. (A) Cartoon depicting the RIM3γ wild-type and targeted allele, in which a gene trap cassette was inserted in the intronic region between exon 3 and 4. The gene trap cassette consists of a promoterless En2 splice acceptor site followed by an IRES-lacZ gene cassette. Splicing of this gene trap cassette to the end of RIM3γ exon 3 results in disruption of RIM3γ expression. (B) Quantitative real time RT-PCR of wild-type, heterozygous and homozygous RIM3γ^−/− mice revealed reduced gene transcript levels in mRNA prepared from hippocampus (HC), cerebellum (CB) and cortex (CX; N = 10 animals per group). (C) Quantification immunoblot (N = 5 animals per group). Significance: two-way ANOVA, Bonferoni post hoc Test *p < 0.05, **p < 0.01 and ***p < 0.001. (D) Immunoblot of RIM3γ^−/− mice, showed a pronounced reduction in protein levels in HC, CB and CX. To verify RIM3γ reactivity of the antibody and the identity of the band analyzed, HEK293 cell lysates overexpressing the RIM3γ were included in the analysis (RIM3γOE). (E) Grand average of auditory brainstem responses (ABR) of RIM3γ^−/− (N = 9) and littermate control (N = 7) animals show normal wave I amplitudes and latencies in the absence of RIM3γ. The shaded area represents the SEM and roman letters indicate the five characteristic ABR waves representing auditory nuclei along the auditory pathway. (F) Mean ABR hearing thresholds of RIM3γ^−/− and control animals show near normal hearing thresholds over the whole frequency range in the absence of RIM3γ. Shaded areas represent the SEM. (G) Mean current-voltage relationship of Ca²⁺-currents in RIM3γ^−/− (n = 20) and littermate control (n = 9) IHCs. Shaded areas represent the SEM. Note that the maximum current amplitude is comparable between genotypes ( p > 0.05, Student’s t-test). (H) Mean Ca²⁺-current trace of RIM3γ^−/− (n = 24) and control (n = 14) IHCs (left) showing comparable inactivation kinetics summarized in single value plot for the residual Ca²⁺-current after 200 ms (right, p < 0.05, Student’s t-test). (I) Mean ± SEM. cell capacitance increments (∆C_m) with respective Ca²⁺-charge (Q_Ca) upon depolarizations of increasing durations of RIM3γ^−/− (n = 24) and control (n = 14) IHCs. Both ∆C_m and Q_Ca are unchanged in the absence of RIM3γ.