Nervous control of ciliary beating by Cl^-, Ca²⁺ and calmodulin in Tritonia diomedea

The control mechanisms for ciliary beating remain unclear. This is despite the fact that beating cilia have numerous and critical roles in most organisms ranging from locomotion in ciliates(Eckert, 1972; Noguchi et al.,1991; Pernberg and Machemer,1995), marine invertebrate larvae(Mackie et al., 1969; Mackie et al., 1976; Wada et al., 1997) and nudibranchs (Audesirk, 1978; Willows et al., 1997) to the chordate development of left-right symmetry, circulation of cerebral spinal fluid spinal in brain ventricles, clearing of mucus in airways and movement of eggs along fallopian tubes (Ibanez-Tallon et al., 2003; Snell et al.,2004).

Ciliary beating may be under hormonal control(Verdugo, 1980; Korngreen and Priel, 1996; Korngreen et al., 1998; Lieb et al., 2002; Barrera et al., 2004) or nervous control via dopamine and neuropeptides(Willows et al., 1997; Woodward and Willows, 2006),serotonin (Goldberg et al.,1994; Christopher et al.,1996; Willows et al.,1997; Nguyen et al.,2001; Doran et al.,2004), acetylcholine (ACh)(Salathe and Bookman, 1999; Zagoory et al., 2001; Zagoory et al., 2002) or depolarization (Aiello and Guideri,1964; Mackie et al.,1969; Mackie et al.,1976; Murakami and Takahashi,1975). Work permitting any distinction between these control mechanisms is sparse and this may be due, in part, to technical challenges presented by vertebrate models. Other systems such as Tritonia diomedea, in which both ciliated epithelia and neurons are accessible,present an opportunity to study both neural and transduction pathways controlling beating of individual ciliated cells as well as coordinating beating across epithelia.

The intracellular pathways involved in transducing excitatory signals into increased ciliary beat frequency (CBF) have begun to be understood. A key step in increasing CBF in multicellular organisms or reversing directions, in e.g. paramecium (Eckert, 1972), is an increase in internal free Ca²⁺(Verdugo, 1980; Villalon et al., 1989; Salathe and Bookman, 1995; Korngreen and Priel, 1996). Studies of CBF have yielded different possible mechanisms for increasing internal free Ca²⁺, including the phospholipase C (PLC) pathway(Christopher et al., 1999; Zagoory et al., 2001), both protein kinase C (PKC) (Gertsberg et al.,1997; Christopher et al.,1999; Barrera et al.,2004) and protein kinase A (PKA)(Braiman et al., 1998; Zagoory et al., 2002; Lieb et al., 2002), inositol 1,4,5-triphosphate (Barrera et al.,2004), nitrous oxide (Uzlander and Priel, 1999; Runer and Lindberg, 1999; Doran et al., 2003) and cAMP(Stommel and Stephens, 1985; Aiello, 1990). The next step in signal transduction after an increase in internal calcium may be direct binding of Ca²⁺ to the dynein motor(Salathe and Bookman, 1999) or Ca²⁺-calmodulin activation of kinases or phosphodiesterases(Zagoory et al., 2001), or possibly both. Unfortunately, little is known of possible electrical control of these pathways, though most early results were consistent with such a hypothesis (Aiello and Guideri,1964; Mackie et al.,1969; Mackie et al.,1976; Murakami and Takahashi,1975; Saimi et al.,1983a; Saimi et al.,1983b). This report provides evidence that CBF may be under electrical control and describes an intracellular pathway necessary for transduction.

In an accompanying paper (Woodward and Willows, 2006), we showed that ciliated pedal epithelial (CPE)cells of Tritonia diomedea possess calcium dependent Cl^-currents, and the direct suppression of these currents by dopamine or TPep-NLS leads to an increase in CBF; this result is mimicked by the Cl^-channel blocker DIDS. However, the mode of action by which Cl^-current blockage increases CBF is not understood. In the present study we test the hypothesis that CPE cells may possess voltage-gated Ca²⁺channels, and that Ca²⁺ influx through voltage-gated channels is necessary for excitation of CBF. Further, we conclude that Ca²⁺influx may be only a trigger for Ca²⁺ dependent Ca²⁺release from intracellular stores via ryanodine receptor (RyR)channels. Finally, we find that Ca²⁺-calmodulin may be involved in CBF excitation through activation of Ca²⁺-calmodulin dependent kinases and phosphodiesterases. In sum we present a model for nervous system control of ciliary beating frequency, mediated by transmitter or neuropeptide release.

Tritonia diomedea Bergh were trawled from Bellingham Bay, WA, USA,or collected using SCUBA in Puget Sound, WA, USA. They were maintained in 10°C seawater aquaria, and fed sea pens (Ptilosarcus gurneyi)either at Friday Harbor Laboratories (open circulation) or at the University of Washington, Department of Biology (recirculating aquaria).

For experiments, explants of the ciliated pedal epithelium were used. Pedal tissue (2-3 mm square) was dissected from the posterior two thirds of the≈20 cm long foot surface. Removal of small foot sections neither reduces the life span, nor alters eating or copulating behaviors. The pedal tissue was pinned out on a Sylgard™ coated Petri dish and the epithelial layer dissected away from the foot musculature. This produced sheets of tissue consisting primarily of epithelial cells. The sheets were further reduced to 100-500 cells (Fig. 1A). After dissection at room temperature, these explants were cooled to 10°C for>30 min before experimentation. The explants were placed on a large coverslip (24 mm×60 mm), immobilized with a matrix of cotton fibers, and covered by a second smaller coverslip (22 mm×22 mm). Using two thin coverslips rather than glass slides increased the depth of field for the 40× objective and increased the heat transfer from the solution for more efficient cooling. Recovery of explants was indicated by beating cilia and absence of mucus secretion.

CPE cells were isolated from explants following published methods(Pavlova and Bakeeva, 1993):explants were dissected as finely as possible, then reduced to individual ciliated cells by drawing them in and out of a pipette repeatedly. The resulting cells were pipetted into 20 μl of seawater on a large coverslip. Cotton in the solution provided a substrate to which cells could stick, and a second coverslip was placed on top. Thus immobilized, cells and their beating cilia could be visualized and recorded for CBF measurements. Individual cells were considered viable if their cilia were still beating spontaneously.

For the CBF experiments we used filtered seawater (FSW) (Millipore, 0.22μM filter), artificial seawater (ASW), 30 mmol l^-1 K⁺seawater, zero Ca²⁺ seawater, and zero Cl^- seawater. The ASW contained (in mmol l^-1): 400 NaCl, 10 KCl, 10 CaCl₂,50 MgCl₂, 10 Hepes, pH 8.0. 30 mmol l^-1 K⁺seawater contained: 380 NaCl, 30 KCl, 10 CaCl₂, 50 MgCl₂, 10 Hepes, pH 8.0. Zero Ca²⁺ seawater contained:400 NaCl, 10 KCl, 60 MgCl₂, 10 Hepes, 5 mmol l^-1 EGTA,pH 8.0. Zero Cl^- seawater contained: 400 sodium gluconate, 10 potassium gluconate, 10 calcium gluconate, 50 magnesium gluconate, 10 Hepes,pH 8.0.

The total volume of solution on the slide was ≈200 μl and solutions were changed for the CBF experiments using a 100 μl pipette. The solutions were delivered at their final concentrations on one side of the coverslip,then an equal amount was drawn off the other side wicking the solution over the explants or individual cells; a total of 500 μl was perfused for each solution change. Nifedipine (Sigma-Aldrich Corp., St Louis, MO, USA) was made up in ethanol in a 50 mmol l^-1 stock solution, which was diluted in seawater to a final concentration of 50 μmol l^-1 (final solution contained 0.1% ethanol). Xestospongin C (Sigma-Aldrich Corp.), calmidazolium(Calbiochem, San Diego, CA, USA) and dantrolene (Sigma-Aldrich Corp.) were all prepared from a stock solution in DMSO, then diluted to final concentrations in seawater, with a final DMSO concentration always <0.1%. The final concentrations for the inhibitors nifedipine, calmidazolium, W-7, and xestospongin C were chosen from the literature, taking into account the proven maximal effective dose and the possible non-specific effects of the drugs at high concentrations. For none of the treatments did we use a concentration lower than half of that proven maximally effective in molluscs.

Slides containing either explants or isolated cells were viewed with a Nikon inverted microscope scope (Nikon USA) at 40× magnification. The slide rested on a stage cooled to 10-12°C by a Peltier device. Beating cilia were recorded using an Elmo CCD camera #TSN401A (Plainview, NY, USA)scanning at 59.94 Hz. The video was then projected on a monitor. A Photonic Sensor (fiber optic device that measures small changes in light intensity near a small probe) was placed in front of the video projected beating cilia and transduced the CBF into an oscillating voltage(Fig. 1B). The voltage signal was amplified 10× and digitally filtered (bandpass 5-30 Hz) to reduce the raster scan signal from the video monitor. The resulting signal was acquired with a DASH-4U (Astro-Med Inc., West Warwick, RI, USA) digital oscilloscope, and a Fast Fourier Transform analysis used to recover the dominant frequency of voltage oscillations(Fig. 1C). We validated these measurements by using a QuickTime (Apple Computer Inc., Cupertino, CA, USA)video of an artificial `cilium' beating at known rates (2-16 Hz). CBF was sampled just after solution changes and then at 3 min intervals. For blocking experiments with dantrolene, xestospongin C and calmidazolium, explants were incubated in the blocking agent for 60, 30 or 30 min, respectively, prior to experimentation. All experiments ended with washout of the applied drug or drugs, and only one experiment was performed on tissue (in the form of a single explant, or single isolated ciliated cell) from each animal.

All graphs were generated in Sigma Plot 2000 6.10 software (SPSS Inc.,Chicago, IL, USA) or in Microsoft Excel software (Microsoft Corporation,Redmon, WA, USA) and all statistical analyses were made using Microsoft Excel software. For most experiments comparisons were made between the mean basal rate of CBF in either FSW or the incubating inhibitor and the maximal change in CBF after treatment, unless otherwise noted. This comparison was performed using a paired Student's t-test. Populations were compared using a two-tailed Student's t-test. Means are reported ± s.e.m., with N values noted.

In an accompanying paper (Woodward and Willows, 2006) we present evidence for voltage-gated Ca²⁺ channels in CPE cells, and hypothesize that the blockage of Cl^- channels by dopamine and TPep-NLS leads to depolarization and increases in CBF. Influx of extracellular Ca²⁺ through voltage-gated Ca²⁺ channels has also been proposed as a method for increasing [Ca²⁺]_in and CBF(Christopher et al., 1996). In addition, a voltage-gated Ca²⁺ channel could link Ca²⁺entry and membrane voltage, thereby providing a mechanism for electrical control of CBF. To test whether depolarization and Ca²⁺ influx through voltage-gated channels is necessary for CBF excitation, we applied 30 mmol l^-1 K⁺ seawater to CPE explants to depolarize the cells. Seawater with 30 mmol l^-1 K⁺ increased CBF 134.7±25.0% above basal levels (N=5; P≤0.006)(Fig. 2A). After 3 min in 30 mmol l^-1 K⁺ the excitation declined to 92.7±30.9%above basal levels (N=5; P≤0.04), and after 6 min of 30 mmol l^-1 K⁺ exposure the CBF declined to 52.0±27.8% above basal levels (N=5), no longer significantly different from seawater controls (P≤0.13).

We repeated the 30 mmol l^-1 K⁺ experiments, but in a zero Ca²⁺ seawater bath, to learn if the depolarization-induced increase in CBF is dependent on the influx of extracellular Ca²⁺(Fig. 2A). Initial exposure to 30 mmol l^-1 K⁺ in the zero Ca²⁺ bath only increased CBF 5.3±11.4% (N=6), a significant decrease from 30 mmol l^-1 K⁺ alone (P≤0.0007). Interestingly,the zero Ca²⁺ seawater alone produced a marginal increase in CBF compared to controls, though only significantly different at one of three time points (viz. after 6 min, P≤0.02). Finally, we used the specific L-type Ca²⁺ channel blocker, nifedipine, to identify the Ca²⁺ channels in CPE cells. 50 μmol l^-1 nifedipine initially caused an increase in CBF (N=5; P≤0.02);however, this maximal CBF excitation is reduced compared to excitation caused by 30 mmol l^-1 K⁺ seawater alone (N=5; P≤0.006) (Fig. 2A). To confirm that the 30 mmol l^-1 K⁺ seawater acts directly, we tested individual isolated CPE cells. Under these circumstances,30 mmol l^-1 K⁺ seawater caused an 83.2±17.3%increase (N=5; P≤0.009) in CBF above basal levels(Fig. 2B). These results confirm that depolarization is acting directly on ciliated cells to cause CBF excitation and that depolarization mimics the excitation mediated by dopamine and TPep-NLS.

Though Ca²⁺ influx may be necessary for depolarization-induced increases in CBF, it may not be required for other CBF excitatory signals. To confirm the role of Ca²⁺ influx for all excitation we used nifedipine to block Ca²⁺ in the presence of a number of known CBF stimulants. Zero external Cl^- bath solution increased CBF 117.1±17.4%, a significant increase(Fig. 3A) (N=5; P≤0.001) that mimics the increases seen with Cl^-channel blockers (Woodward and Willows,2006). However, the excitatory effect of a Cl^--free bath solution is markedly reduced in the presence of 50 μmol l^-1nifedipine (N=5; P≤0.0009)(Fig. 3A), so much so that the residual excitation is indistinguishable from controls(P≤0.10).

Blocking Ca²⁺ influx also reduces excitation by dopamine and TPep-NLS. Nifedipine significantly reduced dopamine (N=5; P≤0.003) and TPep-NLS (N=6; P≤0.01) induced CBF excitation (Fig. 3B). Like nifedipine, removal of external Ca²⁺ also significantly reduced the excitatory effects of TPep-NLS (N=5; P≤0.00009) and dopamine (N=5; P≤0.008)(Fig. 3B). Unlike nifedipine,and especially in the case of dopamine, the effects of zero external Ca²⁺ were highly variable. In some instances, zero Ca²⁺had little or no effect, while in others it resulted in a complete abolition of excitation, yielding a large standard error (±2.54 Hz, corresponding to ±35.6% change from control).

Increases in [Ca²⁺]_in are well established as a prerequisite for CBF excitation (Verdugo,1980; Villalon et al.,1989; Salathe and Bookman,1995; Korngreen and Priel,1996), but the source of the Ca²⁺, e.g. from stores or influx, remains unclear. Having already established that Ca²⁺influx occurs during CBF stimulation we also wanted to explore the possible contribution of endoplasmic reticulum (ER) stores to[Ca²⁺]_in. We used caffeine, known at high concentrations to activate selectively the ryanodine receptor (RyR) channel located in the ER membrane and control Ca²⁺ release from stores. Caffeine applied at 1 mmol l^-1 (Fig. 4A)increased CBF 150.3±18.9% above basal levels (N=5; P≤0.001); however, the excitatory effect waned over time,decreasing to a 64.5±13.1% change by 6 min. However repetition of the experiment with 10 mmol l^-1 caffeine(Fig. 4A) also produced CBF excitation (215.5±24.1% increase above basal levels, N=6; P≤0.0003) and these effects did not diminish after 6 min. We chose therefore to use 10 mmol l^-1 caffeine in all subsequent experiments. We also found that 50 μmol l^-1 nifedipine had little or no effect (N=5; P≤0.16) on the action of caffeine, a result consistent with the hypothesis that internal sources of Ca²⁺ alone are sufficient to produce CBF excitation.

To confirm the specificity of caffeine action on RyR channels we used the cell permeable RyR channel blocker, dantrolene(Blackwell and Alkon, 1999; Kawai et al., 2004). The presence of dantrolene significantly reduced excitation caused by 10 mmol l^-1 caffeine (N=7; P≤0.00004)(Fig. 4B), confirming that caffeine is working specifically to produce Ca²⁺ release via RyR channels. Dantrolene also significantly reduced excitation induced by dopamine (N=6; P≤0.0002) and TPep-NLS(N=6; P≤0.00009) (Fig. 4B), further supporting the hypothesis that Ca²⁺release from stores is required for CBF excitation in addition to Ca²⁺ influx.

A second Ca²⁺ release channel is common in the ER membrane, an inositol 1,4,5-triphosphate (IP₃) mediated Ca²⁺ release channel. We investigated the possible involvement of IP₃ mediated Ca²⁺ release with the potent membrane permeable IP₃mediated Ca²⁺ release blocker, xestospongin C, that has proven effective in marine organisms (Yazaki et al., 2004). The inclusion of xestospongin C with FSW slightly lowered the basal beating rate (P≤0.01)(Fig. 5A), but did not prevent a large increase in CBF with the addition of dopamine. Dopamine's excitatory effect with or without xestospongin C did not diminish over time. In fact, a comparison of the percent increase in CBF induced by dopamine(198.9±16.2%; N=5) (Fig. 5B) revealed that xestospongin C (223.3±12.5%; N=5) had no affect on excitation (P≤0.27). Similarly TPep-NLS induced increase (126.2±9.2%; N=10) was unaffected by the presence of xestospongin C (158.9±11.9%; N=5; P≤0.06) (Fig. 5B). These results suggest the IP₃ pathway is not utilized for dopamine or TPep-NLS induced CBF excitation.

Finally, we wanted to investigate the pathways involved downstream of the[Ca²⁺]_in increase. There is no consensus as to the relationship between second messenger Ca²⁺ and the increases in dynein motor activity and subsequent global increases in CBF. It was proposed that Ca²⁺ might act directly on the axoneme in epithelial cells(Salathe and Bookman, 1999);however, numerous other intermediaries remain possibilities. Calmodulin is one option. Calmodulin is known, once bound with Ca²⁺, to interact directly with the axoneme (Stommel and Stephens, 1985), and also to activate Ca²⁺-calmodulin kinases that can phosphorylate targets in the axoneme including beta tubulin(Fabczak et al., 1999). We studied the potential role of calmodulin in the excitatory response to neurotransmitters and neuropeptides by using a number of calmodulin antagonists that work by inhibiting Ca²⁺-calmodulin dependent kinases and phosphodiesterases, viz. W-7 and calmidazolium. To begin,we tested the calmodulin antagonists on the excitation caused by caffeine.

Caffeine presumably works upstream of calmodulin and so should be affected by an antagonist. W-7 (N=5; P≤0.003) and calmidazolium(N=4; P≤0.006) both significantly reduce the caffeine-induced excitation (Fig. 6). This result is consistent with the placement of calmodulin activity after Ca²⁺ release from stores, not during the Ca²⁺ influx, though it could be active at both times. Further, both W-7 (N=5; P≤0.007) and calmidazolium (N=5; P≤0.001) reduced dopamine-induced excitation as well(Fig. 6). W-7 also proved effective at reducing TPep-NLS excitation of CBF (N=5; P≤0.0001), however, calmidazolium had no effect (N=3; P≤0.17). The failure of calmidazolium to produce an effect is the only experimental result that suggests a difference between the dopamine and TPep-NLS excitatory pathways. The reduction in CBF caused by W-7 and calmidazolium suggest both the involvement of calmodulin in the CBF control pathway, but also that Ca²⁺ acts via one of its phosphodiesterase or kinase targets, not directly on the axoneme.

Our results show that dopamine and TPep-NLS act by modulating Ca²⁺ influx and Ca²⁺ release from RyR gated stores, and by upregulating cilia motor proteins via the Ca²⁺-calmodulin pathway; thereby increasing CBF. The link between blockage of Cl^- currents(Woodward and Willows, 2006)and Ca²⁺ influx is critical. We hypothesize that blockage of Cl^- is linked to depolarization. Depolarization of cells in explants or isolated CPE cells mimics the increase in CBF caused by dopamine and TPep-NLS. Our findings differ from previous observations in other adult molluscs (Murakami and Takahashi,1975; Saimi et al.,1983a; Saimi et al.,1983b), larval molluscs(Mackie et al., 1969; Mackie et al., 1976) and other marine invertebrates (Mackie et al.,1974). However, single cell organisms (e.g. Parameciumand Didinium nasutum) respond to depolarization by reversing cilia stroke direction and increasing beat frequency(Eckert, 1972; Pernberg and Machemer, 1995). We were unable to record changes in CPE membrane potentials using the whole cell patch clamp in current clamp due to the inherently leaky nature of the cells. We therefore relied on high K⁺ extracellular seawater to induce a transient depolarization of CPE cells.

We found that depolarization induced excitation is dependent on Ca²⁺ influx. Both zero external Ca²⁺ and nifedipine block depolarization induced CBF excitation, similar to other observations reported for serotinin (5-HT) induced CBF excitation in Helisomaembryos (Christopher et al.,1996). Our results are consistent with the hypothesis that CPE cells possess a small number of voltage-gated Ca²⁺ channels. This possibility separates molluscan CPE cells from the well-studied ciliated epithelium of vertebrate airways, which show no evidence of excitability, but nonetheless produce elevated plateaus of internal Ca²⁺(Braiman et al., 2000). What appears to be similar is the necessity of Ca²⁺ influx for sustained CBF excitation. We found that excitation by dopamine, TPep-NLS, and high K⁺, are blocked by nifedipine and zero external Ca²⁺,indicating that Ca²⁺ influx is necessary for any CBF excitation. Though the excitable molluscan CPE cells are not similar to vertebrate epithelial cells possessing motile cilia, there are numerous examples of epithelial cells with non motile cilia that possess voltage-gated channels. The pigmented and non-pigmented ciliary epithelial cells of the vertebrate retina possess voltage-gated Ca²⁺ channels, which may aid in the spread of excitability across the retina(Farahbakhsh et al., 1994),and may occur also in kidney epithelial cells(Nauli et al., 2003).

Finally, we also showed a direct link between Cl^- flux and Ca²⁺ influx. Removing external Cl^- excites CBF; however,the excitation is blocked by nifedipine. This provides an alternative to the hypothesis that simple activation of Ca²⁺ channels by transmitters causes Ca²⁺ influx in CPE cells. In addition, it establishes a novel involvement of Cl^- channels in the regulation of motile ciliary beating control.

Ryanodine receptor (RyR) channel-gated internal Ca²⁺ stores contribute significantly to [Ca²⁺]_in increases and CBF excitation. Vertebrate ciliary control also depends on Ca²⁺ stores,often through IP₃-gated stores, not RyR(Braiman et al., 2000). Interestingly, RyR channel-gated stores have not been implicated previously in control of CBF. Our results do not exclude the possibility that both IP₃- and RyR-gated stores are involved in CBF control. Xestospongin C has been used successfully to block IP₃ activated functions widely in animals (Gafni et al.,1997; Barrera et al.,2004; Yazaki et al.,2004), although in our experiments we lacked a positive control for xestospongin C. We did find that xestospongin C depressed basal CBF rates,raising the possibility that the IP₃ pathway contributes to both basal and excited CBF, but not to dopamine or TPep-NLS induced excitation.

Our findings also suggest that Ca²⁺-calmodulin is in the CBF control pathway. We place its role downstream of Ca²⁺ release from internal stores because W-7 and calmidazolium attenuate the observed caffeine excitation. Calmodulin has been implicated in many of the described CBF control pathways (Stommel and Stephens,1985; Zagoory et al.,2001; Zagoory et al.,2002). Some researchers attribute W-7 and calmidazolium reduction of CBF excitation to non-specific chelating effects, instead believing that Ca²⁺ directly binds to ciliary beating mechanism(Salathe and Bookman, 1999). Others support the idea that Ca²⁺-calmodulin dependent kinases and phosphodiesterases influence dynein behavior(Zagoory et al., 2002). While our methods do not differentiate between these two possibilities, our findings do represent yet another CBF control system influenced by Ca²⁺-calmodulin. One result worthy of further investigation is the finding that while W-7 does inhibit TPep-NLS excitation, calmidazolium does not. One possibility is that a higher concentration of calmidazolium would be more effective. However we chose not to use the dose proven to have maximal effect on molluscan neurons (10 μmol l^-1)(Onozuka and Watanabe, 1996)because at higher concentrations calmidazolium may produce non-specific effects (Sugita et al., 1999),rendering results at the higher concentrations questionable. It is also possible that TPep-NLS and dopamine are working through multiple mechanisms,each with a different response to calmidazolium.

In conclusion, our results demonstrate how dopamine and the neuropeptide TPep-NLS control CBF excitation in CPE cells(Fig. 7). Excitatory signals bind receptors that lead directly to a blockage of Ca²⁺ dependent Cl^- currents, currents that also contribute to the resting leak current (Woodward and Willows,2006). Blockage of an inward Cl^- current contributing to V_rest may cause a depolarization of the cell membrane. Fluctuations in membrane potential trigger Ca²⁺ influx through Ca²⁺ channels sensitive to nifedipine. A small Ca²⁺influx in turn triggers Ca²⁺ release from RyR-gated stores. The rise of [Ca²⁺]_in activates calmodulin and it's dependent kinases and phosphodiesterases capable of interacting with the cilia motor proteins to increase CBF. We do not rule out multiple CBF excitation pathways. In most cases significant decreases in CBF excitation through pharmacological treatments did not abolish excitation altogether. This may represent either limited efficacy of the pharmacological agents, or alternatively, parallel pathways controlling Ca²⁺ influx/release and CBF. Further support for multiple pathways includes our observations that excitatory responses to dopamine and TPep-NLS are twofold. There is the easily quantifiable change in CBF, but also a noticeable induced coordination of all the cilia from a single explant. This induced coordination occurs in conjunction with even minor increases in CBF and is not inhibited by the calmodulin inhibitors. Could there be different pathways controlling coordination and CBF increase? We also provide evidence that IP₃ may not be a significant player in CBF excitation, but that does not rule out a PLC dependent pathway using PKC, as found in Helisoma embryos(Christopher et al., 1999; Doran et al., 2004), nor the adenylate cyclase/cAMP/PKA pathway found to be necessary in Mytilus(Stommel and Stephens, 1985; Aiello, 1990) (preliminary findings suggest that cAMP plays no role in CPE cells; O. M. Woodwood,unpublished observation), nor other non-calmodulin dependent Ca²⁺actions (Salathe and Bookman,1999). Electrically induced excitation of CPE cells in Tritonia diomedea is consistent with their role as primary locomotory effectors, and like muscle, their RyR-gated stores and voltage-gated Ca²⁺ channels may be an adaptation permitting parallel CNS control.

List of abbreviations
 
• ACh

  acetylcholine

 
• ASW

  artificial seawater

 
• CBF

  ciliary beat frequency

 
• CNS

  central nervous system

 
• CPE

  ciliated pedal epithelial cells

 
• DA

  dopamine

 
• ER

  endoplasmic reticulum

 
• FSW

  filtered seawater

 
• IP₃

  inositol 1,4,5-triphosphate

 
• PKA, PKC

  protein kinase A/C

 
• PLC

  phospholipase C

 
• RyR

  ryanodine receptor

 
• TPep-NLS

  Tritonia Pedal Peptide

This work was funded by NIH Fellowship 5 F31 NS047922-02 (O.M.W.). We appreciate numerous discussions with Drs R. Wyeth and S. Cain, and thank Drs. W. Moody and M. Bosma for reading earlier versions of the manuscript. A final thanks goes to Tatia Chay Woodward for all other possible types of support.