| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
MEMBRANE TRANSPORTERS, ION CHANNELS, AND PUMPS
2Departments of Anesthesiology and Pharmacology, Digestive Disease Research Center, Vanderbilt University Medical Center, Nashville, Tennessee; and 3Interdepartmental Graduate Program in Neuroscience, Center for Aging and Developmental Biology and 1Nephrology Division, Department of Medicine, University of Rochester Medical Center, Rochester, New York
Submitted 19 July 2007 ; accepted in final form 13 October 2007
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
calcium; oscillation; Caenorhabditis elegans; biosensor
The nematode Caenorhabditis elegans is a genetic model organism that exhibits a rhythmic defecation behavior that is timed by Ca2+ signaling (for review, see Ref. 2). The wild-type hermaphrodite executes a defecation motor program (DMP) at 
45-s intervals in the presence of food (18, 32). The DMP is highly stereotyped, consisting of three consecutive sets of muscular contractions (32). Initiation of the cycle occurs as the intestinal contents are pressurized by a simultaneous contraction of both dorsal and ventral posterior body wall muscles (pBoc). This is followed several seconds later by contraction of the anterior body wall muscles (aBoc), which drives the pharynx in the intestine and moves the contents of the lumen backward. Finally, a group of muscles around the anus is responsible for expulsion (Exp).
The timing and rhythm of the defecation cycle is remarkably consistent, with all of the hallmarks of a molecular clock. A genetic screen for defecation timing mutants identified the gene itr-1, which encodes the single worm inositol 1,4,5-trisphosphate receptor (IP3R), an intracellular Ca2+ release channel (32). The phenotypic effects of itr-1 on defecation were shown to be conveyed via its activity in the 20 epithelia cells that comprise the worm intestine (32), and mutations in the itr-1 gene or overexpression of the IP3R in the intestine affected cycle timing (6), whereas expression of an IP3 "sponge" disrupted defecation (35). Finally, optical recordings using the Ca2+-sensitive dye fura 2 revealed that activation of the DMP correlates with an elevation in Ca2+ in posterior intestinal cells (6), suggesting that Ca2+ signals in the intestine comprise part of the defecation clock.
Recently, several laboratories have employed a combination of genetics and physiology to lend further insight to the molecular mechanisms that regulate rhythmic Ca2+ oscillations in the intestinal epithelia. The laboratory of Dr. K. Strange has developed a system for studying Ca2+ signals in partially isolated intestines. They demonstrated that a Ca2+ wave initiates rhythmically at the posterior end of the intestine and propagates toward the anterior end with a period similar to that of the defecation cycle (7). Furthermore, they showed that two of the six individual worm phospholipase C (PLC) genes, whose products generate inositol 1,4,5-trisphosphate (IP3) and diacylglycerol (DAG) from the hydrolysis of phosphatidylinositol 4,5-bisphosphate (PIP2), help to maintain the defecation rhythm through genetically separate pathways (7). Along similar lines, the laboratory of Dr. K. Iwasaki developed a system for imaging Ca2+ using a yellow cameleon expressed in the intestine of intact worms that have been immobilized on agarose pads and showed, in this experimental model, that posterior-to-anterior Ca2+ waves in the intestine coincided with activation of the DMP (31). The molecular mechanisms that mediate the propagation of this Ca2+ wave are not well understood, although the work of the Teramoto and Iwasaki (31) demonstrated that wave propagation is required to trigger the aBoc and Exp steps of the DMP.
Significantly, the dynamics of Ca2+ wave propagation have not previously been investigated under physiological conditions, and it is currently unknown whether observations from existing models reflect normal physiology. This is an important concern, since the execution of the DMP is known to be regulated in response to mechanical stimuli and food-related sensory cues (3, 18, 29, 32). Here, we describe a system that we have developed for following Ca2+ waves in intact, freely behaving worms using a transgenic YC6.1 biosensor. Under these conditions, worms exhibited a normal, rhythmic activation of the DMP that coincided with Ca2+ wave propagation through the intestine. However, Ca2+ wave dynamics were distinct from those observed in restrained or reduced preparations: the wave initiated simultaneously from both the anterior and posterior intestinal cells and propagated inward. We found that genetic perturbations associated with arrhythmia in the defecation cycle disrupt the propagation of this wave in specific ways. Loss of PLC-B altered both the timing and the site of Ca2+ wave initiation, whereas disrupting KCNQ K+ channel function severely impaired the ability of the wave to propagate. Furthermore, the disruption of CaM-CaMKII signaling increased the frequency of wave propagation, resulting in the reiteration of the DMP at less than a normal cycle length. Surprisingly, we also found that increasing the duration of cellular Ca2+ oscillations through reduction in sarco(endo)plasmic reticulum Ca2+ ATPase (SERCA) function did not extend the cycle period appreciably but instead disrupted the coordination between sequential muscular contractions. Based on our results, we propose a model where Ca2+ waves are shaped spatially by their specific cellular foci of initiation and temporally by an underlying, higher-frequency oscillator. Taken together, these findings demonstrate the power of our system to elucidate the molecular mechanisms controlling ultradian rhythms and the propagation of Ca2+ waves in an awake, behaving animal.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Strains were made available by the C. elegans Genetics Center. The deletion mutants were obtained from S. Mitani at the C. elegans National BioResource Project in Japan. Each deletion strain was crossed to N2 wild-type worms three times before use. The deletion break points have been mapped, which allowed us to formulate a PCR-based genotyping assay that includes three primers for each strain, two of which flank the deletion and one of which anneals within the deleted sequences. For kqt-2 genotyping, the primer sequences are as follows: forward 1 (F1): ACTTTTCCCTTAGTTTCTTCGCAAG; forward 2 (F2): CCGTTGTCTTGAATTCCTTGAGC; and reverse 1: TGGGACTCTGCTGTTTTACGC. The wild-type PCR product(s) are 309 and 1180 nucleotides long, whereas the tm642 allele produces a 379 nucleotide product. For kqt-3 genotyping, the primer sequences are as follows: F1: TTGTCTAGAACGCCCGACTG; R1: TTTGGGAAGCAGCCGACATC; and reverse 2: TCCTAAAACCGCGCCTTGTAG. The wild-type PCR product(s) are 457 and 1458 nucleotides long, whereas the tm542 allele produces a 261 nucleotide product. In most cases, the shorter wild-type genomic PCR product was preferentially amplified over the longer product, as would be expected.
RNA interference. Freshly transformed colonies of the bacterial strain HT115 were grown to midlog phase at 37°C and induced for 1 h with 1 mM IPTG. After fivefold concentration, 80 µl of bacteria were added to the surface of 35- or 60-mm NGM agar or agarose plates supplemented with cholesterol and 0.1 mM IPTG. Later (2 days), L3 larvae were placed on the RNAi plates for 36 h before being moved to a fresh plate, where they were allowed to lay eggs. These progeny were then used for further experiments.
Vectors and creation of transgenic strains. RNAi vectors were created by PCR cloning of from 600 to 1000 nucleotides of the coding region from the target gene in the Fire lab double-strand RNA feeding vector pPD129.36 (Courtesy of A. Fire, Stanford University). The vector pKT2 is a derivative of pFH6.II (14) containing the YC6.1 open reading frame (ORF) in place of green fluorescent protein (GFP). The vector pXXY2004.1, containing the RDE-1 ORF, was a kind gift from K. Strange (Vanderbilt University). The nhx-2 promoter (24) was used in these vector backbones to drive the expression of RDE-1 and YC6.1 specifically in the intestine. Anti-GFP positive and anti-pos-1negative controls were used to verify cell-specific RNAi in this strain (data not shown). Furthermore, vav-1(RNAi) worms were shown to have a modest arrhythmia [coefficient of variation (COV) of 43.9 ± 31.0%], as demonstrated previously (25), whereas itr-1(RNAi) worms were viable, suggesting the absence of RNAi in the gonadal sheath cell, but had defecation frequencies of >5 min because of a lack of Ca2+ oscillations (Table 1).
|
Ca2+ imaging.
Transgenic nematode larva (generally L3-L4 larva) expressing YC6.1 from the nhx-2 promoter in the intestine were imaged on 60-mm NGM-agarose plates (±IPTG for RNAi) seeded with 
the amount of bacteria normally used, since the bacteria, like agar, can increase the background fluorescence. Optical recordings were performed on a Nikon 2000U inverted microscope coupled to a monochromatic light source (Polychrome IV; Till Photonics), CCD camera detection system (Cooke, Germany), and optical beamsplitter (Optical Insights, Tucson, AZ). Nikon Plan Apo x10 [0.5 numeric aperture (NA)] or x20 (0.85 NA) air objectives were used for routine imaging.
For YC6.1 acquisitions, we used a 40- to-100-ms exposure time and 4- to 20-Hz frame rates. The fluorescence intensity of YC6.1 was generally homogenous throughout the intestinal cells, and the reporter generated a signal-to-background ratio of at least five at each emission wavelength. After excitation at 435 nm, the emissions were acquired and saved simultaneously at both 480 and 535 nm using TILLvisION software (TILL Photonics). Background emission images were generated using the same acquisition settings and subtracted from each experimental image on a pixel-by-pixel basis. FRET measurements were calculated using the TILLvisION FRET module, which generates an emission ratio (R) image overlay. With the above acquisition parameters, thresholding the overlay to a minimum pixel density of approximately one-half the background value eliminated nonspecific or weak signals so that only the intestine was visible. Intestinal cells were outlined using a single channel emission image as a guide and then updated in successive images to account for movement. Cellular Ca2+ dynamics were determined by R/R0, using a series of four to eight frames before defection to normalize the emission ratios. pBoc was denoted post hoc from the image series as a pBoc contraction of the intestine and aBoc as a depression of the pharynx in the lumen of the intestine, which displaces the anteriormost segment of int-1. Similarly, Exp was marked by a short (0.25–0.50 s) contraction of the enteric muscles. The ratio change in the YC6.1 reporter was 1.6-fold from minimum to maximum Ca2+ (data not shown).
Analysis of pBoc and pharynx pumping.
pBoc was monitored at room temperature in either L4 larval (Table 1) or 2-day-old adult worms (see 
Fig. 3) swimming on a lawn of OP-50 bacteria grown on agar plates. Worms were imaged using a Carl Zeiss MicroImaging Stemi SV11 M2BIO stereo dissecting microscope (Kramer Scientific) equipped with a DAGE-MTI DC2000 CCD camera or Nikon SMZ 800 stereo dissecting microscope. pBoc rhythmicity in individual worms was assessed by calculating the coefficient of variance (CV), which is the SD of the pBoc period expressed as a percent of the mean, from 10 successive contractions.
|
|
|
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Previous studies using isolated intestines or immobilized animals have shown that a Ca2+ wave propagates through the intestine from the posterior to anterior end (7, 31). However, we observed inconsistent muscular contractions and poorly propagating waves when the worms were immobilized (Supplemental Fig. S1; Supplemental data for this report may be found on the American Journal of Physiology: Cell Physiology web site.). We therefore measured Ca2+ using low-power air objectives to follow live nematodes on an agarose plate under normal physiological conditions, which we term "free range" imaging. Like the system implemented by Faumont et al. (8, 9) to characterize behavioral responses to Ca2+ flux in chemosensory neurons of partially restrained worms, this protocol allowed us to observe the integration of Ca2+ flux, muscular contractions, and neural modulation of the DMP in intact, living animals.
Imaging was facilitated by crossing an "uncoordinated" unc-31(n422ts) allele in the Ca2+ reporter strain. Unc-31 encodes a Ca2+ activator of protein secretion that functions in the G
s pathway in cholinergic motor neurons to regulate locomotion (4). The n422ts allele is temperature sensitive, produces a relatively mild deficit at 20°C, the temperature of our assays, and has normal defecation cycle timing and execution (Table 1). In addition, intestinal calcium dynamics in the unc-31(n422ts) background were indistinguishable from wild-type controls (data not shown).
Execution of the DMP and intestinal Ca2+ oscillations occurred simultaneously (Fig. 1A and Table 1), as predicted from the existing model (6, 7, 25, 31). Muscular contractions could be readily observed post hoc, and, as shown in Table 1, the average defecation period and rhythm calculated from these calcium oscillations were indistinguishable from those determined visually under a dissecting microscope.
The ratio map of yellow fluorescent protein (YFP)/cyan fluorescent protein (CFP) fluorescent emissions from YC6.1 in Fig. 1B is representative of the Ca2+ wave routinely observed during defecation (n > 50 worms). Although it has been suggested that a single Ca2+ wave progresses posterior-to-anterior through the intestine (31), we instead observed two waves that initiated at both the anterior and posterior ends of the intestine and converged (Fig. 1B and Supplemental Movie S1). To quantify the temporal and spatial characteristics of these waves, regions of interest were drawn around groups of intestinal cells using a single channel emission image as a guide, and R/R0 values for the YFP/CFP emission ratio were determined (Fig. 1, B and C). These regions of interest were updated manually as the worm moved or as contractions occurred until defecation was complete. The muscular contractions were denoted post hoc by examining single channel emission recordings as stated.
Ca2+ waves initiated from both ends of the intestine nearly simultaneously and propagated through the entire organ in <1 s (Fig. 1, B and C). Ca2+ remained elevated for 4 s, with the anterior and posterior cells returning to baseline before cells located more centrally. pBoc coincided with the elevated Ca2+ in the posterior cells, whereas aBoc appeared to occur as Ca2+ returned to baseline in the anterior cells, suggesting that contractions of the aBoc are not simply triggered by high Ca2+. Evidence of FRET was reflected by the inverse relationship between the single channel emission intensities during a rise in Ca2+ (Fig. 1D) in both the anterior and posterior cells, suggesting that the "backward" wave does not reflect some unaccounted for environmental sensitivity of YC6.1. Moreover, we generated a strain expressing the cameleon YC2.12 such as was used previously (31) and, employing our free range technique, found the same pattern of dual initiation and convergent propagation as shown in Fig. 1 (Supplemental Fig. S2); in fact, we observed convergent Ca2+ waves in every mutant background used in this manuscript (see below). At present, we cannot reconcile these observed differences, other than to comment on distinct imaging techniques. However, we agree fully with the previous observation that the posterior-to-anterior wave is necessary for coupling and execution of the complete DMP, as shown below.
PLCβ regulates the site of Ca2+ wave initiation.
Recent reports have identified several gene products that function upstream of the IP3R in the intestine to regulate the production of IP3, and reducing their function can lead to cycle length defects as well as arrhythmia (7, 16, 25). These include PLC
, which generates IP3 and DAG from the hydrolysis of membrane lipid PIP2, and loss of the PLC coding gene plc-3 can be suppressed by a gain-of-function mutation in itr-1 that reduces reliance upon upstream signaling pathways (8). However, the nematode PLCβ ortholog EGL-8 apparently works through a mechanism distinct from PLC-3, since an itr-1 gain-of-function cannot suppress arrhythmia in the egl-8(n488) mutant.
Intestinal Ca2+ oscillations were arrhythmic and occurred with reduced amplitude in the egl-8(n488) mutant (Fig. 2A), similar to the oscillations observed previously in isolated intestines (7), and consistent with arrhythmic defecation (Table 1). However, instead of coordinated initiation at the anterior and posterior ends of the intestine, the mutant worms exhibited random increases in Ca2+ that propagated from cell to cell in the form of a wave, but often in the wrong direction, as shown by the representative image series in Fig. 2C. Out of approximately 20 mutant worms imaged, all displayed seemingly random oscillations, with approximately an equal probability of initiating at the anterior end, the posterior end, or in the middle of the intestine (data not shown). We hypothesize that previous observations of diminished muscular contractions in egl-8 mutants (17) may have to do with reverse orientation waves such as these being able to elicit a pBoc contraction, albeit weakly. However, we never observed aBoc or Exp following a reverse wave. These observations strongly suggest that both the orientation and propagation of the Ca2+ wave(s) are required for contraction coupling. We further postulate that all intestinal cells are capable of initiating Ca2+ waves and that PLCβ reinforces the supremacy of the posterior cells as the defecation pacemaker.
EGL-8 has also been shown to participate in signaling at neuromuscular junctions (17). To directly test if EGL-8 regulates Ca2+ oscillation/wave initiation in intestinal cells through cell-autonomous mechanisms, we used cell type-specific RNA interference of intestinal EGL-8 as reported previously (7). In short, the rde-1(ne219) mutant is resistant to RNAi: we rescued RDE-1 activity exclusively in the intestine by driving its expression via the nhx-2 promoter and included the YC6.12 reporter in the transgenic array to be able to measure Ca2+. Both the parent and transgenic rde-1 strains had slightly increased defecation periods compared with N2 worms, although the difference barely reached significance (P = 0.048). Control RNAi treatments that demonstrate the efficacy and specificity of the strain are given in Table 1 and in Fig. 6. We found that intestinal egl-8(RNAi) caused the same defecation phenotype as the egl-8(n488) mutation (Fig. 2B and Table 1). This suggests that EGL-8 limits intestinal cell Ca2+ wave initiation through cell intrinsic activity.
|
Both the kqt-2(tm642) and kqt-3(tm542) deletion alleles were associated with an increase in the defecation period (Fig. 3, A–C and Table 1). In contrast to the regular intervals at which Ca2+ oscillated in wild-type worms (Fig. 3D), both of these kqt loss-of-function mutations caused irregular oscillations, correlating with the increased COV for the defecation period (Fig. 3, E and F). However, the duration that Ca2+ remained elevated was longer than the wild type as well, and Ca2+ oscillations that did not provoke DMP execution were noted in both mutants (Figs. 3, E and F). These data are consistent with the idea that KQT channels have an important role in promoting Ca2+ oscillations in the intestine, possibly by controlling the electrical driving force on Ca2+ influx. The lengthened duration may reflect a slower influx of Ca2+, whereas the nonproductive intermediate oscillations could result from failed wave propagation.
To test this hypothesis, we examined Ca2+ wave propagation. Significantly, the intermediate oscillations were found to be the result of Ca2+ waves initiating at the anterior end of the intestine and progressing slowly toward the posterior end, but without the corresponding posterior-to-anterior wave initiating (Fig. 4 and Supplemental Fig. S3), consistent with our observations of egl-8 mutants in supporting the idea that the posterior-to-anterior wave is necessary to trigger DMP execution. Even more telling was the fact that, during normal defecation, posterior Ca2+ oscillations in the kqt mutants were invariably preceded, often at great length, by anterior oscillations (Fig. 3, H and I). This phenomenon could reflect inherent differences in the regulation of wave generation in the anterior and posterior cells of the intestine, such as a more prominent role for KQT channels in promoting generation of the posterior wave. Alternatively, there could be some dependency between the anterior and posterior oscillation events. Although we are unable to ascribe a physiological function to the reverse wave, the fact that it uniformly precedes the posterior-to-anterior wave suggests that, if not necessary, it may at least be a harbinger of the DMP.
|
The traces shown in Fig. 5, A and B, show intestinal Ca2+ oscillations over a single cycle from two representative unc-43(sa200) mutants. To portray the dynamics of the Ca2+ waves that occurred in these mutants, YC6.1 ratio emission images were extracted at the times indicated in each of the traces, as shown to the bottom. In one of the mutants (Fig. 5A), normal DMP execution (stars) was followed by reiteration of the DMP, or a shadow contraction, one-third of a cycle length later (thick arrow). The Ca2+ waves that occurred during this shadow contraction were indistinguishable from those that coincided with normal DMP execution. However, there was extensive heterogeneity, both among worms and among successive cycles in the same worm, in the degree to which the shadow contractions occurred (18). The mutant in Fig. 5B had a very weak, almost unnoticeable, pBOC contraction, aBOC/Exp failed to occur, and the Ca2+ waves were reduced compared with the norm.
|
To ask whether CaMKII acts in the intestine to regulate Ca2+ oscillations, we first attempted intestine-restricted RNAi to limit loss-of-function to this tissue alone. However, we were unable to phenocopy the unc-43(sa200) mutation using feeding RNAi, which is unsurprising given that unc-43 RNAi has failed to elicit any phenotype from the many global RNAi screens performed in worms and may thus be refractory to RNAi. To provide alternate suggestive evidence that UNC-43 may act in the intestine to regulate defecation, we adopted the following approach: we targeted intestinal cmd-1, the single worm calmodulin ortholog. The cycle period was extended by this treatment, and the oscillations were increased in duration, which is unsurprising given that cmd-1 likely regulates multiple targets (including the kqt channels through an inhibitory binding site). However, similar intracycle oscillations were observed as in the CaMKII mutant (Fig. 5C), consistent with a cell-autonomous role for CaM-CaMKII signaling in regulating Ca2+ flux.
Together, our data suggest that cryptic intracycle Ca2+ oscillations underlie the shadow cycles of unc-43 mutants. We speculate that the defecation rhythm is maintained by a pacemaker activity that operates at several times the frequency of the cycle period and that CaMKII normally suppresses the functional response to this pacemaker, perhaps via modulating the activity of the IP3 receptor.
The SERCA SCA-1 is required for contraction coupling during the DMP. Membrane Ca2+ ATPases are responsible for extruding Ca2+ from the cell and, in the specialized case of the sarco(endo)plasmic isoform SERCA, refilling endoplasmic reticulum (ER) stores following depletion. These proteins would be predicted to help restore intestinal cytoplasmic Ca2+ to baseline levels following defecation. C. elegans contains four membrane Ca2+ ATPase genes (5, 38). To test how Ca2+ clearance rates affect defecation, we assessed the contribution of each of the four mca (sca) genes to defecation timing using RNAi. Because loss of sca-1(mca-4), the worm SERCA ortholog, results in embryonic and larval lethality (5, 38), we examined defecation using cell-type specific RNAi as described above to limit loss-of-function to the intestine.
To test our strain, we first treated the worms with itr-1 RNAi. We were able to phenocopy the effects of itr-1(lf), as shown in Table 1, and intestinal Ca2+ oscillations were suppressed entirely relative to the control-treated animals (Fig. 6, G and H). Because itr-1 regulates defecation via its activity in the intestine, this confirms the susceptibility of the intestine to RNAi in this rescued stain. However, the itr-1 RNAi-treated worms were fertile (data not shown), whereas an itr-1 deletion strain is inviable and forms dead "rod-like" larva, and N2 worms treated with itr-1 RNAi have a pleiotropic phenotype, including larval arrest. These results further suggest that the effects of RNAi were limited to the intestine in our strain.
The sca-1(RNAi) worms grew to adulthood, and, of the four genes tested, only the sca-1(RNAi) animals had a defecation phenotype (data not shown). As might be predicted from the role of SERCA in refilling ER stores, most of the sca-1(RNAi) worms did not defecate nor did they exhibit rhythmic Ca2+ oscillations. Also, as demonstrated previously, the majority of animals died of an unfolded protein response (37). However, because RNAi is rarely 100% effective, a fraction of the worms continued to defecate, even as adults. For these worms, the cycle times were slightly extended compared with controls (Fig. 6, A–C), but the average COV was nearly the same, arguing against arrhythmia (Table 1). However, the interval between pBoc and aBoc, normally around 5 s, was greatly extended (Fig. 6, D–F; Table 1). Interestingly, pBoc-to-aBoc intervals were relatively consistent within a single worm from cycle to cycle, although the average pBoc-to-aBoc interval could differ dramatically between worms, which may arise from heterogeneity in the effectiveness of the RNAi treatment.
We next followed Ca2+ levels in the intestine and assessed when each stage of the motor program was executed relative to the changes in Ca2+. The amount of time that Ca2+ was elevated over baseline in the sca-1(RNAi) worms (Fig. 6I) was significantly greater than in control worms (Fig. 6G), consistent with a role for SERCA in Ca2+ clearance. In fact, although the defecation period was relatively unaltered, Ca2+ could remain elevated in the intestinal cells of the sca-1(RNAi) worms for up to 
of the period. We also found that the pBoc contraction persisted longer than in control worms (data not shown) and that aBoc was coordinated with a reduction of Ca2+ levels specifically in the anterior intestinal cells, regardless of the length of the pBoc-aBoc interval (Fig. 6, J and K). The amount of time that Ca2+ remained elevated in the anterior intestinal cells correlated well with the pBoc-to-aBoc interval between these two contractions (Fig. 6M).
Because the effects of RNAi were limited to the intestine in the strain used for the experiments shown in Fig. 6, the argument can be made that the physiological role of SCA-1 in coordinating pBoc and aBoc occurs cell autonomously in the intestine. However, whether the reduction in Ca2+ that coincides with aBoc reflects an actual consequence of store refilling, cytoplasmic Ca2+ reduction, or is a result of some other facet of SERCA activity is currently unknown.
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
EGL-8: insight into the molecular control of bidirectional Ca2+ wave propagation. We have shown that unrestrained nematodes exhibit Ca2+ waves that initiate simultaneously at both the anterior and posterior end of the intestine and converge around the midbody (Fig. 1). This is in contrast to previous work observing only a posterior-to-anterior Ca2+ wave, either coinciding with DMP execution (31) or in dissected intestinal preparations (7). The discrepancy could be because of the nature of the experimental setup; in the free-range model system, worms are completely unrestrained and able to move normally on an agarose plate containing their preferred food source. We conclude that, under normal physiological conditions, two physically separate sets of intestinal cells are able to coordinately oscillate, although we are at present unable to suggest a physiological role for the reverse wave.
Our observations that the loss of PLCβ leads to ectopic Ca2+ waves further suggest that most, if not all, intestinal cells are capable of initiating Ca2+ oscillations (Fig. 2). EGL-8 has been shown to act at neuromuscular junctions in a presynaptic pathway that includes GOA-1 (Gq) and UNC-13 (17) and additionally regulates several facets of male mating behavior upstream of the IP3R gene itr-1 (12). However, egl-8 is also expressed throughout the intestine (23), with the highest expression levels observed in the posterior cells (17), and both we and others have shown that intestinal RNAi phenocopies the loss-of-function mutation, suggesting cell-intrinsic activity. Unlike its role in male mating behavior, Espelt and colleagues (7) reported that defecation arrhythmias associated with the loss-of-function mutation could not be suppressed by a gain-of-function mutation in itr-1, and the failure of exogenous DAG to suppress the arrhythmia argued against a function analogous to its role in neurotransmission, as well.
So how does EGL-8 regulate defecation? In many systems, PLCβ acts downstream of G protein-coupled receptors. Also, in fact, the Rho/Rac family GEF VAV-1 has been shown to regulate multiple rhythmic behaviors in worms, including defecation. Moreover, the Rho/Rac small GTPases ced-10, mig-2, and rho-1 also have redundant functions in structuring the defecation rhythm (25). However, the loss of VAV-1 can be suppressed by an itr-1 gain-of-function mutation and VAV-1 has been shown to act upstream of PIP5K, which generates PIP2, the substrate for IP3 production (25). These results suggest that VAV-1 may function in the same pathway as PLC to generate substrate for the IP3R, rather than the PLCβ pathway. In support of this idea, we find that intestinal RNAi of vav-1 results in arrhythmia (Table 1) but that Ca2+ oscillations in the intestine nonetheless initiate in both anterior and posterior cells, as normal; the same is true for ced-10, mig-15, and rho-1 (data not shown). Instead, EGL-8 may act primarily to deplete PIP2 rather than to generate IP3. PIP2 (along with PIP3) has been shown to target many proteins to the plasma membrane, including Ras, Rab, Arf, and Rho proteins (15).
It has also been suggested that PLC acts as a positive feedback mechanism during stimulation-induced Ca2+ oscillations because its activity is increased by cytosolic Ca2+ and that this feedback would result in regenerative IP3 oscillations that fuel spikes in Ca2+ (22). The egl-8(lf) phenotype is consistent with such a role for this gene. If egl-8 activity served to set a threshold for regenerative spiking, spatial heterogeneity in the activity of this gene would be expected to result in preferred sites of wave initiation. Indeed, egl-8 is most highly expressed in the posterior intestinal cells, which have been consistently identified as a primary site of wave initiation in numerous studies. Furthermore, a reduction of egl-8 activity would be expected to result in spatially unpredictable and arrhythmic Ca2+ spikes, as is indeed observed with egl-8(lf). However, it remains unclear as to why egl-8 would preferentially play this role instead of plc-3.
A very recent report describes the role of the worm intestinal pannexin gap-junction protein INX-16 on Ca2+ wave oscillations (27). Intriguingly, the arrhythmia and random or reverse wave propagation described in this report mirrors closely the ectopic Ca2+ oscillations that we observe in the egl-8(lf) mutant. Although this is an important independent confirmation of our hypothesis that all intestinal cells are capable of initiating a Ca2+ wave, it also raises the question of whether one role of EGL-8 may be to regulate INX-16 or gap-junction behavior. In future work, the Ca2+ oscillation phenotype that we have identified in egl-8 mutants will undoubtedly help to differentiate between mechanistically distinct physiological regulators of the defecation rhythm and elucidate the signaling pathways in which they function.
Ca2+ oscillation dynamics and rhythmicity in the defecation cycle: the role of KQT channels. In addition to IP3-mediated release of Ca2+ from internal stores, Ca2+ influx from outside the cell is also required for generation of Ca2+ oscillations in the intestine, although the molecular pathway that mediates Ca2+ entry remains unknown (19, 37). Noting the defecation and Ca2+ oscillation phenotypes observed in the kqt mutants, we postulate that KQT channels regulate Ca2+ entry through the generation of a strong electrical driving force. We propose the following model: Ca2+ entry and/or IP3 generation by PLC evokes Ca2+-induced Ca2+ release (CICR) and a spike in intracellular Ca2+ levels. Following this increase, Ca2+ inhibits KQT channel activity through CaM, which is bound to the IQ motif present in these channels. Feedback inhibition of KQT channel activity then depolarizes the resting potential and suppresses Ca2+ influx. Reduced Ca2+ entry and sequestration/buffering of intracellular Ca2+ relieves Ca2+-dependent KQT channel inhibition and restores the driving force for plasma membrane Ca2+ entry for the next cycle of IP3-evoked CICR from internal stores.
Three predictions follow from our working model. If KQT channels regulate plasma membrane Ca2+ influx required for activation and termination of CICR, then disrupting KQT channel function should 1) augment the duration of Ca2+ oscillations and 2) diminish the onset and 3) diminish the amplitude of Ca2+ spikes. As shown in Fig. 3, Ca2+ oscillation duration in kqt mutants is two to three times longer than that of wild-type animals (also see supplemental Fig. S4). Consistent with the role of CaM in our model, the oscillations following cmd-1 RNAi appear to be extended compared with the wild type, as well (Fig. 5). We also observe a slow onset of the Ca2+ increase in both of the kqt mutants, potentially reflecting the uncoupling of plasma membrane Ca2+ influx and IP3 generation and dysregulation of CICR. Finally, our model is further supported by the observation of small-amplitude Ca2+ signals, which often failed to evoke body wall muscle contractions. We postulate that these nonproductive Ca2+ transients represent Ca2+ release from internal stores that fail to develop into full-amplitude Ca2+ oscillations. This could be attributed to an uncoupling of plasma membrane Ca2+ influx and Ca2+ liberation from the ER. We note that these nonproductive Ca2+ oscillations in particular are often more persistent than oscillations that are linked to the DMP (for example, see Fig. 5), suggesting that the speed of Ca2+ increase is directly related to the time it remains elevated.
Intestinal Ca2+ waves coordinate execution of the DMP. Our observations suggest that the posterior-to-anterior wave is a necessary precursor to execution of the complete DMP and that wave directionality is important for functional coupling of the muscular contractions (Fig. 2). In none of the strains examined in our study did we observe the initiation of the DMP in the absence of the posterior-to-anterior wave [although a weak pBoc alone was triggered in the egl-8(lf) mutant during reverse waves that traveled through the posterior intestine, but in the wrong direction]. Moreover, the extent of posterior-to-anterior wave propagation was positively correlated with successful execution of the full DMP. Indeed, examination of the shadow cycle unc-43(sa200) mutants suggested the existence of dependencies between the anterior and posterior oscillations and the execution of the DMP. Strong Ca2+ oscillations in the anterior cells correlated with Ca2+ oscillations in the posterior cells and the execution of pBoc, whereas weak oscillations in the anterior cells did not. Furthermore, pBoc was followed by aBoc only when Ca2+ oscillations in the posterior cells propagated in the form of a wave. Similar observations were made regarding the "abortive" anterior oscillations of kqt mutants. These results are consistent with the observation that injection of heparin, an IP3R inhibitor, in int5 prevented Ca2+ wave propagation as well as the latter steps in the DMP (31). However, because anterior-to-posterior waves were not observed in that study, it seems either that they are not strictly required for posterior-to-anterior wave propagation or execution of the DMP or that the model system in use did not facilitate their detection.
Further insight into the role of Ca2+ in coordinating the DMP in free-range worms was gained by investigating the effects of RNAi on the four C. elegans membrane Ca2+ ATPases. Of these four genes (mca-1, mca-2, mca-3, and sca-1), only sca-1, the worm SERCA ortholog, appeared to control Ca2+ dynamics in the intestine and the defecation phenotype. Because sca-1 RNAi resulted in a high degree of lethality, as reported previously (5, 38), we used intestine-restricted RNAi to test its function specifically during Ca2+ oscillations associated with defecation. As expected, knock down of the SERCA ortholog increased the duration of cytoplasmic Ca2+ oscillations. To our surprise, however, the defecation period was only slightly affected by sca-1 RNAi, regardless of how long Ca2+ remained elevated (Fig. 6). Instead, we observed that the timing between pBoc and aBoc was affected. We had noted previously that aBoc appeared to coincide with the return of Ca2+ in the anterior intestinal cells to baseline levels (Fig. 1). In the sca-1(RNAi) worms, the prolonged elevation in Ca2+ resulted in a lengthened interval between pBoc and aBoc (Fig. 6). Furthermore, aBoc always occurred as Ca2+ in the anterior cells returned to normal levels, regardless of the amount of time that it was elevated (Fig. 6). We made similar observations in kqt mutant worms, which also have slightly but significantly increased pBoc-to-aBoc intervals (Table 1) that are coincident with prolonged Ca2+ oscillations (Fig. 3).
We suggest a model where the temporal coupling between successive muscular contractions is determined, at least in part, by how quickly Ca2+ diminishes following oscillations. Because reduction-of-function following sca-1 RNAi was restricted to the intestine, it is unlikely that the alteration in DMP timing is the result of Ca2+ dynamics being misregulated in other tissues. How this relates to the signals that pass between the intestine and the neurons to signal the anterior body wall and enteric muscles to contract is at this time unknown but likely rules out a conventional Ca2+-induced vesicle fusion event.
"Arrhythmic" phenotypes and the case for an underlying pacemaker.
Investigation of the relationships between altered Ca2+ wave dynamics and arrhythmic defecation cycle phenotypes has led us to hypothesize the existence of an underlying pacemaker activity. This hypothesis has been motivated in part by our observation of unexpected regularities in the defecation cycle of arrhythmic mutants. We have found that arrhythmic events occur with a consistent relationship vis-à-vis the normal defecation rhythm in several mutant backgrounds. In both unc-43(sa200) mutants and animals treated with cmd-1 RNAi, we observed shadow cycles that occurred at 
the duration of the normal defecation cycle (Fig. 4). In kqt mutant animals, we observed that abortive anterior Ca2+ oscillations were accompanied by an offset of the DMP equivalent to of the defecation period (Fig. 3). Whether the cycle is extended or abbreviated, arrhythmic events fell in reproducible intervals at a fraction the length of the typical defecation cycle with unexpected frequency. Together, these observations suggest that the defecation phenotypes of these mutants results from uncoupling of the pacemaker from the DMP, rather than disruption of pacemaker activity itself, implying that Ca2+ oscillations may be in fact a response to an underlying, higher-frequency pacemaker activity.
Model of a molecular pacemaker. These observations have also led us to question how Ca2+ regulates the defecation period. It is presumed that there must be an underlying biological activity that unfolds with a time course appropriate to yield the periodicity of the defecation cycle. It is notable that, although intestinal Ca2+ oscillations have been observed to correlate with the defecation cycle, these oscillations are discrete in nature and thus represent at best an incomplete observation of the biological process serving as the pacemaker. Indeed, to date, no continuously varying biological signal has been identified as a candidate pacemaker activity for the defecation cycle.
It is possible that fluctuations in Ca2+ levels below the level of detection of our YC biosensor constitute such a signal. Espelt et al. (7) recently reported that, under specific voltage-clamp conditions, isolated intestinal cells exhibit a continuously oscillating TRP-like current, offering a possible mechanism through which subrosa Ca2+ oscillations may contribute to the pacemaker activity. It is also of note that the itr-1(sa73) mutation that causes an extended defecation period occurs in a putative Ca2+-binding domain and can be compensated for by increasing extracellular Ca2+ (7), consistent with the idea that undetected Ca2+ flux contributes to timing the defecation cycle. However, given that high levels of cytosolic Ca2+ are thought to repress IP3R activity, our sca-1 results suggest that this repression at least is not an essential component of the pacemaker.
In light of our analysis of arrhythmic mutants, we favor a model where a distinct underlying pacemaker oscillation triggers Ca2+ influx when a threshold value is reached, as via temporal summation. Given that the influx of Ca2+ requires activation of the IP3R, it is possible that oscillations in IP3 itself act as the pacemaker. Of course, because low levels of Ca2+ can enhance the activity of the IP3R, it is also possible small rhythmic increases in Ca2+ occur such that both IP3 and Ca2+ oscillations contribute to the pacemaking activity. As noted above, PLC could act as part of a positive feedback mechanism during Ca2+ oscillations (24), and our observations of ectopic wave initiation in egl-8 mutants are consistent with this idea. However, our sca-1 data hint that Ca2+ stimulation of PLC activity may not be primarily responsible for determining the temporal oscillatory pattern in the defecation system. Critical assessment of the validity of this model will require the use of transgenic strains expressing IP3-sensitive biosensors such as LIBRA or IRIS (20, 30).
|
GRANTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
2. Branicky R, Hekimi S. What keeps C. elegans regular: the genetics of defecation. Trends Genet 22: 571–579, 2006.[CrossRef][Web of Science][Medline]
3. Branicky R, Shibata Y, Feng J, Hekimi S. Phenotypic and suppressor analysis of defecation in clk-1 mutants reveals that reaction to changes in temperature is an active process in Caenorhabditis elegans. Genetics 159: 997–1006, 2001.
3a. Brenner S. The genetics of Caenorhabditis elegans. Genetics 77: 71–94, 1974.
4. Charlie NK, Schade MA, Thomure AM, Miller KG. Presynaptic UNC-31 (CAPS) is required to activate the G alpha(s) pathway of the Caenorhabditis elegans synaptic signaling network. Genetics 172: 943–961, 2006.
5. Cho JH, Bandyopadhyay J, Lee J, Park CS, Ahnn J. Two isoforms of sarco/endoplasmic reticulum calcium ATPase (SERCA) are essential in Caenorhabditis elegans. Gene 261: 211–219, 2000.[CrossRef][Web of Science][Medline]
6. Dal Santo P, Logan M, Chisholm A, Jorgensen E. The inositol triphosphate receptor regulates a 50-second behavioral rhythm in C. elegans. Cell 98: 757–767, 1999.[CrossRef][Web of Science][Medline]
7. Espelt M, Estevez A, Yin X, Strange K. Oscillatory Ca2+ signaling in the isolated Caenorhabditis elegans intestine: role of the inositol-1,4,5-trisphosphate receptor and phospholipases C β and . J Gen Physiol 126: 379–392, 2005.
8. Faumont S, Lockery SR. The awake behaving worm: simultaneous imaging of neuronal activity and behavior in intact animals at millimeter scale. J Neurophysiol 95: 1976–1981, 2006.
9. Faumont S, Miller AC, Lockery SR. Chemosensory behavior of semi-restrained Caenorhabditis elegans. J Neurobiol 65: 171–178, 2005.[CrossRef][Web of Science][Medline]
10. Ferrari MB, Spitzer NC. Calcium signaling in the developing Xenopus myotome. Dev Biol 213: 269–282, 1999.[CrossRef][Web of Science][Medline]
11. Gomez TM, Snow DM, Letourneau PC. Characterization of spontaneous calcium transients in nerve growth cones and their effect on growth cone migration. Neuron 14: 1233–1246, 1995.[CrossRef][Web of Science][Medline]
12. Gower NJ, Walker DS, Baylis HA. Inositol 1,4,5-trisphosphate signaling regulates mating behavior in Caenorhabditis elegans males. Mol Biol Cell 16: 3978–3986, 2005.
13. Granato M, Schnabel H, Schnabel R. pha-1, a selectable marker for gene transfer in C. elegans. Nucleic Acids Res 22: 1762–1773, 1994.
14. Hagen FK, Nehrke K. cDNA cloning and expression of a family of UDP-N-acetyl-D-galactosamine:polypeptide N-acetylgalactosaminyltransferase sequence homologs from Caenorhabditis elegans. J Biol Chem 273: 8268–8277, 1998.
15. Heo WD, Inoue T, Park WS, Kim ML, Park BO, Wandless TJ, Meyer T. PI(3,4,5)P3 and PI(4,5)P2 lipids target proteins with polybasic clusters to the plasma membrane. Science 314: 1458–1461, 2006.
16. Jee C, Lee J, Lee J, Lee W, Park B, Yu J, Park E, Kim E, Ahnn J. SHN-1, a Shank homologue in C. elegans, affects defecation rhythm via the inositol-1,4,5-trisphosphate receptor. FEBS Lett 561: 29–36, 2004.[CrossRef][Web of Science][Medline]
17. Lackner M, Nurrish S, Kaplan J. Facilitation of synaptic transmission by EGL-30 G(q)alpha and EGL-8 PLC beta: DAG binding to UNC-13 is required to stimulate acetylcholine release. Neuron 24: 335–346, 1999.[CrossRef][Web of Science][Medline]
18. Liu D, Thomas J. Regulation of a periodic motor program in C. elegans. J Neurosci 14: 1953–1962, 1994.[Abstract]
19. Lorin-Nebel C, Xing J, Yan X, Strange K. CRAC channel activity in C. elegans is mediated by Orai1 and STIM1 homologues and is essential for ovulation and fertility. J Physiol 580: 67–85, 2007.
20. Matsu-ura T, Michikawa T, Inoue T, Miyawaki A, Yoshida M, Mikoshiba K. Cytosolic inositol 1,4,5-trisphosphate dynamics during intracellular calcium oscillations in living cells. J Cell Biol 173: 755–765, 2006.
21. Mello CC, Kramer JM, Stinchcomb D, Ambros V. Efficient gene transfer in C. elegans: extrachromosomal maintenance and integration of transforming sequences. EMBO J 10: 3959–3970, 1991.[Web of Science][Medline]
22. Meyer T, Stryer L. Molecular model for receptor-stimulated calcium spiking. Proc Natl Acad Sci USA 85: 5051–5055, 1988.
23. Miller K, Emerson M, Rand J. G(o)alpha and diacylglycerol kinase negatively regulate the G(q)alpha pathway in C. elegans. Neuron 24: 323–333, 1999.[CrossRef][Web of Science][Medline]
24. Nehrke K, Melvin JE. The NHX family of Na+-H+ exchangers in Caenorhabditis elegans. J Biol Chem 277: 29036–29044, 2002.
25. Norman K, Fazzio R, Mellem J, Espelt M, Strange K, Beckerle M, Maricq A. The Rho/Rac-family guanine nucleotide exchange factor VAV-1 regulates rhythmic behaviors in C. elegans. Cell 123: 119–132, 2005.[CrossRef][Web of Science][Medline]
26. Ozil JP, Swann K. Stimulation of repetitive calcium transients in mouse eggs. J Physiol 483: 331–346, 1995.
27. Peters MA, Teramoto T, White JQ, Iwasaki K, Jorgensen EM. A gap-junction-mediated calcium wave coordinates a rhythmic behavior in C. elegans. Curr Biol doi:10.1016/j. cub.2007.08.031, 2007.
28. Reiner D, Newton E, Tian H, Thomas J. Diverse behavioural defects caused by mutations in Caenorhabditis elegans unc-43 CaM Kinase II. Nature 402: 199–203, 1999.[CrossRef][Medline]
29. Siklos S, Jasper J, Wicks S, Rankin C. Interactions between an endogenous oscillator and response to tap in C. elegans. Psychobiology 28: 571–580, 2000.[Web of Science]
30. Tanimura A, Nezu A, Morita T, Turner RJ, Tojyo Y. Fluorescent biosensor for quantitative real-time measurements of inositol 1,4,5-trisphosphate in single living cells. J Biol Chem 279: 38095–38098, 2004.
31. Teramoto T, Iwasaki K. Intestinal calcium waves coordinate a behavioral motor program in C. elegans. Cell Cal 40: 319–327, 2006.[CrossRef][Web of Science][Medline]
32. Thomas J. Genetic analysis of defecation in Caenorhabditis elegans. Genetics 124: 855–872, 1990.[Abstract]
33. Truong K, Sawano A, Mizuno H, Hama H, Tong KI, Mal TK, Miyawaki A, Ikura M. FRET-based in vivo Ca2+ imaging by a new calmodulin-GFP fusion molecule. Nat Struct Biol 8: 1069–1073, 2001.[CrossRef][Web of Science][Medline]
34. Uhlâen P, Laestadius A, Jahnukainen T, Sèoderblom T, Bèackhed F, Celsi G, Brismar H, Normark S, Aperia A, Richter-Dahlfors A. Alpha-haemolysin of uropathogenic E coli induces Ca2+ oscillations in renal epithelial cells. Nature 405: 694–697, 2000.[CrossRef][Medline]
35. Walker D, Ly S, Lockwood K, Baylis H. A direct interaction between IP3 receptors and myosin II regulates IP3 signaling in C. elegans. Curr Biol 12: 951–956, 2002.[CrossRef][Web of Science][Medline]
36. Wei AD, Butler A, Salkoff LA, Department of Neurobiology. KCNQ-like potassium channels inCaenorhabditis elegans conserved properties and modulation. J Biol Chem 280: 21337–21345, 2005.
37. Yan X, Xing J, Lorin-Nebel C, Estevez AY, Nehrke K, Lamitina T, Strange K. Function of a STIM1 homologue in C. elegans: evidence that store-operated Ca2+ entry is not essential for oscillatory Ca2+ signaling and ER Ca2+ homeostasis. J Gen Physiol 128: 443–459, 2006.
38. Zwaal RR, Van Baelen K, Groenen JT, van Geel A, Rottiers V, Kaletta T, Dode L, Raeymaekers L, Wuytack F, Bogaert T. The sarco-endoplasmic reticulum Ca2+ ATPase is required for development and muscle function in Caenorhabditis elegans. J Biol Chem 276: 43557–43563, 2001.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
) and posterior (
) intestine, with the regions-of-interest corresponding to black and white boxes diagramed in the image series (n = 3). Values taken from a region near the middle of the intestine are plotted as a solid line. The arrows denote the start of posterior body wall (pBoc; white) and anterior body wall (aBoc; black) muscular contractions. D: ratio of emissions (
) from YC6.1 reflects an inversely proportional relationship between single channel emission intensities (480 nm in gray, 535 nm in black), which are characteristic of FRET, and a conformation change that occurs in the biosensor as Ca2+ rises.
