CaVβ-subunit dependence of forward and reverse trafficking of CaV1.2 calcium channels

Auxiliary CaVβ subunits interact with the pore forming CaVα1 subunit to promote the plasma membrane expression of high voltage-activated calcium channels and to modulate the biophysical properties of Ca2+ currents. However, the effect of CaVβ subunits on channel trafficking to and from the plasma membrane is still controversial. Here, we have investigated the impact of CaVβ1b and CaVβ2a subunits on plasma membrane trafficking of CaV1.2 using a live-labeling strategy. We show that the CaVβ1b subunit is more potent in increasing CaV1.2 expression at the plasma membrane than the CaVβ2a subunit and that this effect is not related to modification of intracellular trafficking of the channel (i.e. neither forward trafficking, nor recycling, nor endocytosis). We conclude that the differential effect of CaVβ subunit subtypes on CaV1.2 surface expression is likely due to their differential ability to protect CaV1.2 from degradation.


Introduction
Calcium influx generated by voltage-gated calcium channels plays a critical role in neuronal functions such as excitability and gene transcription [1,2]. High-voltageactivated (HVA) Ca 2+ channels are formed by assembly of several subunits including the pore forming Ca V α 1 subunit and auxiliary Ca V α 2 δ and Ca V β subunits [3].
Among HVA channels, Ca V 1.2 is the most abundant in the mammalian brain [4] and it is localized in clusters in dendritic shafts and spines [5,6]. Auxiliary Ca V β subunits are key modulators of channel biophysical properties and their targeting to the plasma membrane [7,8]. Four Ca V β isoforms have been identified and they are all expressed in the brain. They contain 2 highly conserved domains: a Src Homology 3 domain and a guanylate kinase domain and variable N-terminal, Hook and C-terminal domains. The guanylate kinase domain binds to the α-interacting domain (AID) in the intracellular I-II loop of Ca V α 1 subunits. The binding of Ca V β to the AID protects Ca V α 1 from degradation by the proteasome [9]. However, the impact of Ca V β subunits on the trafficking of Ca V 1.2 to and from the plasma membrane is still a matter of debate [10][11][12].
In this study, we investigated the effects of two Ca V β subunits, the cytoplasmic Ca V β1b subunit and the membrane anchored (palmitoylated) Ca V β2a subunit [13,14], on plasma membrane trafficking of Ca V 1.2 using a livelabeling strategy based on a Ca V 1.2 construct tagged with bungarotoxin binding sites. We show that Ca V β1b is more potent in increasing Ca V 1.2 expression at the plasma membrane than Ca V β2a and that this effect is not linked to modification of either forward trafficking, recycling or endocytosis. We suggest that the effect of different Ca V β subunits on Ca V 1.2 surface expression is likely due to their differential ability to protect Ca V 1.2 from degradation.

Molecular biology
Alpha-bungarotoxin binding sites (BBS: WRYYESSLEP-YPD) were inserted between the S5 and the P loop of domain II of Ca V 1.2 (downstream Q683: FDEMQ-BBS-TRRST) using standard molecular biology techniques. Briefly, two oligonucleotides coding for a BBS and flanked by MluI restriction sites (oligo A: 5ʹ-ACG CGT CGG ACC GGT TGG AGA TAC TAC GAG AGC TCC CTG GAG CCC TAC CCT GAC CGT A-3ʹ; oligo B: 5ʹ-CGC GTA CGG TCA GGG TAG GGC TCC AGG GAG CTC TCG TAG TAT CTC CAA CCG GTC CGA-3ʹ) were synthesized, annealed and cloned into pMT2 Ca V 1.2 construct (rat brain Ca V 1.2 from T. Snutch; GenBank: M67515.1) linearized with MluI. Correct orientation and location of oligonucleotide cloning were confirmed by sequencing the plasmids. A triple BBS construct was generated.

Electrophysiology recordings
Twenty-four hours after transfection, tsA-201 cells were transferred to a 30 °C incubator for 48 h before being used for experiments. Whole-cell patch-clamp recordings were performed and analyzed as described previously [15]. Briefly, currents were recorded at room temperature (22-24 °C) using an Axopatch 200B amplifier and pClamp 9.2 software. Patch pipettes were filled with a solution containing the following (in mM): 130 CsCl, 2.5 MgCl 2 , 10 HEPES, 5 EGTA, 3 Na-ATP, 0.5 Mg-GTP, pH 7.4. The external solution contained the following (in mM): 132.5 CsCl, 1 MgCl 2 , 10 HEPES, 5 BaCl 2 , 10 glucose, pH 7.4. Current-voltage relationships were obtained by applying 250 ms pulses ranged from − 50 to + 50 mV in 5 mV increment from a holding potential of − 100 mV. Current density-voltage relationships were fitted with a modified Boltzmann equation as follows: I = (G max × (V − V rev ))/(1 + exp(−(V − V 50,act )/k)), where I is the current density (in pA/pF), G max is the maximum conductance (in nS/pF), V rev is the reversal potential, V 50,act is the midpoint voltage for current activation and k is the slope factor.

Trafficking assays and confocal microscopy
tsA-201 cells were plated onto glass-bottomed dishes (MatTek Corp., Ashland, MA) precoated with polyl-lysine and transfected as described above. After 3 days expression, cells were washed twice with Krebs-Ringer solution with HEPES (KRH) (in mM; 125 NaCl, 5 KCl, 1.1 MgCl 2 , 1.2 KH 2 PO 4 , 2 CaCl 2 , 6 Glucose, 25 HEPES, 1 NaHCO 3 ). For endocytosis experiments, cells were incubated with 10 µg/ml α-bungarotoxin Alexa Fluor ® 488 conjugate (BTX488) (Thermo Fisher Scientific) at 17 °C for 30 min. The unbound BTX488 was removed by washing with KRH, and the labelled cells were returned to 37 °C. Endocytosis was terminated by fixing the cells with cold 4% PFA in PBS for 5 min, and then permeabilized with 0.05% Triton X-100 in PBS for 10 min. Cells were blocked with 10% FBS in PBS for at least 30 min and incubated with the primary Ab (rabbit anti-Ca V 1.2, 1:200, Alomone labs) for 1 h at room temperature. Samples were washed and incubated with secondary conjugated Ab anti-rabbit AF594 (1:500; Thermo Fisher Scientific) for 1 h at room temperature. After washing, samples were covered with SlowFadeTM Gold antifade mountant (Thermo Fisher Scientific). For the forward trafficking assay, the cells were incubated with 10 μg/ml unlabeled α-bungarotoxin (BTX; Thermo Fisher Scientific) at 17 °C for 30 min. The unbound BTX was washed off with KRH, and the cells were then incubated with 10 μg/ml BTX488 in KRH at 37 °C. To stop the reaction, cells were washed twice with cold KRH and then fixed with 4% PFA in PBS. Brefeldin A (BFA; 200 ng/ml (0.71 μM); Sigma-Aldrich) in 0.4% DMSO was added to the cells in FBS-free culture medium for 4 h before the experiment, and during the experiment in KRH buffer. Cells were examined on a Leica SP8 confocal microscope using a 63×/1.4 numerical aperture oil-immersion objective in 16-bit mode. Acquisition settings, chosen to ensure that images were not saturated, were kept constant for each experiment.

Statistical analysis
Data are given as mean ± SEM. Statistical comparisons were performed using paired and unpaired Student's t tests, as appropriate, using SigmaPlot 14.5 or Prism GraphPad. Differences were considered to reach statistical significance when p < 0.05.

Results and discussion
To monitor the trafficking of Ca V 1.2 to the plasma membrane we introduced α-bungarotoxin binding sites in the extracellular loop of the channel (Fig. 1a). This construct, Ca V 1.2-BBS, remained functional and generated Ba 2+ currents with a density similar to the WT channel (− 9.7 ± 2.0 pA/pF, n = 9, vs − 13.5 ± 2.5 pA/pF, n = 18, for Ca V 1.2-BBS and WT, respectively; Fig. 1b). However, the insertion of the tag induced a slight increase in the slope of the activation curve (from − 9.3 ± 0.9 mV, n = 9, to − 7.1 ± 0.3 mV, n = 18, for WT and Ca V 1.2-BBS, respectively, p = 0.007) and a depolarizing shift of the reversal potential (from 37.5 ± 2.6 mV, n = 9, to 48.4 ± 2.4 mV, n = 18, for WT and Ca V 1.2-BBS, respectively, p = 0.01). The V 50,act and the G max remained unchanged (V 50,act = − 14.5 ± 2.2 mV, n = 9, and − 14.7 ± 0.6 mV, n = 18; G max = 0.34 ± 0.08 nS/ pF, n = 9, and 0.30 ± 0.05 nS/pF, n = 18, for WT and Ca V 1.2-BBS, respectively). The effect on reversal potential may be indicative of an effect of the BBS modification on permeability, but this should have little bearing on the utility of this construct for trafficking studies.
We subsequently checked the cell surface expression of Ca V 1.2-BBS and compared the effect of co-expressing different types of auxiliary Ca V β subunits. Three days after transfection, tsA-201 cells were live-labelled with α-BTX-488 for 30 min at 17 °C, fixed and the fluorescence was quantified ( Fig. 1c and d). We found that . Ca V 1.2-BBS was co-expressed with Ca V α 2 δ-1 (left, no β) and either Ca V β1b (center) or Ca V β2a (right). Cells were incubated at 17 °C with BTX-AF488 for 30 min and fixed. The cells were then permeabilized and stained with a rabbit anti-Ca V 1.2 Ab and secondary Ab anti-rabbit AF594 (bottom panels). Scale 20 µm. d Average Ca V 1.2-BBS surface expression co-transfected with Ca V α 2 δ-1 and either Ca V β1b (black bar), or Ca V β2a (open bar), or empty vector (gray bar). Bars are mean (± SEM) normalized to Ca V β1b mean. ***p < 0.001, n = 14; $$$ p < 0.001, n = 5; paired t-test, n numbers correspond to independent experiments the co-expression of Ca V β1b induced a 60% increase in Ca V 1.2-BBS surface expression compared with no Ca V β. Additionally, although the co-expression of Ca V β2a also increased Ca V 1.2-BBS cell surface expression, its effect was not as marked as Ca V β1b since only a 30% increase of fluorescence was recorded ( Fig. 1c and d). This result is in good agreement with a study showing differential effects of Ca V β subunits on Ca V 1.2-generated current densities, although it is important to note that interpretations of electrophysiological measurements can be confounded by effects on channel biophysics [17].
We then aimed to gain insight into the Ca V β subunitdependent mechanism(s) responsible for the effects on Ca V 1.2 surface expression. Plasma membrane expression of Ca V 1.2 results from the balance between the incorporation of newly synthetized Ca V 1.2 from the ER, recycled Ca V 1.2 from endosomal compartments and the removal of channels from the plasma membrane by endocytosis [18]. We first monitored the impact of Ca V β subunits on Ca V 1.2 endocytosis by comparing the rate of internalization of Ca V 1.2-BBS (Fig. 2). We showed that Ca V 1.2-BBS, either co-expressed with Ca V β1b or Ca V β2a, exhibited similar kinetics of endocytosis with a time constant of ~ 6 min (Fig. 2b). This is in line with previous studies on N-type calcium channels [19,20] and measurements on cardiac cell lines [21]. We note that when Ca V β subunits are not co-expressed, no reduction of Ca V 1.2-BBS fluorescence is detected over the duration of BTX incubation (after 20 min, Ca V 1.2-BBS fluorescence represented 100 ± 12% of the initial fluorescence, n = 3), suggesting that Ca V β subunits promote Ca V 1.2 endocytosis. This conclusion is different from that of a recent study showing that stabilizing the Ca V 1.2-Ca V β2a interaction via the creation of a concatemer increases the retention time of Ca V 1.2 at the plasma membrane in HEK-293 and HLA-1 cells [12]. However, in our experimental conditions, the starting level of Ca V 1.2-BBS fluorescence without Ca V β subunit is very low, close to the detection limit, and we cannot exclude that some endocytosis of channels may take place even in the absence of Ca V β. It was previously shown that Ca V 1.2 internalization is dynamin-dependent [21]. We took advantage of the dominant negative effect of the dynamin mutant K44E [16] to show that Ca V 1.2 internalization depends on dynamin, regardless of coexpressed Ca V β subtype (Fig. 2c and d).
Next, we examined the effect of Ca V β subunits on net forward trafficking by monitoring the insertion of Ca V 1.2-BBS into the plasma membrane as a function of time ( Fig. 3a and b). In the no Ca V β condition, Ca V 1.2-BBS surface expression doubled after 10 min and stayed stable during the next 40 min. However, when Ca V β subunits were co-expressed, we recorded an increase of Ca V 1.2-BBS surface expression that reached a plateau after 40 min. The increases were comparable for both Ca V β subunits and represented ~ 5 times the starting amount of Ca V 1.2-BBS surface expression. This is . The data were compared using an unpaired t-test. The data were fitted with single exponentials. The time constants of the fits were 10.6 ± 6.6 min and 11.6 ± 8.1 min for Ca V β1b and Ca V β2a (n = 7), respectively surprising, given that the steady state level of Ca V 1.2 is higher in the presence of the Ca V β1b isoform (see below).
Finally, we used BFA to disrupt the Golgi apparatus and prevent the transfer of newly synthesized channels from the ER to the plasma membrane ( Fig. 3c and d). Using this strategy, we were able to estimate that 70% of the surface Ca V 1.2-BBS originated from a recycling pathway [21] and we could also attribute 30% to a forward trafficking process. These contributions remained identical irrespective of whether Ca V 1.2-BBS was co-expressed with Ca V β1b or Ca V β2a subunits. Interestingly, the contributions recycling/forward trafficking for Ca V 2.2 cell surface expression were reported to be closer to 50% [19,20], although it is important to highlight the fact that these studies were performed in a neuronal cell line and that further investigations would be needed to rule out a celldependent effect.
Altogether, we showed that the Ca V β subunit subtype dependent effect on Ca V 1.2 surface expression was not associated with any modifications of the kinetics for forward trafficking, endocytosis and recycling. These results suggest that the level of Ca V 1.2 (available in the ER to be trafficked to the plasma membrane) is differentially increased in the presence of different Ca V β subunits. Such a mechanism is supported by the conclusions of a previous study from our group that showed that Ca V β subunits protect Ca V 1.2 from ubiquitination and degradation by the proteasome [9]. If so, then the fact that the observed effects were greater with Ca V β1b than with Ca V β2a suggests the possibility that a difference in the amount of Ca V 1.2 in the ER could be due to a differential ability of the two Ca V β subunits to protect Ca V 1.2 from degradation. This could potentially be due to the selective palmitoylation and membrane anchoring of Ca V β2a compared to the pure cytoplasmic expression of Ca V β1b which may have better access to the pore forming subunit in the ER. Alternatively, it is possible that the latter may be expressed at higher levels than the former and thus more effective in protecting the channel from degradation. We also consider the possibility that the protective effects of the Ca V β subunit may not be dependent on a physical interaction with the Ca V 1.2, but instead act by regulating calcium channel expression at the transcriptional or translational level. For example, for Ca V 3 calcium channels, coexpression of ancillary subunits promotes current densities despite an absence of a physical interaction [22]. On the other hand, mutation of Ca V 1.2 residue W440 which prevents the physical association with the Ca V β subunit leads to compromised membrane expression of the channel [9,23], thus arguing against a diffuse effect on Ca V 1.2 protein expression. Overall, our data are consistent with a mechanism by which Ca V β subunits are more important for regulating the levels of Ca V 1.2 channels at the level of the ER, rather than directly altering the forward and reverse plasma membrane trafficking of the channel complexes. This does not negate the possibility that these subunits may be involved in modulating the targeting of Ca V 1.2 channels to specific sub-loci within the plasma membrane.