Posterior body wall muscle contraction (pBoc) in the nematode Caenorhabditis elegans occurs rhythmically every 45–50 s and mediates defecation. pBoc is controlled by inositol-1,4,5-trisphosphate (IP3)–dependent Ca2+ oscillations in the intestine. The intestinal epithelium can be studied by patch clamp electrophysiology, Ca2+ imaging, genome-wide reverse genetic analysis, forward genetics, and molecular biology and thus provides a powerful model to develop an integrated systems level understanding of a nonexcitable cell oscillatory Ca2+ signaling pathway. Intestinal cells express an outwardly rectifying Ca2+ (ORCa) current with biophysical properties resembling those of TRPM channels. Two TRPM homologues, GON-2 and GTL-1, are expressed in the intestine. Using deletion and severe loss-of-function alleles of the gtl-1 and gon-2 genes, we demonstrate here that GON-2 and GTL-1 are both required for maintaining rhythmic pBoc and intestinal Ca2+ oscillations. Loss of GTL-l and GON-2 function inhibits IORCa ∼70% and ∼90%, respectively. IORCa is undetectable in gon-2;gtl-1 double mutant cells. These results demonstrate that (a) both gon-2 and gtl-1 are required for ORCa channel function, and (b) GON-2 and GTL-1 can function independently as ion channels, but that their functions in mediating IORCa are interdependent. IORCa, IGON-2, and IGTL-1 have nearly identical biophysical properties. Importantly, all three channels are at least 60-fold more permeable to Ca2+ than Na+. Epistasis analysis suggests that GON-2 and GTL-1 function in the IP3 signaling pathway to regulate intestinal Ca2+ oscillations. We postulate that GON-2 and GTL-1 form heteromeric ORCa channels that mediate selective Ca2+ influx and function to regulate IP3 receptor activity and possibly to refill ER Ca2+ stores.
Introduction
The genetic model organism Caenorhabditis elegans provides numerous experimental advantages for developing an integrative genetic and molecular understanding of fundamental physiological processes (Barr, 2003;Strange, 2003). These advantages include a short life cycle, forward genetic tractability, a fully sequenced and well-annotated genome, and relative ease and economy of characterizing gene function using transgenic and RNA interference methods.
C. elegans intestinal epithelial cells generate rhythmic inositol 1,4,5-trisphosphate (IP3)–dependent Ca2+ oscillations that control posterior body wall muscle contraction (pBoc) (Dal Santo et al., 1999; Espelt et al., 2005; Teramoto and Iwasaki, 2006; Peters et al., 2007). pBoc is part of a motor program that mediates defecation and can be observed readily through a dissecting microscope making it amenable to forward and reverse genetic screening (Thomas, 1990; Liu and Thomas, 1994; Iwasaki et al., 1995). Intestinal Ca2+ signaling can be quantified by imaging methods in isolated intestines (Espelt et al., 2005; Teramoto and Iwasaki, 2006; Peters et al., 2007) or in vivo using genetically encoded Ca2+ indicators (Teramoto and Iwasaki, 2006; Yan et al., 2006; Peters et al., 2007). Recent development of primary cell culture methods (Christensen et al., 2002; Strange et al., 2007) has made it possible to characterize intestinal ion channels using patch clamp methods (Estevez et al., 2003; Estevez and Strange, 2005; Yan et al., 2006; Lorin-Nebel et al., 2007). The ability to combine direct physiological measurements of IP3–dependent oscillatory Ca2+ signals and associated ion channel activity with forward and reverse genetic screening is unique to C. elegans. The worm intestinal epithelium thus provides a powerful model system in which to define the genetic and molecular details and integrative physiology of oscillatory Ca2+ signaling in nonexcitable cells.
Intestinal Ca2+ oscillations are strictly dependent on Ca2+ release from the ER via ITR-1, the single IP3 receptor encoded by the C. elegans genome (Dal Santo et al., 1999; Espelt et al., 2005; Teramoto and Iwasaki, 2006). Extensive studies in vertebrate (for reviews see Venkatachalam et al., 2002; Parekh and Putney, 2005; Hogan and Rao, 2007) and Drosophila cells (Yeromin et al., 2004) have demonstrated that depletion of ER Ca2+ stores activates store-operated Ca2+ channels (SOCCs). SOCCs are widely believed to be an essential and ubiquitous component of Ca2+ signaling pathways, functioning to refill ER Ca2+ stores and modulate intracellular Ca2+ signals (e.g., Venkatachalam et al., 2002; Parekh and Putney, 2005; Hogan and Rao, 2007). The most extensively studied and characterized SOCC is the Ca2+ release–activated Ca2+ (CRAC) channel (Parekh and Putney, 2005). The CRAC channel pore is comprised of Orai1/CRACM and channel activation is mediated by STIM1, which functions as an ER Ca2+ sensor (for reviews see Hogan and Rao, 2007; Lewis, 2007; Putney, 2007).
C. elegans intestinal cells express robust CRAC channel activity (Estevez et al., 2003). RNAi silencing of orai-1 or stim-1, which encode worm Orai1/CRACM and STIM1 homologues, dramatically reduces CRAC channel expression and function, but surprisingly has no effect on intestinal Ca2+ signaling (Lorin-Nebel et al., 2007;Yan et al., 2006). These findings suggest that CRAC channels are not essential components of IP3-dependent Ca2+ signaling in the intestine and indicate that other Ca2+ entry mechanisms must function to maintain intestinal Ca2+ oscillations.
In addition to CRAC channels, intestinal cells express a store-independent outwardly rectifying Ca2+ (ORCa) channel that has biophysical properties resembling those of mammalian TRPM channels (Estevez et al., 2003). Three TRPM homologues are encoded by the C. elegans genome, GON-2, GTL-1, and GTL-2 (Kahn-Kirby and Bargmann, 2006; Baylis and Goyal, 2007). GFP reporter studies have demonstrated that intestinal cells express gon-2 and gtl-1 (Teramoto et al., 2005; cited as unpublished observations in Baylis and Goyal, 2007; WormBase, http://www.wormbase.org/). The goal of the present study was to define the roles these genes play in intestinal Ca2+ signaling. Our results demonstrate that GON-2 and GTL-1 are both required for ORCa channel activity and for maintaining rhythmic Ca2+ oscillations. We propose that gon-2 and gtl-1 encode the ORCa channel. We also suggest that ORCa channels comprise a major Ca2+ entry pathway in intestinal epithelial cells and that they function to regulate IP3 receptor activity and refill ER Ca2+ stores.
MATERIAL AND METHODS
C. elegans Strains
Nematodes were cultured using standard methods on nematode growth medium (NGM) (Brenner, 1974). Wild-type worms were the Bristol N2 strain or elt-2∷gfp worms that express a transcriptional GFP reporter in intestinal cell nuclei. Worms homozygous for the gon-2 loss-of-function allele gon-2(q388) or the gtl-1 deletion allele gtl-1(ok375) were used for studies of GON-2 and GTL-1 function. gon-2;gtl-1 double mutant worms were generated by crossing the gtl-1(ok375) and gon-2(q388) strains (Teramoto et al., 2005). The gon-2;gtl-1 double mutant worm strain exhibits greatly slowed larval development on NGM. To improve development and fertility sufficiently for experiments to be performed, double mutants were grown on NGM supplemented with 20 mM Mg2+ (see Teramoto et al., 2005). All worm strains were maintained at 16–20°C. Growth temperatures used in specific experiments are described below.
Characterization of pBoc Cycle
gon-2(q388) is a temperature sensitive allele and the mutant phenotype is observed at growth temperatures of 25°C (Sun and Lambie, 1997). For posterior body wall muscle contraction (pBoc) measurements, eggs from wild-type and mutant worm strains were cultured in a 25°C incubator until adulthood. The times required for wild-type, gon-2 mutant, gtl-1 mutant, and double mutant worms to reach adulthood at 25°C were 2–3 d, 3–4 d, 3–4 d, and 5 d, respectively.
pBoc was monitored by imaging worms on growth agar plates using a Carl Zeiss MicroImaging Inc. Stemi SV11 M2BIO stereo dissecting microscope (Kramer Scientific Corp.) equipped with a DAGE-MTI DC2000 CCD camera. A minimum of 10 pBoc cycles were measured in each animal. Measurements were performed at a room air temperature of 22–23°C. Agar temperature was monitored during the course of pBoc measurements using a thermistor (Model 4600, Yellow Springs Instruments) and was 24–25°C.
Dissection and Fluorescence Imaging of Intestines
Worms were cultured as described above for pBoc measurements. Calcium oscillations were measured in isolated intestines as described previously (Espelt et al., 2005). In brief, worms were placed in control saline (137 mM NaCl, 5 mM KCl, 1 mM MgCl2, 1 mM MgSO4, 0.5 mM CaCl2, 10 mM HEPES, 5 mM glucose, 2 mM l-asparagine, 0.5 mM l-cysteine, 2 mM l-glutamine, 0.5 mM l-methionine, 1.6 mM l-tyrosine, 28 mM sucrose, pH 7.3, 340 mOsm) and cut behind the pharynx using a 26-gauge needle. The hydrostatic pressure in the worm spontaneously extruded the intestine, which remained attached to the rectum and the posterior end of the animal. Isolated intestines were incubated for 15 min in bath saline containing 5 μM fluo-4 AM and 1% BSA. Imaging was performed using a Nikon TE2000 inverted microscope, a Superfluor 40X/1.3 N.A. oil objective lens, a Photometrics Cascade 512B cooled CCD camera (Roper Industries), and MetaFluor software (Molecular Devices Corporation). Room temperature was maintained at 25–26°C. Fluo-4 was excited using a 490-500BP filter and a 523-547BP filter was used to detect fluorescence emission. Fluorescence images were acquired at 0.2 or 1 Hz. Changes in fluo-4 intensity were quantified in posterior-to-anterior moving Ca2+ waves using region-of-interest selection and MetaFluor software (Molecular Devices Corporation).
C. elegans Embryonic Cell Culture and Patch Clamp Electrophysiology
Newly hatched wild type and mutant worm L1 larvae were cultured at 25°C until adulthood. Embryonic cells were cultured for 2–3 d at 25°C on 12-mm diameter acid-washed glass coverslips using established methods (Christensen et al., 2002; Strange et al., 2007). To maximize suppression of GON-2 activity, cells isolated from gon-2 and gon-2;gtl-1 double mutant worms were cultured in the presence of gon-2 double stranded RNA (dsRNA) using methods described previously (Yan et al., 2006; Lorin-Nebel et al., 2007). gon-2 dsRNA was synthesized from a 640-bp (4041–4681-bp) gon-2 cDNA that was amplified from a C. elegans cDNA library.
Coverslips with cultured embryo cells were placed in the bottom of a bath chamber (model R-26G; Warner Instrument Corp.) that was mounted onto the stage of a Nikon TE2000 inverted microscope. Bath temperature was maintained at 25°C using a Warner Instruments model SC-20 dual in-line heater/cooler, a model CL-100 bipolar temperature controller, and a PHC series heater/cooler jacket for the bath chamber. Cells were visualized by fluorescence and video-enhanced DIC microscopy. Intestinal cells were identified in culture by expression of the intestine-specific reporter elt-2∷GFP or by morphological characteristics (Fukushige et al., 1998; Estevez et al., 2003).
Patch electrodes were pulled from soft glass capillary tubes (PG10165-4, World Precision Instruments) that had been silanized with dimethyl-dichloro silane. Pipette resistance was 4–7 MΩ. Bath and pipette solutions contained 145 mM NaCl, 1 mM CaCl2, 5 mM MgCl2, 10 mM HEPES, 20 mM glucose, pH 7.2 (adjusted with NaOH), and 147 mM sodium gluconate (NaGluconate), 0.6 mM CaCl2, 1 mM MgCl2, 10 mM EGTA, 10 mM HEPES, 2 mM Na2ATP, 0.5 mM Na2GTP, pH 7.2 (adjusted with CsOH), respectively. The osmolality of bath and pipette solutions were adjusted to 345–350 mOsm and 325–330 mOsm using sucrose.
Whole cell currents were recorded using an Axopatch 200B (Axon Instruments) patch clamp amplifier. Command voltage generation, data digitization, and data analysis were performed on a 2.79 GHz Pentium computer (Dimension 9150; Dell Computer Corp.) using a Digidata 1322A AD/DA interface with pClamp 10 software (Axon Instruments). Electrical connections to the amplifier were made using Ag/AgCl wires and 3 M KCl/agar bridges.
Currents were elicited using a ramp or step voltage clamp protocol. For the ramp protocol, membrane potential was held at 0 mV and ramped from −80 to +80 mV at 215 mV/s every 5 s. Step changes in whole cell current were elicited by stepping membrane voltage from −80 to +80 mV in 20-mV steps from a holding potential of 0 mV. Voltage steps were maintained for 400 ms. Cell capacitances for all cells studied ranged from 1 to 4 pF.
As we described previously, IORCa is outwardly rectifying with a strongly positive reversal potential (Estevez et al., 2003). In the present study, we also observed that currents in gon-2 and gtl-1 mutant cells reversed at strongly positive membrane potentials and exhibited outward rectification. Outwardly rectifying currents with reversal potentials <10 mV were deemed to be excessively contaminated with nonspecific leak current and were rejected from final datasets.
Ion substitution studies were performed by replacement of bath Na+ with various test cations. Cells were patch clamped initially in control bath solution until whole cell current had stabilized and then switched to a Ca2+- and Mg2+-free medium containing 1 mM EGTA. Changes in reversal potential (Erev) were measured after replacement of 150 mM bath NaCl with 150 mM NMDG-Cl, 130 mM NMDG-Cl, and 10 mM CaCl2 or 130 mM NMDG-Cl and 10 mM MgCl2. Liquid junction potential changes were calculated using pClamp 10. Reversal potentials during ion substitution experiments were corrected for liquid junction potentials. Relative permeabilities were calculated from Erev changes as described previously (Estevez et al., 2003).
Induction of RNA Interference by Double Strand RNA Feeding
RNA interference was induced by feeding gon-2;gtl-1 double mutant worms bacteria producing dsRNA (e.g., Kamath et al., 2000; Rual et al., 2004) homologous to PLCγ or PLCβ. RNAi bacterial strains were engineered as described previously (Yin et al., 2004). Bacterial strains were streaked to single colonies on agar plates containing 50 μg/ml ampicillin and 12.5 μg/ml tetracycline. Single colonies were used to inoculate LB media containing 50 μg/ml ampicillin and cultures were grown at 37°C for 16–18 h with shaking. 300 μl of each bacterial culture were seeded onto 60-mm NGM agar plates containing 20 mM Mg2+, 50 μg/ml ampicillin, and 1 mM IPTG to induce dsRNA synthesis. After seeding, plates were left at room temperature overnight. Eggs were transferred to the RNAi feeding plates and grown at 25°C.
Statistical Analysis
Data are presented as means ± SEM. Statistical significance was determined using Student's two-tailed t test for unpaired means. When comparing three or more groups, statistical significance was determined by one-way analysis of variance with a Bonferroni post-hoc test. P values of ≤0.05 were taken to indicate statistical significance. The rhythmicity of the pBoc cycle and intestinal Ca2+ oscillations is quantified as coefficient of variance, which is the standard deviation expressed as a percentage of the sample mean.
Results
Removal of Extracellular Ca2+ Causes Rapid Cessation of Intestinal Ca2+ Oscillations
Calcium is taken up into the ER via the sarco/endoplasmic reticulum Ca2+ ATPase (SERCA) while plasma membrane pumps and exchangers continuously extrude Ca2+ from the cell (Berridge et al., 2003; Hogan and Rao, 2007). Because of the presence of plasma membrane Ca2+ extrusion mechanisms, some Ca2+ will be lost from the cell during ER Ca2+ release. Repeated and/or prolonged ER Ca2+ release will eventually deplete ER Ca2+ stores and prevent further IP3-dependent Ca2+ signals unless plasma membrane Ca2+ entry mechanisms are also active. To determine whether such Ca2+ entry mechanisms are required for IP3-dependent Ca2+ signaling in the intestine, we monitored Ca2+ oscillations during removal of bath Ca2+. As shown in Fig. 1, total intracellular fluo-4 fluorescence dropped and Ca2+ oscillations ceased rapidly when extracellular Ca2+ was removed. Calcium oscillations recovered when Ca2+ was added back to the bath. These results demonstrate that Ca2+ entry mechanisms are active in the intestine and that Ca2+ oscillations are strictly dependent on extracellular Ca2+ influx. Calcium entry almost certainly functions to refill ER stores. In addition, Ca2+ influx may modulate IP3 receptor activity and/or contribute to the total increase in cytoplasmic Ca2+ concentration during Ca2+ oscillations.
The TRPM Channels GTL-1 and GON-2 Are Required for Normal Intestinal Ca2+ Signaling
As discussed in the Introduction, loss of function of CRAC channels and the ER Ca2+ sensor STIM-1 has no effect on oscillatory Ca2+ signaling in the C. elegans intestine (Lorin-Nebel et al., 2007; Yan et al., 2006). Other channels must therefore mediate Ca2+ entry. Given that gon-2 and gtl-1 are expressed in the intestine (Teramoto et al., 2005; cited as unpublished observations in Baylis and Goyal, 2007; WormBase, http://www.wormbase.org/), we quantified pBoc and intestinal Ca2+ oscillations in animals harboring loss-of-function mutations in these genes. gtl-1(ok375) is a 2,714-bp deletion allele that deletes all of the predicted transmembrane domains of GTL-1 and is almost certainly null. gon-2(q388) is a point mutation in which glutamate 955 is mutated to lysine (West et al., 2001). Glutamate 955 is highly conserved in human, mouse, Drosophila, and C. elegans TRP channels and mutation to lysine most likely causes temperature-sensitive disruption of a step in GON-2 synthesis (West et al., 2001). The E955K mutation induces a severe loss-of-function phenotype when worms are grown at 25°C (Sun and Lambie, 1997; Church and Lambie, 2003). As noted earlier, the gon-2;gtl-1 double mutant was derived from a cross of gtl-1(ok375) and gon-2(q388) worms (Teramoto et al., 2005).
Fig. 2 A shows pBoc cycles in individual wild-type and channel mutant worms. Coefficients of variance were calculated as a measure of cycle rhythmicity. Wild-type worms exhibited a highly rhythmic pBoc cycle with coefficients of variance for individual animals ranging from 2 to 5%. In striking contrast, loss of activity of either channel disrupted pBoc rhythmicity. Coefficients of variance ranged from 3 to 33% and 7 to 28% for GTL-1 and GON-2 mutant worms, respectively. Disruption of pBoc was more severe in the double mutant worms where coefficients of variance ranged from 10 to 67%.
pBoc cycle data are summarized in Fig. 2 B. Mean cycle periods and coefficients of variance were increased significantly (P < 0.05) in gtl-1 mutant, gon-2 mutant, and double mutant worms. In addition, the mean coefficient of variance was significantly (P < 0.01) greater in the double mutant worms compared with either GTL-1 or GON-2 mutant animals.
As discussed in the Materials and methods section, double mutant worms develop and reproduce poorly unless the Mg2+ concentration in the growth agar is increased to 20 mM. To determine whether high Mg2+ has any effect on the pBoc cycle, we grew wild-type worms for one generation on high Mg2+ plates. Mean ± SEM pBoc period and coefficient of variance were 43 ± 1 s and 3.5 ± 0.7% (n = 6), respectively, and were not significantly (P > 0.3) different from those of worms grown on standard NGM agar (see Fig. 2 B).
Intestinal IP3-dependent Ca2+ oscillations drive pBoc through a yet to be defined mechanism (Dal Santo et al., 1999; Espelt et al., 2005; Teramoto and Iwasaki, 2006; Peters et al., 2007). To determine whether GTL-1 and GON-2 function in Ca2+ signaling, we quantified Ca2+ oscillations in intestines dissected from wild-type and mutant animals. Calcium oscillations were arrhythmic in intestines isolated from GTL-1, GON-2, and double mutant worms (Fig. 3 A). Mean coefficients of variance were increased significantly (P < 0.05) by 2.3–3.2-fold in the single and double mutants (Fig. 3 B). Due to intracycle and animal-to-animal variability, the mean oscillation periods were not significantly (P > 0.05) different for the three groups of mutant worms and wild-type animals (unpublished data). Oscillation kinetics as measured by oscillation rise and fall times were unaffected (P > 0.05) by channel mutations (unpublished data). We conclude from data shown in Figs. 2 and 3 that GTL-1 and GON-2 are both required for maintaining the rhythmicity of Ca2+ oscillations in the C. elegans intestinal epithelium.
GTL-1 and GON-2 Mediate Whole Cell Outwardly Rectifying Ca2+ Currents
We suggested previously that IORCa may play an important role in generating intestinal Ca2+ oscillations (Estevez et al., 2003; Estevez and Strange, 2005). To determine whether the ORCa channel is encoded by gon-2 and/or gtl-1, we characterized whole cell cation currents in intestinal cells cultured from wild-type, gon-2 mutant, gtl-1 mutant, and gon-2;gtl-1 double mutant worms. IORCa in wild-type intestinal cells is constitutively active and undergoes additional slow activation for 1–2 min after whole cell recording is initiated (Estevez et al., 2003). Mean ORCa current density at +80 mV measured 4–5 min after membrane rupture in wild-type cells was 266 pA/pF (Fig. 4 A). The mean ± SEM reversal potential (Erev) of IORCa was 18 ± 1 mV (n = 22). The positive reversal potential is expected for a Ca2+-selective channel (Estevez et al., 2003).
Whole cell current density was strikingly and significantly (P < 0.01) suppressed in intestinal cells cultured from both gon-2 and gtl-1 mutant worms. In both groups of cells, the majority of currents we observed were outwardly rectifying with a strongly positive Erev similar to that of IORCa. In 2 out 11 gon-2 mutant cells, whole cell current exhibited an Erev close to zero and a near-linear current-to-voltage relationship. We interpreted these observations as indicating that loss of function of gon-2 in these cells completely suppressed IORCa and that whole cell conductance was due largely to a nonselective leak current. Mean current density was 26.5 pA/pF in gon-2 cells and 83.5 pA/pF in gtl-1 cells (Fig. 4 A). Currents recorded from all gtl-1 cells showed outward rectification and had a mean ± SEM Erev of 19 ± 1 mV (n = 21). The mean ± SEM Erev value for the outwardly rectifying currents observed in gon-2 mutant cells was 18 ± 2 mV (n = 9). Mean reversal potentials of outwardly rectifying currents in gon-2 and gtl-1 mutant cells were not significantly (P < 0.05) different from that observed in wild-type cells.
In five out of five gon-2;gtl-1 double mutant cells, a small current with a near-linear current-to-voltage relationship was detected. The mean ± SEM Erev for this current was 1.1 ± 2.7 mV (n = 5), which is not significantly (P > 0.7) different from 0 (Fig. 4 B). To determine whole cell current properties in the absence of IORCa, we patch clamped wild-type intestinal cells and bathed them with 100 μM La3+, which completely inhibits ORCa channel activity (see Fig. 6 A). A small near-linear current with an Erev (mean ± SEM = −1.6 ± 1.5 mV; n = 5) not significantly (P > 0.3) different from 0 was recorded in these cells (Fig. 4 B). We define this current as nonselective leak current. Mean ± SEM whole cell currents measured at −80 mV and +80 mV in gon-2;gtl-1 double mutant cells and wild-type cells treated with 100 μM La3+ were −3.5 ± 1.8 pA/pF and 4.0 ± 1.8 pA/pF (n = 5) and −1.9 ± 3.2 pA/pF and 3.1 ± 0.4 pA/pF (n = 5), respectively, and were not significantly (P > 0.6) different (Fig. 4 B). Treatment of gon-2;gtl-1 mutant cells with 100 μM La3+ had no significant (P > 0.2) on whole cell current amplitude (mean ± SEM whole cell currents measured at −80 and +80 mV were −4.9 ± 1.0 and 8.3 ± 3.8 pA/pF, respectively; n = 3). These results demonstrate that combined loss of GON-2 and GTL-1 activity completely suppresses IORCa. We therefore conclude that IORCa is mediated by the function of both channels.
Functional Properties of GON-2 and GTL-1 Mediated Whole Cell Currents
The inhibitory effects of loss of GON-2 or GTL-1 alone on IORCa are not additive; whole cell current density was reduced ∼90% and ∼70% in gon-2 and gtl-1 mutant cells, respectively (Fig. 4 A). These results indicate that (a) GON-2 and GTL-1 can function independently as ion channels, but (b) their functions in mediating IORCa are somehow interdependent (see Discussion). We define the currents observed in gon-2 and gtl-1 mutant cells as IGTL-1 and IGON-2, respectively.
To further define the roles of GON-2 and GTL-1 in mediating IORCa, we characterized the biophysical properties of IGTL-1 and IGON-2. Fig. 5 shows representative ORCa (i.e., wild type), GON-2, and GTL-1 currents and relative current-to-voltage relationships. All three currents show similar outward rectification. However, relative inward GTL-1 currents at −20 to −80 mV were slightly but significantly (P < 0.001) greater than that of IORCa (Fig. 5 B).
IORCa was inhibited by extracellular La3+ with a mean ± SEM IC50 of 3.7 ± 0.6 μM (n = 6). The La3+ dose–response relationships for IGON-2 and IGTL-1 were superimposable with that of IORCa (Fig. 6 A). Mean ± SEM La3+ IC50 values were 5.7 ± 1.8 μM (n = 6) and 5.3 ± 1.5 μM (n = 4) for IGON-2 and IGTL-1, respectively, and were not significantly (P > 0.05) different from that of IORCa.
Fig. 6 B shows cation permeabilities measured under bi-ionic conditions of the ORCa, GON-2, and GTL-1 channels relative to Na+ (i.e, Pcation/PNa). The PNMDG/PNa, PCa/PNa, and PMg/PNa for the channels were not significantly (P > 0.05) different and ranged between 0.07 and 0.1, 57 and 66, and 3 and 6, respectively.
Increasing intracellular Mg2+ concentration inhibits IORCa (Fig. 6 C) (Estevez et al., 2003). The Mg2+ dose–response relationships for IORCa, IGON-2, and IGTL-1 were similar (Fig. 6 C). IC50 values derived from fits to mean values in the datasets were 420 μM for IORCa, 440 μM for IGON-2, and 260 μM for IGTL-1. Comparison of the fits indicated that the three datasets were not significantly (P > 0.05) different from one another.
It has been suggested that GON-2 and GTL-1 play a central role in intestinal Mg2+ uptake (Teramoto et al., 2005). The ORCa, GON-2, and GTL-1 channels clearly have measurable Mg2+ permeabilities under bi-ionic conditions. However, given that the relative Ca2+ permeabilities of the channels are at least an order of magnitude greater than that of Mg2+ (Fig. 6 B and Teramoto et al., 2005), a more physiologically relevant question is whether significant Mg2+ permeation occurs when Ca2+ is present in the extracellular medium. To address this question, we patch clamped wild-type intestinal cells in a modified standard bath solution containing 130 mM NaCl and 30 mM NMDG-Cl and the normal Ca2+ and Mg2+ concentrations of 1 mM and 5 mM, respectively. When current amplitude had stabilized, the NMDG-Cl was replaced with 15 mM MgCl2. In the presence of 1 mM bath Ca2+, the mean ± SEM shifts in Erev and current density at −80 mV observed when bath Mg2+ levels were raised fourfold were 0.7 ± 0.5 mV (n = 4) and −1.6 ± 1.7 pA/pF (n = 4), respectively (Fig. 7). These values were not significantly (P > 0.3) different from zero suggesting that Mg2+ permeation through the ORCa channel is very low in the presence of Ca2+. Studies designed to directly quantify net Mg2+ influx through the ORCa channel under physiologically relevant conditions are needed to fully define its role in intestinal Mg2+ uptake and whole animal Mg2+ homeostasis.
Physiological Roles of GON-2 and GTL-1
As shown in Figs. 2 and 3, loss of gon-2 and/or gtl-1 activity dramatically disrupts pBoc rhythmicity and intestinal Ca2+ signaling. Teramoto et al. (2005) observed that the pBoc cycle was prolonged and apparently arrhythmic in gon-2;gtl-1 double mutant worms and that the defect was fully rescued by increasing the Mg2+ concentration of the growth agar to 40 mM. They suggested that the altered defecation cycle was due to an alteration in the physiological state of the intestine resulting from Mg2+ deficiency. In our hands, the pBoc defect in double mutant worms was unaffected by external Mg2+ levels of either 20 (Fig. 2) or 40 mM (unpublished data).
Given the lack of effect we observed of high external Mg2+ concentration on pBoc and the high relative Ca2+ permeabilities of the ORCa, GON-2, and GTL-1 channels (Fig. 6 B), it is reasonable to postulate that they play a direct role in regulating and/or maintaining IP3-dependent intestinal Ca2+ oscillations. To address this possibility, we performed genetic epistasis experiments. PLCγ and PLCβ homologues function together to regulate pBoc and generate Ca2+ oscillations in the C. elegans intestine. Loss of function of either enzyme causes striking arrhythmia of both pBoc and oscillatory Ca2+ signaling. Combined loss of function of both enzymes is additive giving rise to severe Ca2+ signaling and pBoc defects (Espelt et al., 2005). These results suggest that PLCγ and PLCβ function in different signaling pathways. Epistasis analysis using mutant alleles predicted to elevate intracellular IP3 levels indicates that PLCγ functions primarily to generate IP3 and regulate IP3 receptor activity whereas PLCβ functions in a distinct and yet to be defined signaling pathway required for normal Ca2+ signaling (Espelt et al., 2005). The localization of PLCβ to sites of cell–cell contact (Miller et al., 1999) suggests that the enzyme may play a role in regulating intestinal Ca2+ waves that coordinate muscle contractions required for defecation (Peters et al., 2007).
To determine whether GON-2 and GTL-1 may play a role in the IP3 receptor signaling pathway, we fed gon-2;gtl-1 double mutant worms bacteria expressing dsRNA homologous to either PLCγ or PLCβ. As shown in Fig. 8 A, PLCγ RNAi had no additive effect on the pBoc arrhythmia induced by loss of function of both channels. Mean ± SEM pBoc period and coefficient of variance for gon-2;gtl-1;PLCγ(RNAi) worms were 73 ± 7 s and 49 ± 7% (n = 14). These values were not significantly (P > 0.09; see Fig. 2, A and B) different from that observed in gon-2;gtl-1 double mutant worms fed normal bacteria. In contrast, knockdown of PLCβ in gon-2;gtl-1 mutant worms induced a pBoc defect that was much more severe than that observed with the channel mutations alone (Fig. 8, B and C). Over a 20-min measurement period, no more than seven pBocs were observed in any of the gon-2;gtl-1;plcβ(RNAi) worms. The mean number of pBocs observed in 20 min in these animals was three (n = 12). One of the 12 animals examined exhibited no pBocs in this time period. This phenotype is remarkably similar to that induced by combined loss of function of PLCγ and PLCβ (Espelt et al., 2005) and suggests that GON-2 and GTL-1 function together with PLCγ to regulate IP3 receptor activity and ER Ca2+ release.
Discussion
The ORCa Channel Is Encoded by the TRPM Homologues gon-2 and gtl-1
The TRP cation channel superfamily is subdivided into TRPC, TRPV, TRPM, TRPML, TRPP, TRPN, and TRPA subfamilies. All TRP channels are comprised of six predicted transmembrane domains and intracellular N and C termini. Functional TRP channels are formed from homomeric or heteromeric association of four TRP subunits. TRP channels function in diverse physiological processes including sensory transduction, epithelial transport of Ca2+ and Mg2+, Ca2+ signaling, and modulation of membrane potential (Owsianik et al., 2006; Nilius et al., 2007).
The mammalian TRPM subfamily consists of TRPM1–8 (Kraft and Harteneck, 2005). GON-2 and GTL-1 share ∼23% identity with TRPM1, TRPM3, TRPM6, and TRPM7 (Baylis and Goyal, 2007). The conserved structural motifs in these channels are the transmembrane domains, the TRP domain, and portions of the cytoplasmic N terminus.
Amino acids that comprise the pores of TRPM6 and TRPM7 have been identified by mutagenesis and patch clamp analysis (Chubanov et al., 2007; Li et al., 2007; Topala et al., 2007). The homologous pore domains are nearly identical in GON-2 and GTL-1. This is consistent with our findings that the two channels have similar biophysical properties (Figs. 5 and Figs.6).
As shown in Fig. 4 A and Fig. 5 A, IORCa is dramatically inhibited by loss-of-function mutations in either gon-2 or gtl-1. Loss of function of both genes completely eliminates the current (Fig. 4 B). There are two possible explanations for these results. Either the ORCa whole cell current is comprised of independent GON-2 and GTL-1 currents, or the ORCa channel is a GON-2/GTL-1 heteromer. Our results suggest that the function of GON-2 and GTL-1 are interdependent. The combined inhibition of IORCa observed in gon-2 and gtl-1 mutant cells is ∼160% (Fig. 4 A). This finding indicates that GON-2 and GTL-1 can function independently as ion channels, but that maximal IORCa activity requires a functional interaction between them. One possibility is that the ORCa channel is a GON-2/GTL-1 heteromer. Alternatively, loss of either GON-2 or GTL-1 alone may disrupt the trafficking, expression, and/or regulation of the other channel.
Numerous studies have provided evidence that closely related TRP channels heteromultimerize (Owsianik et al., 2006; Nilius et al., 2007). Heteromultimers of TRPM6 and TRPM7, homologues of GON-2 and GTL-1, have been described (Chubanov et al., 2004; Li et al., 2006). At present, we favor the hypothesis that the ORCa channel is formed by association of GON-2 and GTL-1 subunits. However, extensive additional work including heterologous expression, mutagenesis, and subcellular localization of the two channels in vivo is required to test this idea.
Our electrophysiological findings differ from those of Teramoto et al. (2005). These investigators saw no effect of the gtl-1 deletion allele on whole cell current, whereas the gon-2 mutation reduced La3+-inhibitable outward current at +100 mV ∼75%. Current reduction was similar in intestinal cells cultured from gon-2 and the gon-2;gtl-1 double mutant worms. They also observed that the IC50 value for inhibition of the wild-type current by intracellular Mg2+ was 4.7-fold higher than that of the current observed in gtl-1 mutant cells. In contrast, we found that IORCa, IGON-2, and IGTL-1 exhibit similar sensitivities to intracellular Mg2+ (Fig. 6 C). Teramoto et al. (2005) concluded that GON-2 mainly mediates the outwardly rectifying current and that GTL-1 functions mainly to regulate current Mg2+ responsiveness. The reasons for the differences in our findings are unclear.
Role of GON-2 and GTL-1 in Oscillatory Ca2+ Signaling
Most TRP channels described to date have no or relatively low selectivity for Ca2+ over Na+ (Owsianik et al., 2006). The exceptions to this generalization are TRPV5 and TRPV6, which have PCa/PNa values >100 and play important roles in epithelial Ca2+ absorption (Vennekens et al., 2000; Yue et al., 2001; Owsianik et al., 2006). GON-2, GTL-1, and the ORCa channels exhibit a >60-fold selectivity for Ca2+ over Na+ (Fig. 6 B; Estevez et al., 2003). Mammalian TRPM channels are either impermeable to Ca2+ (TRPM4 and TRPM5) or have PCa/PNa values of 0.1–10 (Owsianik et al., 2006). Heterologously expressed Drosophila TRP and TRPL have relative Ca2+ permeabilities of 10–12 (Xu et al., 1997). Studies of the native TRP current in wild-type Drosophila photoreceptor cells indicate that the channel(s) responsible for the current are ∼40-fold more permeable to Ca2+ than monovalent cations (Hardie and Minke, 1992; Reuss et al., 1997). The endogenous Ca2+ conductances in trp and trpl mutant photoreceptor cells have PCa/PNa values of ∼4 and ∼86, respectively (Hardie and Minke, 1992; Reuss et al., 1997). Thus, together with mammalian TRPV5/6 and possibly Drosophila TRP, GON-2, GTL-1, and the ORCa channels have the highest Ca2+ selectivity of all characterized TRPs.
Given their exceptionally high Ca2+ selectivity and essential roles in maintaining pBoc and Ca2+ signaling rhythmicity (Fig. 2 and Figs.3), what possible functions could GON-2 and GTL-1 be performing? Data in Fig. 8 suggest that the channels function in a signaling pathway together with PLCγ to regulate IP3 receptor activity. Our previous studies failed to identify a significant role for the canonical store-operated CRAC channel in maintaining intestinal Ca2+ oscillations (Lorin-Nebel et al., 2007; Yan et al., 2006). Thus other Ca2+ channels must provide a Ca2+ entry pathway that allows for store refilling. It is conceivable that GON-2 and GTL-1 function in part to refill ER Ca2+ stores. However, even in the absence of these channels Ca2+ oscillations continue albeit arrhythmically (Figs. 2 and 3). This indicates that other Ca2+ entry pathways must function in the intestine to refill stores under these experimental conditions.
An attractive possibility is that the GON-2 and GTL-1 channels play a direct role in modulating IP3 receptor activity and controlling oscillation frequency. It is well established that IP3 receptors are regulated in a biphasic manner by intracellular Ca2+; low levels of Ca2+ activate the channels whereas high Ca2+ levels feedback and inhibit channel activity (Foskett et al., 2007). Foskett and coworkers (Mak et al., 1998; Foskett et al., 2007) have argued that Ca2+ is a true IP3 receptor agonist and that IP3 functions only to relieve Ca2+ inhibition. In excitable cells, plasma membrane Ca2+ influx through voltage- and ligand-gated Ca2+ channels can trigger intracellular Ca2+ release through ryanodine receptors via a process termed Ca2+-induced Ca2+ release (CICR) (Berridge et al., 2003). Plasma membrane Ca2+ influx can also trigger CICR via IP3 receptors (e.g., Kukuljan et al., 1997; Kapur et al., 2001; Gordienko et al., 2007).
The disruption of Ca2+ oscillation rhythmicity in gon-2 and gtl-1 mutants (Figs. 2 and 3) suggests that the channels function as part of the timekeeping apparatus that regulates cycle periodicity. We have shown previously that under conditions of low intracellular Ca2+ buffering, ORCa channel activity oscillates. Oscillating channel activity is due to a Ca2+ feedback mechanism similar to that observed with the IP3 receptor (Estevez and Strange, 2005). Such oscillating channel activity could provide a source of extracellular Ca2+ that functions to modulate IP3 receptor activity. Specifically, Ca2+ influx through ORCa channels could trigger IP3 receptor–mediated Ca2+ release via CICR. Rising cytoplasmic Ca2+ levels would feedback on both the IP3 receptor and ORCa channels functioning initially to increase and than eventually inhibit their activity. Calcium influx through ORCa channels would raise Ca2+ levels in channel microdomains and may also contribute to the overall increase in cytoplasmic Ca2+. Microdomain Ca2+ increases as well as the amplitude of the cytoplasmic Ca2+ increase would likely play a role in triggering downstream cellular functions.
Several TRP channels are known to be regulated by intracellular Ca2+ and play important roles in Ca2+ signaling. For example, the nonselective cation channel TRPM4 is activated by increases in intracellular Ca2+ (Launay et al., 2002). In T cells, TRPM4-mediated membrane depolarization modulates Ca2+ influx via CRAC channels and controls oscillatory Ca2+ signaling (Launay et al., 2004). TRPM5 is activated by intracellular Ca2+ concentrations of 0.3–1 μM and inhibited by higher Ca2+ levels and may function to couple intracellular Ca2+ release to membrane electrical activity (Prawitt et al., 2003). TRPC3 shows modest Ca2+ selectivity and initiates Ca2+ oscillations when activated by OAG. Increasing intracellular Ca2+ levels inhibits the channel (Grimaldi et al., 2003). Extensive additional studies using Ca2+ imaging, patch clamp electrophysiology, molecular biology, and forward and reverse genetics are needed to define the precise roles played by GON-2 and GTL-1 in intestinal Ca2+ signaling.
In conclusion, we have demonstrated that IORCa requires the combined function of the TRPM genes gon-2 and gtl-1. GON-2 and GTL-1 are highly Ca2+-selective channels and are essential for maintaining rhythmic Ca2+ oscillations in the C. elegans intestine. We postulate that GON-2 and GTL-1 form a heteromeric channel that selectively mediates Ca2+ influx and functions primarily to regulate IP3 receptor activity and possibly to refill ER Ca2+ stores.
Acknowledgments
We thank Dr. Jerod Denton for critically reviewing the manuscript and for many helpful discussions, Dr. Joel Rothman for providing the elt-2∷GFP-expressing worm strain, and Dr. Eric Lambie for providing the gon-2(q388) and gon-2;gtl-1 double mutant worm strains. Other strains used in this work were provided by the Caenorhabditis Genetics Center (University of Minnesota, Minneapolis, MN).
This work was supported by National Institutes of Health grant GM74229 to K. Strange.
Lawrence G. Palmer served as editor.
References
A. Estevez's current address is Biology Department, St. Lawrence University, Canton, NY 13617.
Abbreviations used in this paper: CICR, Ca2+-induced Ca2+ release; CRAC, Ca2+ release–activated Ca2+; dsRNA, double-stranded RNA; IP3, inositol-1,4,5-trisphosphate; ORCa, outwardly rectifying Ca2+; pBoc, posterior body wall muscle contraction; SERCA, sarco/endoplasmic reticulum Ca2+ ATPase; SOCC, store-operated Ca2+ channel.