Asn-Linked Glycosylation Contributes to Surface Expression and Voltage-Dependent Gating of Ca_v1.2 Ca²⁺ Channel

Abstract

The Ca_v1.2 Ca²⁺ channel is essential for cardiac and smooth muscle contractility and many physiological functions. We mutated single, double, and quadruple sites of the four potential Asn (N)-glycosylation sites in the rabbit Ca_v1.2 into Gln (Q) to explore the effects of Nglycosylation. When a single mutant (N124Q, N299Q, N1359Q, or N1410Q) or Ca_v1.2/WT was expressed in Xenopus oocytes, the biophysical properties of single mutants were not significantly different from Ca_v1.2/WT. In comparison, the double mutant N124,299Q showed a positive shift in voltage-dependent gating. Furthermore, the quadruple mutant (QM; N124,299,1359,1410Q) showed a positive shift in voltage-dependent gating as well as a reduction of current. We tagged EGFP to the QM, double mutants, and Ca_v1.2/WT to chase the mechanisms underlying the reduced currents of QM. The surface fluorescence intensity of QM was weaker than that of Ca_v1.2/WT, suggesting that the reduced current of QM arises from its lower surface expression than Ca_v1.2/WT. Tunicamycin treatment of oocytes expressing Ca_v1.2/WT mimicked the effects of the quadruple mutations. These findings suggest that Nglycosylation contributes to the surface expression and voltage-dependent gating of Ca_v1.2.

Introduction

Voltage-activated Ca2+ channels (VACCs) are heteromultimeric membrane proteins composed of a pore-forming alpha1 (α1 ) subunit and its auxiliary subunits (α2δ, β, and γ). The largest α1 subunits act as voltage-sensing and poreforming machinery controlling external Ca2+ influx, mainly determining biophysical and pharmacological characteristics of VACCs, while the auxiliary subunits modulate the expression, drug binding sensitivity, and current kinetics of α1 subunits [2, 12]. Molecular cloning and expression experiments have identified 10 α1 subunit genes for VACCs of which functional types are assigned as follows. L-type Ca2+ channels are encoded by four Cav1 isoforms (Cav1.1-1.4). Non-L-type Ca2+ channels (N-, P/Q-, and R-type channels) are encoded by Cav2 (Cav2.1-2.3), respectively. Low voltage-activated T-type Ca2+ channels are encoded by three Cav3 isoforms (Cav3.1-3.3) [2, 12].

Cav1.2 is a member of the L-type Ca2+ subfamily (Cav1.1- 1.4), which is predominantly expressed in heart and blood vessels. Electrical stimulation activates L-type Cav1.2 channel complex formed with auxiliary subunits, through which external Ca2+ entry induces Ca2+ release from ryanodine receptors in cardiac myocytes and activates calmodulin in smooth muscle, consequently signaling cascades leading to muscle contraction of heart and blood vessels. Additionally, expression of Cav1.2 was broadly detected in neuronal tissues, endocrine tissues, lung, trachea, and ovary, having suggested its functional implications in neuronal excitability, hormone secretion, activation of enzymes, and gene expression [7, 17]. The broad involvement of Cav1.2 in important physiological functions was supported by the identification of a missense mutation (G406R) of human Cav1.2, which is linked with Timothy syndrome, showing diverse dysfunctions in multi-organs, including cardiac arrhythmia, immune deficiency, hypoglycemia, abnormal cognition, and autism [15].

Ion channels in the plasma membrane have been reported to be profoundly modified after translation by multiple types of post-translational processes, including glycosylation, formation of disulfide bond(s), enzyme digestion, phosphorylation(s), and ubiquitination. Asparagine (Asn, N)-linked glycosylation (N-glycosylation) has been reported to potently affect the structural folding, membrane targeting, expression level, stability, and voltage-dependent properties of many ion channels in the plasma membrane [14]. In Kv1.5, for example, removal of the N-glycosylation site by mutagenesis or lessening of N-glycans caused to positively shift voltage-dependency for channel gating [13]. Modification of N-glycosylation was reported to alter the activation and inactivation kinetics of Kv3.1 currents [6]. In TRESK and Kv1.2, N-glycosylation of the channels was shown to affect channel expression in the plasma membrane [3, 5]. Similarly, reduction or elimination of N-glycosylation of the Cav3.2 Ttype Ca2+ channel was reported to decrease channel activity and expression with altered current kinetics [11, 16]. Furthermore, Orestes et al. [11] further suggested that deglycosylation treatments of Cav3.2 can be a potential method to relieve pain from diabetic peripheral neuropathy via a decrease of Cav3.2 channel activity.

In this study, we analyzed the amino acid sequence of the rabbit Cav1.2 α1 subunit and found four potential sites for N-glycosylation (corresponding to “Asn-X-Thr/Ser”) on the extracellular loops of the rabbit Cav1.2 subunit [9]. We examined functional roles of the N-glycosylation candidates by site-directed point mutations. Single point mutations of the four Asn residues did not cause any significant alterations, but the quadruple mutations of them caused to significantly reduce current amplitude, and to positively shift the voltage-dependent gating. These effects were mimicked by treatment of tunicamycin inhibiting N-glycosylation. These findings suggest that Nglycosylation plays critical roles in conferring its functional channel expression and biophysical properties to the Cav1.2 channel.

 

Materials and Methods

Chemicals

Most of the chemicals were purchased from Sigma–Aldrich (St. Louis, MO, USA). A PCR kit used to generate site-directed mutagenesis was purchased from Vivagen (Seoul, Korea).

Point Mutations of Potential N-Glycosylation Sites in Cav1.2

The rabbit Cav1.2 (GenBank Accession No. NM_001136522.1) contains four potential Asn-glycosylation sites: N124, N299, N1359, and N1410. The four Asn (N) sites were mutated into Gln (Q) using the PCR-based overlap extension method. Mutated site(s) were confirmed by sequencing analysis. The restriction enzyme sites were marked by numbers in parentheses. In total, seven mutants, including four single-site mutants, two double-site mutants, and one quadruple-site mutant, were generated.

N124Q. The forward primer to amplify the upper cassette for N124Q was AGCCATCGATGCGGCC and the reverse primer was AATTGGTGGCCTGGGAGTCATCTTCTGGAAAG. The forward primer to amplify the lower cassette for N124Q was AGA TGACTCCCAGGCCACCAATTCCAACCTG and the reverse primer was CATGAAGAGCTCCAGGCC. The upper and lower cassettes were connected by second-step PCR. N124Q was constructed by ligating the ClaI- and SacI-digested PCR fragments into the ClaI- (448, Cav1.2) and SacI-digested (1147, Cav1.2) plasmid Cav1.2 pGEM.

N299Q. The forward primer to amplify the upper cassette for N124Q was TTGCCAATTGTGTGGCCT and the reverse primer was ACACGGTGCCCTGCTGGCACTG CCGCCCG. The forward primer to amplify the lower cassette for N124Q was GCAGTG CCAGCAGGGCACCGTGTGCAAGCC and the reverse primer was TATTCCACCGGCGCCAG. The upper and lower cassettes were connected by second-step PCR. N124Q was constructed by ligating the MfeI- and SfiI-digested PCR fragments into the MfeI-(689, Cav1.2) and SfiI-digested (1759, Cav1.2) plasmid Cav1.2 pGEM.

N1359Q and N1410Q. The forward primer to amplify the upper cassettes for N1359Q and N1410Q was AAAGCACGTGGTTCA GTG and the reverse primers were CCGTGGTGTCCTGCAGGG CGATTTTCCCAAA and CTTCTGTGCTCTGGTGGGGCTCAG ACTCTG, respectively. The forward primers to amplify the lower cassettes for N1359Q and N1410Q were AATCGCCCTGCAGGA CACCACGGAGATCAAC and TGAGCCCCACCAGAGCAC AGAAGGGGAGACC, respectively, and the reverse primer was TTGCAAGCCACACGGTGA. The upper and lower cassettes were connected by second-step PCR. N1359Q and N1410Q were constructed by ligating the BsrGI- and DraIII-PCR digested fragments, MfeI (689, Cav1.2)-AflII (2880, Cav1.2) fragment, and AflII (2880, Cav1.2)-BsrGI (3445, Cav1.2) fragment into the MfeI- (689, Cav1.2) and DraIII-digested (4892, Cav1.2) plasmid Cav1.2 pGEM.

N124,299Q. N124,299Q was constructed by ligating the SacI (1147, N299Q)-XmaI (6590, N299Q) fragment into the SacI- (1147, N124Q) and XmaI-digested (6590, N124Q) plasmid N124Q pGEM.

N1359,1410Q. N1359,1410Q was constructed by introducing the N1410Q mutation into N1359Q. The upper and lower cassettes were obtained from N1359Q by PCR amplification using the same primer sets used to construct N1410Q. The upper and lower cassettes were combined by second-step PCR. N1359,1410Q was finally constructed by ligating the BsrGI- and DraIII-digested PCR fragment, MfeI (689, Cav1.2)-AflII (2880, Cav1.2) fragment, and AflII (2880, Cav1.2)-BsrGI (3445, Cav1.2) fragment into the MfeI- (689, Cav1.2) and DraIII-digested (4892, Cav1.2) plasmid Cav1.2 pGEM.

N124,299,1359,1410Q (quadruple mutant; QM). The quadruple mutant was constructed by ligating the MfeI (689, N124,299Q)-AflII (2880, N124,299Q) fragment into the MfeI- (689) and SacI-digested (1147) plasmid N1359,1410Q pGEM.

Expression of Cav1.2 and Mutant Channels in Xenopus Oocytes

The GenBank accession numbers of rabbit Cav1.2 and rat β3 are NM_001136522.1 and M88751, respectively. In vitro synthesis of cRNAs and their expression in Xenopus oocytes were reported previously [8]. Briefly, the Cav1.2 and mutant cDNAs were linearized at their 3’ ends by HindIII, while the β3 cDNA was linearized by SacII. cRNA transcripts were in vitro synthesized using T7 RNA polymerase (Ambion, Austin, TX, USA). Female Xenopus laevis frogs were obtained from Hallym University (Gangwon-do, Chuncheon, Korea). Several ovary lobes were surgically cut from frogs under anesthesia and manually torn into small clusters of 4-6 eggs in a standard oocyte solution (SOS; in mM: 100 NaCl, 2 KCl, 1.8 CaCl2, 1 MgCl2, 5 HEPES, 2.5 pyruvic acid, and 50 µg/ml gentamicin; pH 7.6). Follicle membranes were eliminated by shaking the oocyte clusters treated with collagenase (10 mg/ml; Gibco-BRL, Gaithersburg, MD, USA) in a Ca2+ -free solution (in mM: 82.5 NaCl, 2.5 KCl, 1 MgCl2, and 5 HEPES; pH 7.6) for 40-60 min. Each oocyte was injected with ~5 ng of α1 subunit cRNA and 2.5 ng of β3 cRNA in a volume of 50 nl using a Drummond Nanoject pipette injector (Parkway, PA, USA) attached to a Narishige micromanipulator (Tokyo, Japan) under a stereo microscope.

Electrophysiological Recordings in Oocytes and Data Analysis

Recording of Ba2+ currents from oocytes were performed between 3 and 5 days after cRNA injection at room temperature by a twoelectrode voltage clamping method, for which a two-electrode voltage-clamp amplifier (OC-725C; Warner Instruments, Hamden, CT, USA) was utilized. Glass microelectrodes were pulled from capillaries (Warner Instruments, Hamden, CT, USA) using a pipette puller (P97; Sutter Instrument Co., Novato, CA, USA), and filled with 3 M KCl, and the electrode resistance was ~1.0 MΩ. The recording solution contained 10 mM Ba(OH)2, 90 mM NaOH, 1 mM KOH, and 5 mM HEPES (pH 7.4 adjusted with methanesulfonic acid). The currents were sampled at 5 kHz and low pass-filtered at 1 kHz using the pClamp system (Digidata 1322A and pClamp 8; Molecular Devices, Palo Alto, CA, USA). Properties of the currents were analyzed using Clampfit software (Axon Instruments).

Activation curves were generated by the chord conductance method, in which the cord conductance (G) was calculated by dividing the current amplitude by the driving force (reversal potential–test potentials), and normalized to the maximum conductance. Activation curves were from fitting the normalized data to the Boltzmann equation (G = 1/[1+exp(V50,act −V)/k], where V50,act is the half-activation voltage, and k is the slope factor). The protocol for steady-state inactivation consisted of 10 sec prepulses between -90 mV and +30 mV from a holding potential of -80 mV, followed by a test potential of +10 mV. Peak currents were calculated to ratios to the maximal current elicited after a 10 sec prepulse potential of -90 mV and the ratios were plotted against prepulse potentials. The curve for steady-state inactivation was from fitting the data points with a Boltzmann equation (1/[1+exp (V50,inact –V)/k], where V50,inact is the potential for half-inactivation and k is the slope factor).

Graphical representation of the data was achieved with Prism software (GraphPad, San Diego, CA, USA). Data are given as the mean ± SEM. Differences were tested for significance using Student's unpaired t tests and one-way ANOVA Tukey tests, with p < 0.05 as the level of significance.

 

Results

Potential N-Glycosylation Sites in Rabbit Cav1.2

The amino acid sequence of the rabbit Cav1.2 α1 subunit (GenBank No. NM_001136522.1) was analyzed to look for the potential N-glycosylation motif (Asn-X-Thr/Ser) in its extracellular loops, in which the 1st Asn (N) residue is separated from the 3rd Ser or Thr residue by an amino acid (X), which can be all the amino acid residues except Pro. The rabbit Cav1.2 contains four potential N-glycosylation sites (N124, N299, N1359, and N1410), which are marked on the schematic diagram of the rabbit Cav1.2 (Fig. 1). N124 and N299 are positioned in the S1-S2 and S5-S6 linkers in domain I of the Cav1.2. Both N1359 and N1410 are positioned in the S5-S6 linker in domain IV.

Fig. 1.Potential N-glycosylation sites in rabbit Cav1.2. The schematic diagram of the rabbit Cav1.2 α1 subunit (GenBank # NM_001136522.1) exhibits the locations of four potential Nglycosylation sites in the extracellular surface: N124, N299, N1359, and N1410. N124 and N299 are in domain I of the Cav1.2, and N1359 and N1410 are in domain IV.

Single Site Mutations Did Not Alter Biophysical Properties of the Cav1.2

We first performed experiments to confirm the expression of Cav1.2 L-type channels in the oocyte system using nifedipine, which is known to be an L-type Ca2+ channel antagonist. Ba2+ currents from oocytes expressing Cav1.2 channels were evoked by +10 mV step pulses from a holding potential of -80 mV. Application of 1 µM nifedipine potently inhibited Cav1.2 currents, and the average inhibition percentage was about 80% (Fig. S1). This nifedipinesensitive blockade supports that the Cav1.2 L-type channels were expressed in oocytes.

To examine individual functions of the four potential Nglycosylation sites, we generated N124Q, N299Q, N1359Q, and N1410Q by mutating each of the four Asn (N) residues into Gln (Q). An equal amount (5 ng) of Cav1.2/WT cRNA or single-site mutant cRNA was co-injected with β3 cRNA (2.5 ng) into Xenopus oocytes. Their expression was detected as robust inward currents from the oocytes from the 3rd day after cRNA injection. Their representative current traces elicited by serial step pulses to -40, -10, +10, and +30 mV from a holding potential of -80 mV are displayed (Figs. 2A-2E). Their average current-voltage (I-V) data appeared to be almost overlapped (Fig. 2F), suggesting that the relevant properties of the single-site mutants are similar to those of Cav1.2/WT, including average current amplitude at various test potentials, voltage-dependency for activation, and reversal potential. When their cord conductance data derived from I-V data were plotted and analyzed (Fig. 2G), the V50,act values (50% activation potential) for Cav1.2/WT, N124Q, N299Q, N1359Q, and N1410Q were -7.2 ± 0.7, -8.4 ± 0.8, -7.5 ± 0.8, -7.7 ± 0.8, and -6.8 ± 0.7 mV, and their slope factors were 6.2 ± 0.4, 6.4 ± 0.5, 6.1 ± 0.5, 6.2 ± 0.5, and 6.7 ± 0.3 mV, respectively. The V50,act values and slope factors were not significantly different among them (one-way ANOVA Tukey tests, n = 9-14). Analysis of steady-state inactivation data showed that the V50,inact values for Cav1.2/ WT, N124Q, N299Q, N1359Q, and N1410Q were -17.7 ± 0.7, -18.0 ± 0.6, -17.7 ± 0.9, -19.5 ± 0.7, and -18.7 ± 0.6 mV, and their slope factors were -10.5 ± 0.6, -9.8 ± 0.5, -11.2 ± 0.8, -11.3 ± 0.7, and -10.1 ± 0.5 mV, respectively (Fig. 2H). The V50,inact values and slope factors were not significantly different among the groups (one-way ANOVA Tukey tests, n = 7-9). These results strongly suggest that disruption of any single site among the four potential N-glycosylation sites does not significantly affect the biophysical properties, including current amplitude and voltage-dependency for activation and steady-state inactivation.

Fig. 2.Electrophysiological properties of single N-glycosylation site mutants. (A-E) Representative current traces through Cav1.2/WT (A), N124Q (B), N299Q (C), N1359Q (D), and N1410Q (E). In an extracellular 10 mM Ba2+ solution, current traces were elicited from oocytes expressing Cav1.2/WT or each single mutant with β3 by a voltage protocol composed of depolarizing step pulses in 10 mV increments, from a holding potential of -80 mV. Among them, current traces evoked by step pulses to -40, -10, +10, and +30 mV from a holding potential of -80 mV are shown. (F) Average current-voltage (I-V) relationships of Cav1.2/WT ( ○ ), N124Q ( □ ), N299Q ( △ ), N1359Q ( ▽ ), and N1410Q (◇). Averaged current values (± SEM) were plotted against test potentials. Significant differences between average currents of Cav1.2/WT and the single mutants were not detected over most of test potentials (one-way ANOVA Tukey tests, p > 0.05, n = 9-14). (G) Activation curves of Cav1.2/WT and single mutants. Activation curves were generated by the chord conductance method (refer to the Methods section). The V50,act values and slope factors for Cav1.2/WT and the single mutants are not different among them (one-way ANOVA Tukey tests, p > 0.05, n = 9-14). (H) Steady-state inactivation curves of Cav1.2/WT and single mutants. The inactivation protocol and generation of steady-state inactivation curves are described in the Methods section. The V50,iact values and slope factors for Cav1.2/WT and the single mutants were not different among them (one-way ANOVA Tukey tests, p > 0.05, n = 7-9).

Quadruple Mutation for N-Glycosylation Sites Altered the Biophysical Properties of Cav1.2

We additionally generated double-site mutants (N124,299Q and N1359,1410Q) and a quadruple-site mutant (N124,299,1359,1410Q) to examine whether these mutations of the N-glycosylation sites have effects on channel expression and biophysical properties. When an equal amount (5 ng) of each cRNA for Cav1.2/WT, the double mutants, and QM was coinjected with β3 cRNA (2.5 ng) into oocytes, strong inward currents were recorded from the oocytes from the 3rd day after cRNA injection (Figs. 3A-3D). Ba2+ currents of the double mutants (N124,299Q and N1359,1410Q) evoked by the I-V protocol were not significantly different from Cav1.2/WT currents in amplitude, whereas Ba2+ currents of QM (N124,299,1359,1410Q) were significantly smaller than Cav1.2/WT currents (Figs. 3A-3E; one-way ANOVA Tukey tests, p < 0.05, n = 9-14).

Fig. 3.Current traces and I-V relationships of the double and quadruple mutants. (A-D) Representative current traces of Cav1.2/WT (control), and double and quadruple mutant channels. In 10 mM Ba2+ solution, currents were evoked by a voltage protocol composed of serial voltage steps ranging from -40 mV to +60 mV by 10 mV increments, from a holding potential of -80 mV. Superposed current traces of Cav1.2/WT (A), N124,299Q (B), N1359,1410Q (C), and QM (D) elicited by test potentials of -40, -10, +10, and +30 mV from a holding potential of -80 mV are represented. (E) Current-voltage (I-V) relationships of the Cav1.2/WT, double mutants, and QM. The average current amplitudes of Cav1.2/WT (○, n = 14), N124,299Q (□, n = 9), N1359,1410Q (△, n = 9), and QM (△, n = 9) were plotted against test potentials, as described in Fig. 2F. Statistical differences are indicated with asterisks (*, p < 0.05, one-way ANOVA Tukey tests, n = 9-14). (F) Normalized current-voltage (I-V) relationships of double and quadruple mutants and Cav1.2/WT. Current amplitudes of Cav1.2/WT, N124,299Q, N1359,1410Q, or QM elicited by a voltage protocol were normalized to the maximal current amplitude observed and then normalized average values were plotted against test potentials. (G) Activation curves of Cav1.2/WT, double mutants, and QM. Activation curves were derived from fitting cord conductance (G) as explained in the Methods section. The V50,act values (50% activation potential) and slope factors (k) for Cav1.2/WT, N124,299Q, N1359,1410Q, and QM were estimated from the fittings using the Boltzmann equation. The V50,act values for N124,299Q and QM a significantly different from those of Cav1.2/WT and N1359,1410Q (one-way ANOVA Tukey tests, p < 0.05, n = 9-14). (H) Steady-state inactivation curves of Cav1.2/WT, double mutants, and QM. The V50,inact values for Cav1.2/WT, N124,299Q, N1359,1410Q, and QM were estimated to be -18.4 ± 0.9, -16.8 ± 0.9, -18.5 ± 1.1, and -17.0 ± 0.9 and their slope factors (k) were -10.5 ± 0.7, -10.2 ± 0.6, -11.3 ± 1.1, and -10.4 ± 0.9, respectively. Significant differences were not detected among them (one-way ANOVA Tukey tests, n = 8-12).

When the I-V data of the double and quadruple mutants were normalized to that of Cav1.2/WT for comparison (Figs. 3E and 3F), the I-V relationships of N124,299Q and QM were shown to be positively shifted, compared with that of Cav1.2/WT. Their activation curves were from fitting their cord conductance(s) calculated from the I-V data (Fig. 3G). The V50,act values for Cav1.2/WT, N124,299Q, N1359,1410Q, and QM were -6.45 ± 0.15, -0.98 ± 0.13, -6.18 ± 0.52, and -0.91 ± 0.24 mV, respectively. Statistical analysis showed that the V50,act values for N124,299Q and QM were significantly different from that of Cav1.2/WT (one-way ANOVA Tukey tests, p < 0.05), but their slope factors were not significantly different from Cav1.2/WT. This finding consistently suggests that the activation curves of N124,299Q and QM were significantly shifted to the depolarizing direction. In comparison, the double mutations in domain IV did not cause any significant alterations in current amplitude and voltage-dependent gating.

The steady-state inactivation curves of Cav1.2/WT, N124,299Q, N1359,1410Q, and QM were obtained from fitting the data to the Boltzmann equation (Fig. 3H). The V50,inact values and slope factors for Cav1.2/WT, N124,299Q, N1359,1410Q, and QM were not significantly different, suggesting that the voltage dependency for steady-state inactivation was not significantly altered by the double or quadruple mutations.

Taken together, electrophysiological characterizations of the mutant channels indicate that the single-site mutations did not significantly alter the biophysical properties of Cav1.2, but the double mutation in domain I caused a positive shift of the activation curve. Moreover, the quadruple mutation caused a positive shift of the activation curve and a decrease of current.

Less Surface Expression of EGFP-QM than EGFP-Cav1.2/WT

To explore the mechanism(s) underlying the reduced currents of QM compared with Cav1.2/WT and the double mutants, we engineered to tag EGFP cDNA to the 5’ end of Cav1.2/WT cDNA, N124,299Q cDNA, N1359,1410Q cDNA, and QM cDNA. An equal amount (5 ng) of EGFP-Cav1.2/ WT cRNA, EGFP-N124,299Q cRNA, EGFP-N1359,1410Q cRNA, or EGFP-QM cRNA was coinjected with β3 cRNA (2.5 ng) into oocytes. Localization of EGFP-Cav1.2/WT, EGFP-N124,299Q, EGFP-N1359,1410Q, and EGFP-QM in the oocyte plasma membrane was observed via their fluorescence images under a confocal microscope on the 4th day after cRNA injection (Fig. 4A). Analysis of the fluorescence images at the oocyte membrane showed that the average fluorescence intensity for EGFP-QM was significantly weaker than those for EGFP-Cav1.2/WT, EGFP-N124,299Q, and EGFP-N1359,1410Q (Fig. 4B). These findings suggest that the smaller current of QM than the Cav1.2/WT and double-mutant channels may arise from the lower surface expression of QM than Cav1.2/WT.

Fig. 4.Surface expression of EGFP-Cav1.2/WT, EGFP-double mutants, and EGFP-QM. (A) Representative confocal images of fluorescence of oocytes expressing EGFP-Cav1.2/WT, EGFP-N124,299Q, EGFP-N1359,1410Q, or EGFPQM. Surface fluorescence images were taken from oocytes injected with cRNA (5 ng) of EGFP-Cav1.2/WT, EGFP-N124,299Q, EGFPN1359,1410Q, or EGFP-QM with β3 cRNA (2.5 ng), using a Zeiss confocal microscope. (B) Relative fluorescent intensities of EGFPtagged Cav1.2 constructs at the oocyte surface. The fluorescence intensities at the surface of oocytes expressing EGFP-Cav1.2/WT, EGFPN124,299Q, EGFP-N1359,1410Q, or EGFP-QM were quantitated using the ImageJ program. Their average fluorescence values (arbitrary unit) were 124.6 ± 18.3, 114.6 ± 15.7, 135.6 ± 17.8, and 64.1 ± 7.1, respectively, which are represented in the bar graphs. Significant differences were detected between the fluorescence intensities of EGFP-QM and the other groups (one-way ANOVA Tukey tests, p < 0.05, n = 8-10).

Tunicamycin Treatment of Cav1.2/WT Mimics the Quadruple Mutation Effects

To examine whether treatment of tunicamycin, a specific inhibitor of N-linked glycoslylation via inhibiting Nacetylglucosamine transferase, can affect the biophysical properties of Cav1.2/WT, we prepared two sets of oocytes. A control set of oocytes were injected with only a cRNA mixture of Cav1.2/WT and β3 , whereas the other set of oocytes were coinjected with tunicamycin (3 ng/oocyte) and cRNA mixture of Cav1.2/WT and β3. From 3-4 days after injection, we recorded currents from the two groups of oocytes. Analysis of I-V data (Fig. 5A) showed that the currents of the tunicamycin-treated group were significantly smaller than those of the control Cav1.2/WT group over most of the test potentials (Student's unpaired t tests, p < 0.05, n = 10-12), implying that the current reduction may be caused by disruption of N-glycosylation on Cav1.2/WT by the drug. In addition, the I-V of the tunicamycin-treated group seems to be positively shifted, compared with that of the control group (Fig. 5A). This positive shift of I-V curves was consistently detected in their activation curves (Fig. 5C). The V50,act values for the Cav1.2/WT and the tunicamycintreated Cav1.2/WT were -6.4 ± 0.3 mV and -2.0 ± 0.4 mV, and their slopes were 6.7 ± 0.3 and 6.8 ± 0.3, respectively. The V50,act value for the tunicamycin-treated Cav1.2/WT was significantly bigger than that of the Cav1.2/WT, suggesting that tunicamycin positively shifted the activation curve of Cav1.2/WT.

Fig. 5.Effects of tunicamycin on Cav1.2/WT and QM. A control group of oocytes were injected with only a cRNA mixture of Cav1.2/WT (A) or QM (B) and β3 , whereas the other group of oocytes were injected with tunicamycin (3 ng/oocyte) as well as cRNA mixture of Cav1.2/WT or QM and β3. (A-B) Effects of tunicamycin on the I-V curves of Cav1.2/WT and QM. Ba2+ currents through Cav1.2/WT and QM were elicited by the same voltage protocol used in Fig. 1. They were averaged and then plotted against test potentials. The symbols and sample numbers are as follows: Cav1.2/WT control (○, n = 12); tunicamycin-treated Cav1.2/WT (●, n = 10); QM control (□, n = 8); tunicamycin-treated QM (■, n = 7). Statistical differences were marked with asterisks (*, p < 0.05, Student’s unpaired t-tests). (C) Activation curves of the control groups and the test groups treated with tunicamycin. Activation curves were obtained as described in the Methods section. Their V50,act values were significantly different (Student’s unpaired t test, p < 0.05, n = 12, 10). The V50,act values for the QM control and the tunicamycin-treated QM were -1.5 ± 0.4 mV and -1.2 ± 0.4 mV and their slope factors 6.6 ± 0.4 and 6.9 ± 0.6, respectively. No significant difference was detected (Student's unpaired t test, p > 0.1, n = 8, 7). (D) Tunicamycin effects on the steady-state inactivation properties of Cav1.2/WT and QM. The V50,inact values for Cav1.2/WT, tunicamycin-treated Cav1.2/WT, QM, and tunicamycin-treated QM were -18.0 ± 0.9, -16.8 ± 1.1, -16.2 ± 1.0, and -16.3 ± 0.9 and their slope factors (k) were -10.7 ± 0.7, -10.2 ± 0.6, -11.3 ± 1.1, and -10.4 ± 0.9, respectively. Significant differences were not detected among the parameters (Student’s unpaired t tests, n = 10-12).

Next experiments were performed to compare steadystate inactivation properties between the control Cav1.2/WT group and the tunicamycin-treated Cav1.2/WT. Analysis of their steady-state inactivation curves (Fig. 5D) showed that the V50,inact values of the control Cav1.2/WT group and the tunicamycin-treated Cav1.2/WT groups were -18.0 ± 0.9 and -16.8 ± 1.1, and their slope factors (k) were -10.7 ± 0.7 and -10.2 ± 0.6, respectively. Significant differences were not found between the relevant values of the two groups (Student's unpaired t tests). These results suggest that tunicamycin treatment mimics the effects caused by the quadruple mutations of potential N-glycosylation sites, supporting that N-glycosylation can significantly affect the surface expression and voltage-dependent gating of Cav1.2.

To elucidate the decrement effect on Cav1.2/WT currents by treatment of tunicamycin (Fig. 5A), we examined whether tunicamycin can affect the surface expression of EGFP-tagged Cav1.2/WT in oocytes. Analysis of fluorescence images in the membrane regions of oocytes showed that the average fluorescent intensity (117.6 ± 18.6) of the control group injected with the cRNA mixture (EGFPCav1.2/WT and β3) was significantly greater than the intensity (70.6 ± 11) of the test group injected with the cRNA mixture plus tunicamycin (Fig. S2). The reduced fluorescence intensity of the tunicamycin group suggests that the drug might disrupt N-glycosylation of the channels, resulting in a significant decrease in the surface expression of EGFP-Cav1.2/WT. This may be a possible underlying mechanism for the reduction effects on Cav1.2/WT currents by tunicamycin.

We finally examined whether tunicamycin can alter the biophysical properties of QM. Analysis of the I-V data showed that tunicamycin did not alter the current amplitudes of QM over most of the test potentials (Fig. 5B). Analysis of the activation curves and the steady-state inactivation curves displayed that the drug did not significantly change the voltage dependency for activation and steady-state inactivation (Figs. 5C and 5D). These results suggest that tunicamycin did not affect the biophysical properties of QM in which potential N-glycosylation sites were abolished by site-directed mutagenesis.

 

Discussion

We here showed that single mutations of the potential Nglycosylation sites on Cav1.2 did not have significant effect(s) on the biophysical properties of Cav1.2. It is thus limited to specify function(s) of each of the four potential N-glycosylation sites. In comparison, double mutations of the two sites in domain I, but not in domain IV, caused a depolarizing shift in voltage-dependent gating. Furthermore, the quadruple mutation caused a strong reduction of current amplitude as well as a positive shift of voltagedependent gating. Consistently, the tunicamycin treatment mimicked the biophysical effects led by the quadruple mutation, whereas it did not affect those of QM in which all the potential N-glycosylation sites were abolished by sitedirected point mutations. Taken together, these findings suggest that N-glycosylation would contribute to the surface expression and voltage-dependent gating of Cav1.2/WT.

Our finding that disruption of the two N-glycosylation sites in domain I of the Cav1.2 induced a positive shift in voltage-dependent gating is consistent with the effects found in Nav1.4 [1]. Negatively charged glycan(s) are attached to nitrogen atoms of Asn residues in N-linked glycosylation, potentially contributing to the creation of a negative potential on the extracellular surface. Because elimination of N-glycosylation can reduce the negative surface potential formed by the attachment of glycan(s), depolarized potential needs to be applied for channel gating to overcome the reduced negative surface potential. This may account for the depolarizing shifts in channel gating, which were detected in the quadruple mutant and the double mutant in domain I of Cav1.2.

Among the four potential N-glycosylation sites, three sites (N299, N1359, and N1410) are positioned at the pore loops (S5-S6) of domains I and IV (Fig. 1). Although single mutations of them were not sufficient to alter the functions of Cav1.2, we speculate that the three sites critically influence the voltage-dependent gating and/or the surface expression of Cav1.2 in a distinctive way. Serial construction and characterization of pore N-glycosylation mutants (N124,1359Q, N299,1410Q, and N124,1359,1410Q) remain to be further undertaken in the future, which may provide important information to unveil the underlying mechanisms for the functions of N-glycosylation.

Fluorescence imaging studies of EGFP-QM, EGFP-double mutants, and EGFP-Cav1.2/WT suggest that the reduced currents of EGFP-QM arise from its reduced surface expression probably via decreased membrane targeting of EGFP-QM proteins after translation, compared with the EGFP-Cav1.2/WT and EGFP-double mutants. However, it remains to further investigate the detailed mechanisms for the reduced surface expression, which might involve various steps as follows. First, we speculate that disruption of N-glycosylation sites on the Cav1.2 might negatively affect the structural folding of Cav1.2 proteins after protein synthesis. Second, improperly folded Cav1.2 proteins and/or N-glycosylation-deficient Cav1.2 proteins might be less efficiently targeted to the membrane as well. Third, improperly folded Cav1.2 proteins or N-glycosylationdeficient Cav1.2 proteins might be more labile for degradation.

The pathophysiological relevance of N-glycosylationdeficient Cav1.2 can be found in patients with congenital disorders of N-linked glycosylation (CDG) [4, 10]. The patients would have membrane proteins lacking of Nglycosylation. Considering that Cav1.2 channels are expressed in the cardiovascular system, endocrine system, nervous system, and diverse peripheral systems, CDG patients would be severely implicated in cardiovascular malfunctions and broad types of pathophysiological conditions in the other systems, because N-glycosylation deficiency in Cav1.2 would cause a reduction of current via reduced surface expression and a positive shift in voltage-dependent gating of Cav1.2. In conclusion, our combined approaches of site-directed mutagenesis, electrophysiology, and confocal imaging studies support that N-glycosylation critically contributes to membrane trafficking for the functional expression and voltage-dependent gating of Cav1.2.