The C. elegans T-type calcium channel CCA-1 boosts neuromuscular transmission

Abstract

Low threshold-activated or T-type calcium channels are postulated to mediate a variety of bursting and rhythmic electrical firing events. However, T-type channels' exact physiological contributions have been difficult to assess because of their incompletely defined pharmacology and the difficulty in isolating T-type currents from more robust high threshold calcium currents. A current in C. elegans pharyngeal muscle displays the kinetic features of a T-type calcium channel and is absent in animals homozygous for mutations at the cca-1 locus (see accompanying paper). cca-1 is expressed in pharyngeal muscle and encodes a protein (CCA-1) with strong homology to the alpha1 subunits of vertebrate T-type channels. We show that CCA-1 plays a critical role at the pharyngeal neuromuscular junction, permitting the efficient initiation of action potentials in response to stimulation by the MC motor neuron. Loss of cca-1 function decreases the chance that excitatory input from MC will successfully trigger an action potential, and reduces the ability of an animal to take in food. Intracellular voltage recordings demonstrate that when wild-type cca-1 is absent, the depolarizing phase of the pharyngeal action potential tends to plateau or stall near -30 mV, the voltage at which the CCA-1 channel is likely to be activated. We conclude that the CCA-1 T-type calcium channel boosts the excitatory effect of synaptic input, allowing for reliable and rapid depolarization and contraction of the pharyngeal muscle. We also show that the pharyngeal muscle employs alternative strategies for initiating action potentials in certain cases of compromised MC motor neuron function.

Fig. 1

Structure of the cca-1 gene and CCA-1 calcium channel. (A) Intron–exon structure of cca-1. Coding exons are represented by filled rectangles, non-coding exons by open rectangles. The ad1650 and gk30 deletions are indicated by bars. All cDNAs characterized were found to be trans-spliced at the 5′ end to the SL1 leader RNA (SL1). The most-abundant transcript, cca-1A, is composed of all exons except exon 19. Splice variant cca-1B uses an alternative splice donor in exon 18, which is spliced to exon 19. The cca-1D transcript splices exon 17 directly to exon 19. The inclusion of exon 19 would add amino acid residues to the pore loop region of domain II, and may have significant effects on the activity of the channel. Splice variant cca-1C uses an alternative 5′ splice donor for exon 24, which is predicted to result in the incorporation of an additional 15 amino acids in the III–IV loop. Sequences of the four splice variants have been submitted to GenBank under accession numbers AY313898 (cca-1A), AY313899 (cca-1B), AY313900 (cca-1C) and AY322480 (cca-1D). The PCR products used to generate the cca-1::GFP fusion constructs are diagrammed and labeled with the names of the primers used for amplification (cca35/40 and cca38/34). (B) Domain structure of CCA-1. Open cylinders indicate transmembrane domains, and the positions of the alternative exons found in splice variants B, C and D are shown.

Fig. 2

A cca-1::GFP gene fusion is expressed in the pharyngeal muscle. (A) Transgenic animals carrying an extra-chromosomal array including a full-length cca-1::GFP fusion (strain TS321) exhibit GFP fluorescence in all pharyngeal muscle including cells of the procorpus (pc), metacorpus, the isthmus (i) and in the terminal bulb (tb). The intense staining in the posterior of the terminal bulb is localized to the pm8 muscle cell. The asterisk indicates expression in extrapharyngeal neurons (n) and sheath cells (s) in the head. A differential interference contrast (DIC) image of the same worm is shown in B for comparison. Labeling as in A.

Fig. 3

Loss of cca-1 function or reduction of MC neurotransmission compromises action potential initiation and reduces pharyngeal pumping rates. (A) Effect of cca-1 mutations on the pumping rates of intact worms in the presence of E. coli HB101. Results are shown as the mean number of pumps per minute. Values are means ± standard error of the mean (s.e.m.). N=20 worms for each data point. See Materials and methods for additional details. (B) In electropharyngeograms (EPGs) from wild-type worms, a small EPSP (excitatory post-synaptic potential) precedes each excitation (E-phase) spike. An EPSP is marked with a large black arrow while an E-spike is marked with a small black arrow. The action potential terminates with a large negative repolarization (R-phase) spike (large gray arrow). Smaller negative spikes between E and R represent inhibitory post-synaptic potentials (IPSPs) from the inhibitory motor neuron M3 (Dent et al., 1997; Raizen and Avery, 1994). (C) In EPGs from cca-1 mutants homozygous for either of two loss-of-function alleles (ad1650 or gk30), small, delayed E-phase spikes (small arrows) follow the EPSPs. EPGs from cca-1 mutants also contain interpump phase (I-phase) spikes between pumps (arrowheads). (D) Effect on EPG traces of defects in MC neurotransmission. unc-17 encodes a membrane transporter that loads acetylcholine into synaptic vesicles within the cholinergic neurons, including the MC motor neuron. e245 is a viable missense allele of unc-17 (Alfonso et al., 1993). Null mutations of unc-17 are lethal. EPGs from unc-17(e245) mutants contain occasional I-phase spikes (arrowheads), because the low acetylcholine content of synaptic vesicles reduces the success rate of MC neurotransmission. snt-1 encodes synaptotagmin, a vesicle-associated protein necessary for effective calcium-stimulated release of neurotransmitter. snt-1(md290) is a putative null allele of synaptotagmin (Nonet et al., 1993). EPGs from snt-1(md290) mutant worms contain many I-phase spikes (arrowheads), often occurring in clusters, as a result of uncoordinated neurotransmitter release. Because EPGs recorded from different individual worms show some variation, we have annotated the features that consistently differ between worms of different genotypes. Unmarked differences are likely to be due to individual variation.

Fig. 4

Intracellular recordings reveal altered depolarization in action potentials from cca-1 mutant pharynxes. (A) Sample intracellular voltage recordings from wild-type and cca-1(ad1650) mutant worms. While the depolarization phase of the wild-type action potential usually rises steeply and steadily from resting potential to a peak, action potentials from cca-1 mutant worms often display a notch or flattened area early in depolarization. (B) Average maximal rising phase slopes for the wild-type, cca-1, eat-2 and eat-2; cca-1. 50–100 differentiated and aligned action potentials were averaged for each individual pharynx. A peak slope for each individual pharynx was obtained from the averaged trace. This calculation was performed for 19 wild-type animals, 16 cca-1 mutant animals, 14 eat-2 mutant animals and 11 eat-2; cca-1 double mutant animals. From the peak slopes for each individual, an average peak slope was calculated for each genotype, expressed as mean ± s.e.m. *Significantly different from the wild type; ^†significantly different from all other strains (P<0.005 for all comparisons, t-test). (C) (Top) Superimposed action potentials (with rising phases aligned) from two representative wild-type pharynxes. (Bottom) Time derivatives of the action potentials above. N=63 and 66 for left and right traces, respectively. (D) Same as in C for two representative cca-1(ad1650) worms; N=51 and 74 for left and right panels, respectively. While wild-type action potentials generally display a smooth steep rising phase, the majority of action potentials from cca-1 mutant worms contain an initial depolarization, followed by a flattened area or notch. Plotting the time derivative of each action potential trace reproduces the EPG phenotypes of wild-type and cca-1 mutant worms (compare with Fig.·3) with a small, delayed depolarization spike following each initial rise in membrane potential. (E) Action potentials from two representative eat-2(ad465); cca-1(ad1650) double mutants, labeled as in C and D. These action potentials lack the initial depolarization and plateau seen in cca-1 single mutants, suggesting that early depolarization in a wild-type pharynx or cca-1 mutant pharynx is mediated by the EAT-2-containing nicotinic receptor.

Fig. 5

CCA-1 activity is crucial for the efficient initiation of action potentials in animals with defects in cholinergic neurotransmission. (A) EPGs from unc-17(e245); cca-1(ad1650) double mutants contain many I-phase spikes, reflecting the fact that MC is often unsuccessful in triggering a pharyngeal muscle action potential. The frequency of I-phase spikes is far greater in an unc-17(e245); cca-1(ad1650) double mutant than in an unc-17(e245) single mutant (shown in Fig.·3D). (B) EPGs from snt-1(md290); cca-1(ad1650) mutant worms contain rare action potentials separated by many clusters of I-phase spikes. Again, the rate of failure of action potential generation is far greater in a snt-1(md290); cca-1(ad1650) double mutant than in a snt-1(md290) single mutant (shown in Fig.·3D). EPSPs are marked with large arrows. E-phase spikes are marked with small arrows. I-phase spikes are marked with arrowheads.

Fig. 6

Worms lacking the EAT-2 nicotinic receptor subunit initiate action potentials without MC-stimulated EPSPs. (A) EPGs from eat-2 mutants lack EPSPs and I-phase spikes because eat-2 mutants are defective in nicotinic neurotransmission from MC to the pharyngeal muscle. The ad465 allele of eat-2 contains an early stop codon. (B) EPGs from eat-2; cca-1 double mutants closely resemble those of eat-2 single mutants. E-phase spikes are marked with arrows.

Fig. 7

Worms adapt to the loss of nicotinic MC stimulation by altering resting potential to preserve pharyngeal pumping. (A) Sample intracellular voltage recordings from wild-type worms show that resting potential is about −73·mV. (B) Resting potential in cca-1 single mutants matches that of wild-type worms. (C) In contrast, a recording from an eat-2 mutant reveals that resting potential is elevated by ~13·mV over wild-type resting potential, and has a slight tendency to rise between action potentials. (D) A recording from an eat-2; cca-1 double mutant reveals an elevated resting potential that drifts significantly towards positive potentials between action potentials. (E) Average resting membrane potential for above four strains. Mean ± s.e.m.; N=12 (WT), 10 (cca-1), 15 (eat-2), 11 (eat-2; cca-1). *Significantly different from both the wild type and cca-1 (P<0.005, t-test).

Fig. 8

Proposed role of the CCA-1 T-type channel in the pharyngeal action potential. In wild-type worms, action potentials are triggered by the influx of cations through the EAT-2-containing acetylcholine gated channel. The small rise in membrane potential resulting from cation influx is sufficient to activate the CCA-1 T-type channel. Calcium influx through CCA-1 lifts membrane potential further, until the EGL-19 L-type calcium channels can be activated. In the absence of CCA-1 function, some EPSPs fail to activate EGL-19 channels, or activate them more slowly than in wild-type worms. When both MC EPSPs and CCA-1 function are absent, the pharynx employs an alternative method, possibly a leak conductance and an elevation of resting membrane potential, to trigger EGL-19 activation.