Small GTPase Rab11b regulates degradation of surface membrane L-type Cav1.2 channels

Jabe M. Best, Jason D. Foell, Courtney R. Buss, Brian P. Delisle, Ravi C. Balijepalli, Craig T. January, Timothy J. Kamp


L-type Ca2+ channels (LTCCs) play a critical role in Ca2+-dependent signaling processes in a variety of cell types. The number of functional LTCCs at the plasma membrane strongly influences the strength and duration of Ca2+ signals. Recent studies demonstrated that endosomal trafficking provides a mechanism for dynamic changes in LTCC surface membrane density. The purpose of the current study was to determine whether the small GTPase Rab11b, a known regulator of endosomal recycling, impacts plasmalemmal expression of Cav1.2 LTCCs. Disruption of endogenous Rab11b function with a dominant negative Rab11b S25N mutant led to a significant 64% increase in peak L-type Ba2+ current (IBa,L) in human embryonic kidney (HEK)293 cells. Short-hairpin RNA (shRNA)-mediated knockdown of Rab11b also significantly increased peak IBa,L by 66% compared when with cells transfected with control shRNA, whereas knockdown of Rab11a did not impact IBa,L. Rab11b S25N led to a 1.7-fold increase in plasma membrane density of hemagglutinin epitope-tagged Cav1.2 expressed in HEK293 cells. Cell surface biotinylation experiments demonstrated that Rab11b S25N does not significantly impact anterograde trafficking of LTCCs to the surface membrane but rather slows degradation of plasmalemmal Cav1.2 channels. We further demonstrated Rab11b expression in ventricular myocardium and showed that Rab11b S25N significantly increases peak IBa,L by 98% in neonatal mouse cardiac myocytes. These findings reveal a novel role for Rab11b in limiting, rather than promoting, the plasma membrane expression of Cav1.2 LTCCs in contrast to its effects on other ion channels including human ether-a-go-go-related gene (hERG) K+ channels and cystic fibrosis transmembrane conductance regulator. This suggests Rab11b differentially regulates the trafficking of distinct cargo and extends our understanding of how endosomal transport impacts the functional expression of LTCCs.

  • ion channels
  • endosomal trafficking
  • electrophysiology
  • heart

calcium ion entry through voltage-gated Cav1.2 L-type channels (ICa,L) regulates numerous physiological processes in diverse tissues including glucose-stimulated insulin secretion from pancreatic β-cells, neurotransmitter release, long-term potentiation in neurons, and excitation-contraction coupling in smooth and cardiac muscle (5, 13, 14, 27, 28, 29, 30, 39). ICa,L must be finely tuned to permit appropriate levels of intracellular Ca2+ for proper cellular function. One critical determinant of ICa,L is the number of functional channels at the plasma membrane, and changes in the density of L-type Ca2+ channels (LTCCs) have been observed in aging and various diseases (3, 9, 10, 18, 21, 40). For example, in congestive heart failure, there is a significant reduction in the number of plasmalemmal LTCCs in ventricular myocytes, which partially underlies the characteristic dysregulation of Ca2+ homeostasis and excitation-contraction coupling (9, 18, 21). More clearly defining the mechanisms that regulate surface membrane density of Cav1.2 LTCCs will contribute to our understanding of the pathophysiology of such prevalent diseases.

Endosomal trafficking represents an important mechanism for maintaining proper plasmalemmal expression of some ion channels (12, 17, 26, 34, 35). Endosomal transport is largely regulated by the Rab family of small GTPases, within which there are over 60 members with varying tissue distribution, and individual Rab GTPases coordinate distinct steps of anterograde and retrograde protein trafficking (38, 44). Previous work in our lab using dominant negative mutants that block endogenous Rab function demonstrated that Rab11b is critical for robust plasmalemmal expression of human ether-a-go-go-related gene (hERG) K+ channels in human embryonic kidney (HEK)293 cells (12). Rab11b is a member of the Rab11 subfamily that includes Rab11a and Rab25 (7) and has been previously shown to regulate endosomal recycling of transferrin (Tfn) receptors to the plasma membrane in various epithelial cell lines (31, 33). Whether Rab GTPases significantly impact surface membrane density of LTCCs remains unknown.

An important role for endosomal trafficking of LTCCs has been suggested by recent studies (16). Sustained depolarization of rat cortical neurons using KCl led to a reversible, Ca2+-dependent internalization of Cav1.2 channels into endosomal vesicles (16). Further analysis revealed these Cav1.2-containing vesicles mimicked the size, localization, and movement of Tfn receptor-containing vesicles, which recycle to the plasma membrane through both sorting endosomal and recycling endosomal compartments (25). Whether Rab GTPases regulate this trafficking and whether these pathways represent a universal mechanism for modulating surface membrane expression of LTCCs in other excitable cells such as cardiac myocytes remain important questions. Because Cav1.2-containing vesicles shared many characteristics with Tfn receptor cargo vesicles in neurons (16), we hypothesized that Rab11b regulates the trafficking and functional expression of Cav1.2 L-type Ca2+ channels. To address this possibility we employed various electrophysiological and biochemical approaches in a HEK293 expression system and in neonatal mouse ventricular myocytes.


Expression plasmids and molecular biology.

Plasmids encoding the human pore-forming α1C,77 (Cav1.2) and auxiliary β2cN4 LTCC subunits have been previously described (15, 36). Hemagglutinin (HA)-Cav1.2 was engineered from α1C,77 to contain an extracellular HA epitope within the S5-H5 loop of transmembrane segment II following a previously outlined strategy (1). Briefly, the EcoRI fragment corresponding to nucleotides 775–2181 of α1C,77 was subcloned into pCRII-TOPO (Invitrogen, Carlsbad, CA) and used as a template for site-directed mutagenesis (QuikChange, Agilent Technologies, Santa Clara, CA) to insert double silent mutations (C>G and G>T, corresponding to nucleotides 2052 and 2055 of α1C,77, respectively), creating a unique MluI restriction enzyme site. Oligonucleotides (sense 5′-CGCGTCATTACCCATACGACGTCCCAGACTACGCTGTCACGTTTGATGAGATGCAGA-3′; antisense 5′-CGCGTCTGCATCTCATCAAACGTGACAGCGTAGTCTGGGACGTCGTATGGGTAATGA-3′) encoding the amino acid sequence TRHYPYDVPDYAVTFDEMQ (HA epitope in bold) were synthesized by Integrated DNA Technologies (Coralville, IA), annealed, and ligated into the newly created MluI site within the pCRII-α1C,77 EcoRI fragment. The EcoRI fragment containing the HA insert was then religated into the α1C,77 backbone. Correct insertion of the HA-encoding oligonucleotide was verified by DNA sequencing (University of Wisconsin Biotechnology Center, Madison, WI), and the integrity of HA-Cav1.2 was tested in HEK293 cells using the whole cell patch-clamp technique, which yielded IBa,L comparable to that observed from wild-type (WT) Cav1.2 channels (data not shown). Constructs encoding WT Rab11b and the dominant negative Rab11b S25N or GTPase-deficient Rab11b Q70L mutants modified to contain green fluorescent protein (GFP) fused at their NH2-termini were kindly provided by Dr. Beate Schlierf (Institut für Biochemie, Universität Erlangen-Nürnberg, Erlangen, Germany) (33), and the GFPpRK5 plasmid expressing the S65T bright GFP variant (15) was used as a transfection control. Rab11b- and Rab11a-specific short-hairpin RNA (shRNA) plasmids containing a GFP marker were obtained from SABiosciences (Frederick, MD).

Cell culture and transfection.

HEK293 cells were transfected with WT Cav1.2 or HA-Cav1.2, β2cN4, and either GFP or GFP-Rab11b fusion constructs in a 1:1: 0.1 molar ratio using Lipofectamine 2000 (Invitrogen) and cultured in DMEM + 10% fetal bovine serum, 2 mM l-glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin at 37°C + 5% CO2. HEK293 cell experiments were performed 40–48 h after transfection unless otherwise indicated. Neonatal mouse ventricular myocytes were isolated by enzymatic digestion as previously described (4) and nucleofected using 2 μg plasmid DNA (Lonza, Walkersville, MD). Myocytes were cultured in DMEM + Medium 199 (at a 4:1 vol/vol ratio) + 0.5% fetal bovine serum, 0.5% horse serum, 2 mM l-glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin at 37°C + 5% CO2 for 72–96 h before experiments.


Whole cell ruptured patch-clamp studies were performed on HEK293 cells and neonatal mouse cardiac myocytes at 21–24°C. The pipette solution consisted of (in mM) 114 CsCl, 10 EGTA, 10 HEPES, 5 Mg-ATP, and 0.1 Li-GTP (to pH 7.2 with CsOH). The recording solution contained (in mM) 133 CsCl, 10 BaCl2, and 10 HEPES (to pH 7.4 with CsOH). Upon electrical access, capacitive transients were obtained by applying a 5-mV hyperpolarizing pulse. Whole cell currents were elicited from a holding potential of −80 mV and recorded using an Axopatch 200B amplifier (Molecular Devices, Sunnyvale, CA). Data were sampled at 25 kHz and filtered at 5 kHz, and leak and capacitive currents were subtracted using a P/4 protocol. Data are presented as current density by normalizing IBa,L to cell capacitance.

Immunodetection and quantitation of plasma membrane HA-Cav1.2.

HEK293 cells were transfected and cultured in 12-well tissue culture plates. Cells were briefly washed in phosphate-buffered saline (PBS) before fixation in 4% paraformaldehyde for 10 min. After being blocked for 1 h in 2% bovine serum albumin + 1% goat serum in PBS, cells were incubated in anti-HA (1:200, Clone 3F10, Roche Applied Science, Indianapolis, IN) for 1 h at room temperature. Cells were then washed four times in blocking solution before incubation in goat anti-Rat Alexa Fluor 680 (1:500, Invitrogen) for 30 min. In parallel control experiments to assess cell density, secondary antibody was omitted and cells were incubated in 5 μg/ml wheat germ agglutinin (WGA) Alexa Fluor 680 (Invitrogen) for 10 min at room temperature. Cells were washed four times before imaging using the LI-COR Odyssey Infrared Imaging System. Images were analyzed using ImageJ (NIH, Bethesda, MD) by selecting regions of interest within each well and calculating total infrared signal. Wells in which primary antibody was omitted were used for background infrared signal subtraction, and the values for each experiment were normalized to the GFP control group. Alternatively, single cell imaging was performed using a Bio-Rad MRC-1024 confocal microscope equipped with a Nikon Diaphot 300 microscope with a ×60 1.3 numerical aperature oil immersion lens. For confocal microscopy, goat anti-Rat Alexa Fluor 568 was used for immunodetection of HA-Cav1.2. Images were acquired using LaserSharp software at 1.0-μm intervals along the z-axis, and representative images were chosen for presentation.

Cell surface biotinylation and degradation assay.

For cell surface biotinylation experiments, transfected HEK293 cells grown in 10-cm tissue culture dishes were cooled to 4°C, washed three times in ice-cold PBS, and biotinylated for 30 min with 1–1.5 mg/ml EZ Link Sulfo-NHS-LC-Biotin (Thermo Scientific Pierce Protein Research Products, Rockford, IL) at 4°C. Cells were then washed twice in ice-cold Tris-buffered saline (TBS) followed by an additional wash in PBS to remove biotinylation reagent. Cells were either harvested directly by scraping or, in the case of the degradation assay, fresh culture media was added to the plates and cells were returned to 37°C + 5% CO2 for indicated times. For these samples, cells were washed twice in PBS before being harvested. Cells were lysed in buffer containing 150 mM NaCl, 50 mM Tris·HCl (pH 7.4), 2 mM EGTA, 1% Triton X-100, 1% IGEPAL, and protease inhibitor cocktail containing 2 mM phenylmethylsulfonyl fluoride, 50 μg/ml aprotinin, 50 μg/ml benzamidin, 50 μg/ml leupeptin, and 5 μM pepstatin A.

Tfn receptor recycling assay.

HEK293 cells transfected with Cav1.2, β2cN4, and either Rab11b WT or Rab11b S25N were loaded with 5 μg/ml human Tfn-Alexa Fluor 594 conjugate (Invitrogen) in Opti-MEM (Invitrogen) for 1 h at 37°C + 5% CO2. Cells were then washed in Opti-MEM and either imaged directly (t = 0) or returned to 37°C + 5% CO2 for 40 min to allow recycling of Tfn receptor and dissociation of Tfn-Alexa Fluor 594 from the cells. Fluorescent confocal microscopy was performed on live cells using a Bio-Rad MRC 1024 confocal microscope.


HEK293 cell lysates (0.5–1 mg protein) were used for immunoprecipitation (IP) with anti-HA (0.5 μg) by incubating with 40 μl Protein G agarose (GE Healthcare Life Sciences, Piscataway, NJ) at 4°C overnight. In some experiments, biotinylated proteins were captured using NeutrAvidin. Soluble protein (1–1.5 mg) was incubated with 75 μl high capacity NeutrAvidin agarose beads (Thermo Scientific Pierce Protein Research Products) for 2 h at 4°C. Protein G or NeutrAvidin agarose beads were washed, and proteins were eluted by heating to 95°C for 5 min in SDS-PAGE sample buffer.

SDS-PAGE and Western blotting.

Adult mouse left ventricular and whole brain tissues were dissected and homogenized in ice-cold lysis buffer containing 150 mM NaCl, 20 mM Tris·HCl (pH 7.4), 1% Triton X-100, 1% deoxycholic acid, and protease inhibitor cocktail. Protein samples were separated using SDS-PAGE and transferred to polyvinylidene difluoride membranes. Nonspecific binding sites were blocked using 5% (wt/vol) dried skim milk in TBS with 0.05% Tween-20 (TBST). Membranes were probed with the following antibodies: rabbit anti-Rab11b (see Ref. 24 and sc-133937, Santa Cruz Biosciences, Santa Cruz, CA), rabbit anti-Rab11a (Invitrogen), rat anti-HA, rabbit anti-GFP (sc-8834, Santa Cruz Biotechnology), goat anti-biotin (ab6643, Abcam, Cambridge, MA), and mouse anti-transferrin receptor (clone H68.4, Invitrogen) overnight at room temperature. Membranes were subsequently washed four times for 10 min before incubation in appropriate horseradish peroxidase-conjugated secondary antibody for 1 h. After additional washes in TBST, immunoreactivity was visualized using ECL Plus (GE Healthcare Life Sciences). Membrane stripping was performed at 55°C for 30 min in buffer containing 62.5 mM Tris·HCl, pH 6.8, 2% SDS, 100 mM β-mercaptoethanol, and 100 mM DTT.


Adult mouse ventricular myocytes were isolated by enzymatic digestion as previously described (4, 42). Total RNA was isolated using RNAzole B (IsoTex Diagnostics, Friendswood, TX) and served as a template for cDNA synthesis using Superscript III Reverse Transcriptase (Invitrogen, Carlsbad, CA). PCR was performed using 5.0 μl cDNA in a reaction containing 20 mM Tris·HCl (pH 8.8), 10 mM KCl, 10 mM (NH4)2SO4, 2.0 mM MgSO4, 0.1% Triton X-100, 0.1 mg/ml BSA, 10 nmol each dATP, dCTP, dGTP, dTTP, 75 pmol of each primer, 5 U Taq Extender Additive (Agilent Technologies), and 2.5 U Taq DNA polymerase (Fisher Scientific, Pittsburgh, PA). PCR reactions were initially denatured at 94°C for 3 min and then cycled at 94°C for 45 s, 55°C for 30 s, and 72°C for 2 min 25 times, followed by a final extension for 7 min at 72°C. Primer sets are as follows (5′ to 3′): fRab11b (NM_008997) TCAGATCTGGGACACTGCTG, rRab11b GGCTGAGGTCTCAATGAAGG, fβ-actin (NM_007393) GATTACTGCTCTGGCTCCTAG, rβ-actin AAGGGTGTAAAACGCAGCTCA, fMAP2 (Mtap2, NM_001039934) CTGGACATCAGCCTCACTCA, and rMAP2 AATAGGTGCCCTGTGACCTG. PCR products were analyzed on a 1% agarose gel with SYBR Safe DNA gel stain (Invitrogen) under UV light.

Statistical analysis.

Data from each group are reported as means ± SE. ANOVA was performed to determine statistical significance between multiple groups followed by post hoc Bonferroni means comparison when applicable to determine whether an experimental group was significantly different from control group using Origin 7.5 software (OriginLab, Northhampton, MA). Statistical significance was defined as P < 0.05.


Inhibition of Rab11b function using dominant negative Rab11b S25N mutant or Rab11b-specific shRNA increases IBa,L.

A subpopulation of Cav1.2 LTCCs has been detected in highly mobile endosomal structures (16), but the mechanisms regulating the endosomal trafficking of the channels are not well defined. Because the small GTPase Rab11b regulates endosomal recycling, we sought to test whether Rab11b regulates functional expression of Cav1.2 L-type channels. We transiently expressed LTCC subunits Cav1.2 and β2cN4 along with GFP, Rab11b WT, or dominant negative Rab11b S25N in HEK293 cells and recorded IBa,L using whole cell ruptured patch-clamp technique. Representative IBa,L measurements show that cells coexpressing Rab11b S25N yielded larger IBa,L when compared with cells cotransfected with GFP or Rab11b WT (Fig. 1A). Rab11b S25N cells demonstrated a 64% larger peak IBa,L at a test potential of 10 mV when compared with GFP control cells (−68 ± 6 pA/pF, n = 12 vs. −42 ± 3 pA/pF, n = 14; *P < 0.05) (Fig. 1B). Based on the mean current-voltage (I-V) relations, Rab11b S25N did not alter the voltage dependence of activation gating for IBa,L when compared with GFP or Rab11b WT cells. Among cells selected for electrophysiological recordings, whole cell capacitance was not significantly different among the three groups (GFP, 19.0 ± 1.4 pF; Rab11b, WT 17.2 ± 0.8 pF; Rab11b S25N, 16.5 ± 1.3 pF), suggesting that Rab11b S25N does not alter average cell membrane surface area.

Fig. 1.

Disruption of endogenous Rab11b function with dominant negative Rab11b S25N mutant increases L-type Ba2+ current (IBa,L). A: representative IBa,L traces recorded from human embryonic kidney (HEK)293 cells transiently expressing wild-type (WT) Cav1.2, β2cN4, and either green fluorescent protein (GFP), Rab11b WT, or Rab11b S25N using whole cell rupture patch-clamp technique. B: mean current-voltage (I-V) relations constructed from GFP (n = 14), Rab11b WT (n = 4), and Rab11b S25N (n = 12) expressing cells. C: representative IBa,L traces from HEK293 cells transfected with WT Cav1.2, β2cN4, and either GFP, Rab11b WT, GTPase deficient Rab11b Q70L, or Rab11b S25N. D: mean I-V relations from GFP (n = 6), Rab11b WT (n = 5), Rab11b Q70L (n = 8), and Rab11b S25N (n = 8) expressing cells. *P < 0.05.

Because blocking Rab11b function with a dominant negative mutant led to a significant increase in IBa,L, we performed a second series of experiments testing a GTPase-deficient (33) Rab11b Q70L mutant, which has been shown to act in a dominant active fashion in T84 intestinal epithelial cells (35). Coexpression of Rab11b Q70L did not significantly alter IBa,L when compared with GFP control cells and likewise overexpression of Rab11b WT had no effect (Fig. 1, C and D). The stimulatory effect of Rab11b S25N on IBa,L was reproducible, however (Fig. 1, C and D).

To investigate the specificity of the Rab11b S25N effects, we next tested shRNA-mediated knockdown of Rab11b as well as knockdown of the related Rab11a protein in HEK293 cells. Rab11a differs from Rab11b only in its final 30 amino acids, and because of this high degree of homology, activated Rab11b and Rab11a share several downstream effectors (7, 20). Despite the similarity between Rab11b and Rab11a, these two proteins have been shown to localize to distinct subcellular vesicular compartments and differentially regulate some transmembrane proteins including cystic fibrosis transmembrane conductance regulator (CFTR) (24, 35). shRNA-mediated knockdown of endogenous Rab11b or Rab11a in HEK293 cells was confirmed by Western blot analysis showing specific knockdown of Rab11b or Rab11a protein, respectively (Fig. 2A). GAPDH was used as a loading control, whereas probing for GFP showed comparable transfection efficiency among the shRNA-transfected samples. Densitometric analysis of multiple samples (n = 3–4) revealed transfection with Rab11b-specific shRNA led to a significant 84% reduction in Rab11b protein level (Fig. 2B), while expression of Rab11a-specific shRNA caused a 72% depletion of Rab11a (Fig. 2C) compared with cells transfected with control shRNA.

Fig. 2.

Short-hair pin RNA (shRNA)-mediated depletion of Rab11b increases IBa,L. A: Western blot analysis of HEK293 cells transfected with control shRNA or shRNA specific for Rab11b or Rab11a. To assess the relative knockdown of Rab11 isoforms, densitometric analysis was performed. Rab11b (B, n = 3) or Rab11a (C, n = 4) levels were normalized to GAPDH and compared with control shRNA. D: IBa,L recordings from HEK293 cells transfected with Cav1.2, β2cN4, and either control shRNA, Rab11b shRNA, or Rab11a shRNA. E: mean I-V relations constructed from control shRNA (n = 12), Rab11b shRNA (n = 13), and Rab11a shRNA (n = 11) expressing cells. *P < 0.05.

To examine the effect of shRNA-mediated depletion of Rab11b versus Rab11a on the functional expression of L-type Cav1.2 channels, whole cell patch-clamp experiments were performed on HEK293 cells transfected with Cav1.2, β2cN4, and either control shRNA, Rab11b shRNA, or Rab11a shRNA. Representative IBa,L traces show that depletion of Rab11b, but not Rab11a, led to increased L-type current density compared with cells expressing control shRNA (Fig. 2D). On average, cells expressing Rab11b shRNA yielded 66% larger inward peak IBa,L compared with control shRNA-expressing cells (Fig. 2E). In contrast, cells expressing Rab11a shRNA did not demonstrate any significant change in peak IBa,L relative to control. Thus loss of Rab11b function by either shRNA or the dominant negative Rab11b S25N leads to an increase in functional expression of LTCCs in this model, and this effect is specific to Rab11b compared with Rab11a.

Dominant negative Rab11b S25N increases surface membrane density of L-type Cav1.2 channels.

Because the primary known function of Rab11b is regulating the trafficking of membrane proteins (12, 31, 33, 35), we hypothesized that the increased IBa,L observed after disruption of Rab11b function was due to enhanced surface membrane expression of Cav1.2 channels. To test this we developed an assay to quantitatively measure Cav1.2 channels inserted into the plasma membrane and test the effect of loss of Rab11b function by expression of Rab11b S25N. This assay is based on immunolabeling intact, nonpermeabilized cells expressing HA-Cav1.2 (extracellular HA epitope within the S5-H5 linker of transmembrane segment II) with anti-HA. A confocal image of a representative cell immunolabeled with anti-HA confirmed that extracellular application of primary antibody specifically labels plasma membrane-associated HA-Cav1.2 channels in punctate clusters without appreciable intracellular labeling (Fig. 3A). Using a similar approach that we previously employed to quantitatively analyze plasma membrane expression of hERG K+ channels (12), we measured HA-Cav1.2 immunoinfrared signal from a large population of cells. Figure 3B (top row) shows a representative image of a 12-well culture plate containing cells coexpressing GFP, Rab11b WT, or Rab11b S25N and immunolabeled with anti-HA. In the middle row are wells in which primary antibody was omitted, which were used for background infrared signal subtraction. Cells expressing WT Cav1.2 did not demonstrate immunolabeling with anti-HA above background levels in either confocal imaging experiments or plate-based infrared imaging experiments (data not shown). Regions of interest, depicted with white dotted circles, were drawn within each well for quantitative analysis of infrared signal. Analysis of four independent experiments demonstrated that Rab11b S25N led to a significant 1.7-fold increase in plasma membrane-associated HA-Cav1.2 immunoinfrared signal compared with GFP control cells (Fig. 3C; *P < 0.05), and coexpression of Rab11b WT had no significant effect. To rule out the possibility that the increase in HA-Cav1.2 immunoinfrared signal observed in Rab11b S25N cells was secondary to increased cell number or size, cells were labeled with WGA conjugated to Alexa Fluor 680, which attaches to sialic acid residues in the plasma membrane and serves as an indirect measure of total cell plasmalemma. A representative infrared image of a 12-well culture plate containing cells coexpressing GFP, Rab11b WT, or Rab11b S25N shows similar WGA labeling among the three groups (Fig. 3B, bottom row), which when quantified was not significantly different (Fig. 3D, n = 4), demonstrating expression of Rab11b S25N does not alter cell membrane surface area.

Fig. 3.

Rab11b S25N increases plasmalemmal expression of Cav1.2. A: confocal image of a nonpermeabilized HEK293 cell transiently transfected with HA-Cav1.2 (containing extracellular HA epitope), β2cN4, and GFP and immunolabeled with anti-hemagglutinin (HA) to visualize surface membrane Cav1.2 channels. White scale bar represents 5 μm. B: representative image of the infrared emission signal obtained from cells grown in a 12-well tissue culture plate. Top row: cells that were immunolabeled with anti-HA; middle row: primary antibody was omitted to depict background infrared signal. Bottom row: cells that were labeled with wheat germ agglutinin (WGA) Alexa Fluor 680 to control for cell density. White dotted circles represent regions of interest used for quantitative analysis. C: mean HA infrared signal intensity (n = 4) in which data were normalized to the GFP control group; *P < 0.05. D: mean WGA infrared signal intensity (n = 4) after normalizing to GFP control group.

The observation that Rab11b S25N led to an increase in IBa,L and plasma membrane density of Cav1.2 channels was unexpected considering this dominant negative mutant significantly decreased plasmalemmal expression of other ion channels (12, 35). Therefore we tested the ability of Rab11b S25N to inhibit the endocytic recycling of Tfn receptors in HEK293 cells transfected with Cav1.2 and β2cN4. Cells coexpressing Rab11b WT (see supplemental Fig. S1A online at the AJP-Cell Physiol website) or Rab11b S25N (supplemental Fig. S1B) comparably internalized Tfn (t = 0). However, after a 40-min incubation to allow endosomal recycling, most Tfn was efficiently recycled from Rab11b WT expressing cells while Rab11b S25N cells still retained most internalized Tfn (t = 40 min). This result is consistent with previous findings (31, 33) and confirms Rab11b S25N blocks Rab11b-mediated endocytic recycling in our expression system. Taken together, these results suggest that disruption of Rab11b function with dominant negative Rab11b S25N modulates functional L-type current density by increasing the surface membrane expression of Cav1.2 channels heterologously expressed in HEK293 cells.

Role of Rab11b in retrograde trafficking of Cav1.2 from the plasma membrane.

The increase in plasma membrane density of Cav1.2 L-type Ca2+ channels observed with coexpression of Rab11b S25N could be due to an increase in anterograde trafficking of Cav1.2 to the surface membrane or, alternatively, decreased retrograde trafficking. If Rab11b S25N impacts Cav1.2 trafficking primarily through enhancing forward trafficking from Golgi compartments, we would anticipate increased surface membrane expression of LTCCs even at early time points after transient transfection as Cav1.2 is synthesized and transported through the secretory pathway to the plasma membrane. To address this possibility we performed cell surface biotinylation of HEK293 cells at 12, 24, and 36 h after transient transfection with HA-Cav1.2, β2cN4, and either GFP, Rab11b WT, or Rab11b S25N. Cell lysates were subjected to immunoprecipitation with anti-HA and analyzed by Western blotting. All three groups demonstrated a similar profile of total HA-Cav1.2 protein expression that peaked at 24 h posttransfection and began to decline by 36 h posttransfection (Fig. 4A, top). When quantified, total HA-Cav1.2 protein levels among the three groups did not significantly differ at any of the time points studied (Fig. 4B). To focus specifically on surface membrane Cav1.2 protein, membranes were stripped and reprobed with anti-biotin (Fig. 4A, bottom). No significant differences in biotinylated HA-Cav1.2 signal were observed among the three groups at 12 or 24 h after transient transfection (Fig. 4A, bottom and 4C). These findings argue against Rab11b S25N increasing LTCC plasma membrane density primarily through enhancing anterograde trafficking of newly synthesized Cav1.2 channels. This is consistent with previous observations that Rab11b resides primarily in the perinuclear recycling endosomal compartment with minimal colocalization with the Golgi-specific marker MPR300 or with ER markers (33). At 36 h posttransfection, however, a significant increase in biotinylated HA-Cav1.2 signal was observed from cells cotransfected with Rab11b S25N compared with GFP and Rab11b WT groups, which had both declined to similar levels from peak plasma membrane expression at 24 h posttransfection (Fig. 4A, bottom and Fig. 4C; *P < 0.05). This result suggests disruption of Rab11b function with dominant negative Rab11b S25N may lead to increased stability of Cav1.2 LTCCs at the plasma membrane.

Fig. 4.

Rab11b S25N increases plasmalemmal expression of Cav1.2 at late, but not early, time points after transient transfection. A: Western blot analysis of cell surface biotinylation studies performed 12, 24, or 36 h posttransfection. Samples were immunoprecipitated with anti-HA antibody and detected by immunoblotting with anti-HA (top) and anti-biotin (bottom). Average densitometric analysis (n = 4) of HA (B) and biotin (C) signal intensities over time for GFP (solid bars), Rab11b WT (shaded bars), and Rab11b S25N (open bars) expressing cells. HA and biotin signal intensities were normalized to their respective values from the GFP control group at 24 h posttransfection. *P < 0.05.

To further investigate the role of Rab11b in regulating retrograde trafficking of plasma membrane Cav1.2 channels, we biotinylated cell surface proteins and studied the degradation of biotinylated HA-Cav1.2 protein over time. At 20–24 h posttransfection, cells were cooled to 4°C, labeled with a pulse of biotin, and then incubated in culture media at 37°C + 5% CO2 for defined times to allow degradation of biotinylated transmembrane proteins. Cell lysates were used for IP with anti-HA or isolation of biotinylated proteins using NeutrAvidin beads and analyzed by SDS/PAGE and Western blot. A representative Western blot and average data from densitometric analysis (n = 3–6 for each time point) are plotted in Fig. 5. Standard IP samples revealed total HA-Cav1.2 protein levels did not significantly change during the 6-h experiment for any of the three groups (Fig. 5A, top, and 5B). However, analysis of the NeutrAvidin-captured samples revealed progressive loss of biotin pulse-labeled HA-Cav1.2 over time in all three groups that was significantly blunted in cells expressing Rab11b S25N at 6 h when compared with GFP control or Rab11b WT cells (Fig. 5A, middle, and 5C; *P < 0.05). Interestingly, analysis of biotin pulse-labeled Tfn receptor protein (Fig. 5A, bottom, and 5D) showed minimal change over the 6-h time course in any of the three groups, consistent with the idea that Rab11b selectively regulates Cav1.2 channels through a unique mechanism that is not shared with all endosomal cargo.

Fig. 5.

Slowed degradation of plasmalemmal Cav1.2 protein by Rab11b S25N. HEK293 cells were pulse-labeled with biotin and then incubated at 37°C + 5% CO2 for 0, 1, 3, or 6 h. Cell lysates were used for IP with anti-HA or incubated with NeutrAvidin and analyzed by SDS/PAGE and Western blotting. A: representative Western blot for total HA-Cav1.2 (top), biotin pulse-labeled HA-Cav1.2 (middle), and biotin-pulse labeled transferrin receptor (Tfn-R, bottom). Densitometric analysis (n = 3–6) was performed to quantitate total HA-Cav1.2 (B), biotin pulse-labeled HA-Cav1.2 (C), and biotin-pulse labeled Tfn-R (D) over time for GFP (■, solid line), Rab11b WT (●, dashed line), and Rab11b S25N (▲, dotted line) groups. Within each group, values were normalized to the signal intensity at t = 0. *P < 0.05.

Rab11b is expressed in ventricular myocardium and regulates endogenous Cav1.2 L-type Ca2+ channels in cardiac myocytes.

Studies performed in the HEK293 cell expression system yielded valuable insight into the mechanism of Rab11b regulation of Cav1.2 L-type Ca2+ channels. We next wanted to determine whether Rab11b regulates endogenous LTCCs in cardiomyocytes. Expression of Rab11b has been demonstrated in a variety of neuronal and epithelial cell types including HEK293 cells (22, 24, 33, 35), but its expression in ventricular myocardium has not been clearly established. To examine expression of Rab11b in the heart, we performed RT-PCR using RNA from freshly isolated adult mouse ventricular myocytes or brain. Primers specific for Rab11b amplified a single band of the predicted 280 base pairs in both brain and ventricular myocytes (Fig. 6A). β-Actin was used as a control and the neuronal-specific MAP2 was included to exclude the possibility of synaptic contamination in the myocyte preparations. Protein homogenates from an adult mouse left ventricle or brain were analyzed by SDS-PAGE and Western blotting using antibody specific for Rab11b (24). Consistent with other reports (22, 23), Rab11b protein was abundant in the brain, and we also readily detected Rab11b in left ventricular myocardium (Fig. 6B).

Fig. 6.

Rab11b is expressed in ventricular myocyardium and regulates IBa,L in cardiomyocytes. A: RT-PCR analysis using RNA extracted from adult mouse brain tissue or freshly isolated left ventricular (LV) myocytes. B: Western blot analysis of adult mouse brain or LV homogenate using antibody specific for Rab11b. C: representative IBa,L recordings from neonatal mouse ventricular myocytes. D: mean I-V relations from myocytes transfected with GFP (n = 5), Rab11b WT (n = 6), and Rab11b S25N (n = 4). *P < 0.05.

To test the role of Rab11b on the functional expression of endogenous cardiac L-type channels, we performed whole cell patch-clamp recordings of neonatal mouse ventricular myocytes transfected with GFP, Rab11b WT, or Rab11b S25N. Representative IBa,L traces recorded from myocytes are shown in Fig. 6C. Mean I-V relation data reveal expression of Rab11b S25N significantly increased peak IBa,L at 0 mV by 98% (P < 0.05) compared with GFP control cells, whereas Rab11b WT did not significantly alter current density compared with GFP control cells (Fig. 6D). These results demonstrate Rab11b is expressed in the myocardium and regulates functional expression of native Cav1.2 L-type Ca2+ channels in cardiac myocytes similarly to the regulation observed for heterologously expressed Cav1.2 channels in HEK293 cells.


In the present study, we provide evidence that Rab11b regulates the degradation of surface membrane Cav1.2 L-type Ca2+ channels in both HEK293 cells and native cardiac myocytes. Disruption of Rab11b function with a dominant negative Rab11b S25N mutant or via shRNA-mediated Rab11b depletion increased the functional expression of Cav1.2 LTCCs based on whole cell patch-clamp studies. Quantitative analysis of immunolabeled surface membrane HA-tagged Cav1.2 suggested this increase in L-type current is due to increased plasmalemmal density of Cav1.2 channels. Cell surface biotinylation studies demonstrated that Rab11b does not play a significant role in the delivery of newly synthesized LTCCs to the plasma membrane but rather regulates the degradation of surface membrane Cav1.2 channels. This mechanism differs from the established role of Rab11b in regulating endosomal recycling of membrane proteins such as Tfn receptor to the surface membrane and suggests Rab11b differentially regulates the targeting of distinct endosomal cargo.

The Rab family of small GTPases coordinates nearly all steps of intracellular protein trafficking. Rab11b has been localized to the perinuclear endosomal recycling compartment in numerous cell types (24, 31, 33, 45), and we and others have demonstrated that Rab11b regulates endocytic recycling of Tfn receptors (31, 33, and supplemental Fig. S1). Existing data have shown that Rab11b is also a positive regulator of ion channel surface membrane expression, because disruption of endogenous Rab11b function through expression of dominant negative mutants decreased plasmalemmal localization and functional expression of hERG K+ channels as well as CFTR channels (12, 35). Taking our findings into consideration, we propose a model in which Rab11b performs dual roles at the endocytic recycling compartment, preferentially recycling Tfn receptors (and other transmembrane proteins) to the plasma membrane while targeting Cav1.2 channels toward a degradative pathway.

Elegant work using total internal reflection microscopy to monitor plasma membrane trafficking of Cav1.2 at high spatiotemporal resolution showed highly mobile Cav1.2-containing vesicles that closely resembled Tfn receptor-containing endosomal vesicles in neurons (16). Because the primary known function of Rab11b is regulating endosomal recycling, we anticipated that blocking Rab11b function with Rab11b S25N would decrease Cav1.2 surface membrane expression and functional IBa,L. However, other studies demonstrated that exposing neurons to glutamate leads to internalization of Cav1.2 and subsequent targeting of Cav1.2 for lysosomal degradation (41). This finding highlights that degradation of internalized Cav1.2 represents an important pathway of LTCC trafficking in neurons, which is hypothesized to serve as protection against Ca2+-induced excitotoxicity. The role that Rab11b plays in trafficking internalized Cav1.2 toward a degradative pathway rather than through endocytic recycling may represent a related protective mechanism to limit plasma membrane density of LTCCs.

Our results show Rab11b S25N blunts the degradation of biotin pulse-labeled plasmalemmal Cav1.2 over time. The resulting increase in IBa,L suggests that either internalization of functional Cav1.2 channels from the plasma membrane is slowed, or alternatively, internalized Cav1.2 channels are readily recycled to the plasma membrane through a Rab11b-independent pathway under conditions when endogenous Rab11b activity is blocked. Direct recycling from early endosomes has been well described (25), and there is evidence that Rab11 blockade actually enhances fast direct recycling of αVβ3 integrin to the plasma membrane in response to platelet-derived growth factor (PDGF) (32), perhaps by increasing the pool available for fast recycling. Ongoing studies are aimed at investigating other potentially relevant LTCC trafficking pathways.

Interestingly, degradation of Cav1.2 channels was not accelerated with overexpression of Rab11b WT or with coexpression of the GTPase-deficient Rab11b Q70L mutant in our studies. Rab11b Q70L has been shown to act as a dominant active mutant in accelerating CFTR recycling in T84 colon epithelial cells (35). However, Rab11b Q70L yielded similar results to dominant negative Rab11b S25N in reducing cell migration of PtK1 cells in a wound healing assay (31). Thus the ability of Rab11b Q70L to accelerate Rab11b-mediated trafficking may be cell-type dependent or cargo specific. Our results suggest that the impact of Rab11b on LTTC trafficking is saturated at the basal levels of expression under our conditions, and further expression of WT Rab11b or Rab11b Q70L does not impact Cav1.2 surface expression and IBa,L any further. This may be due to the lack of other rate-limiting effector proteins such as the Rab11 family interacting proteins (Rab11-FIPs). Consistent with the current findings, we previously demonstrated that overexpression of several WT small GTPases had little effect on the current density or plasmalemmal immunostaining of hERG when compared with control (12).

How Rab11b can recognize and perform different functions on distinct membrane proteins remains an intriguing question. Interestingly, glutamate-induced Cav1.2 lysosomal degradation was dependent on the lipid kinase PIKfyve, whose activity generates phosphatidylinositol (3,5)-bisphosphate, a phospholipid found in endosomal membranes and implicated in membrane trafficking and lyosomal function (41). However, activation of PIKfyve by the serum- and glucocortocoid-induced kinase was shown to promote endocytic recycling of KCNQ1/KCNE1 channels to the plasma membrane in oocytes (34), while protein kinase B phosphorylation of PIKfyve stimulates the recruitment of intracellular GLUT4 vesicles to the plasma membrane in adipocytes (6). These findings suggest that distinct cargo may be differentially targeted within the endosomal transport pathway even under seemingly similar regulation. This may potentially be achieved through unique adapter proteins that associate with specific cargo. Rab11-FIPs consist of five members that interact with Rab11, and there is growing evidence that individual Rab11-FIPs are involved in the trafficking of specific cargo (19). For example, the Rab11-FIP member Rip11 is responsible for the translocation of GLUT4 transporters to the plasma membrane in adipocytes (43). Understanding the precise mechanisms responsible for the Rab11b-mediated degradation of surface membrane Cav1.2 channels will require further studies.

Differential trafficking of multiple ion channels by a single GTPase could provide a mechanism to integrate cellular phenotypic and pathological responses. For example, the ventricular myocyte action potential requires precise orchestration between several depolarizing and repolarizing currents. L-type Ca2+ current (ICa,L) through Cav1.2 channels contributes the depolarizing current to support the plateau phase of the ventricular action potential, but the rapid component of the delayed rectifier K+ current (IKr) carried by hERG channels provides a major repolarizing current. Dysregulation of both ICa,L and IKr have been implicated in the arrhythmogenesis of long- and short-QT syndromes (2, 3, 8, 11, 37). Rab11b blockade leads to increased ICa,L and decreased HERG current (IhERG)(12), and these effects will additively prolong the action potential duration. Therefore, relatively small changes in the function or expression of Rab11b could have a significant impact on the cardiac action potential and theoretically alter the risk of arrhythmias.

There are some limitations to the current study. Many of the experiments were performed in a heterologous HEK293 cell expression system in which proteins are overexpressed, which can make the interpretation of trafficking studies difficult. However, we were able to recapitulate functional results obtained in HEK293 cells by recording endogenous L-type currents in cardiac myocytes. Furthermore, disruption of endogenous Rab11b function in our studies was primarily achieved by use of a dominant negative Rab11b S25N mutant, which may have off-target effects or compete for Rab11 effector proteins. However, depletion of Rab11b protein levels using a specific shRNA construct led to a similar increase in IBa,L to that which was observed with expression of Rab11b S25N, while knockdown of Rab11a had no significant impact on IBa,L. This suggests the effects observed with Rab11b S25N are largely due to disruption of basal Rab11b function. Finally, for simplicity in our heterologous expression system, many of our studies utilized LTCC subunits Cav1.2 and β2cN4. Whereas we acknowledge the complexity of endogenous LTCC composition involving multiple Cavβ, Cavα2δ, and Cavγ subunits and splice variants, our data from neonatal mouse ventricular myocytes suggest that our findings in HEK293 cells apply to endogenous cardiac LTCCs.

In summary we have identified the small GTPase Rab11b as an important regulator of Cav1.2 L-type Ca2+ channel surface membrane density and functional L-type current by modulating Cav1.2 trafficking from the plasma membrane toward a degradative pathway. These results contribute to our knowledge regarding the role of endosomal trafficking in regulating functional expression of LTCCs and provide mechanistic framework to better understand alterations in L-type Ca2+ channel density seen in human disease.


This work was supported by National Heart, Lung, and Blood Institute Grants R01 HL-078878 (to T. J. Kamp) and R01 HL-087039 (to B. P. Delisle), American Heart Association Scientist Development Grant 0730010N (to R. C. Balijepalli), and American Heart Association Predoctoral Fellowships 10PRE2580002 (to J. M. Best) and 0910120G (to J. D. Foell).


No conflicts of interest, financial or otherwise, are declared by the author(s).


We are grateful for technical assistance from Jing Wang and Marlese Koehnlein. We thank Dr. Nikolai Soldatov (National Institute on Aging, National Institutes of Health, Bethesda, MD) for supplying the α1C,77 expression plasmid, Dr. Beate Schlierf (Institut für Biochemie, Universität Erlangen-Nürnberg, Erlangen, Germany) for the gift of the GFP-Rab11b expression constructs, and Drs. Lynn Lapierre and James Goldenring (Vanderbilt University Medical Center, Nashville, TN) for supplying Rab11b rabbit polyclonal antibody.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
View Abstract