INTRODUCTION

Top Abstract Introduction Procedures Results Discussion References

There is abundant evidence that the phosphoinositide cascade is involved in invertebrate phototransduction; however, the specific roles of inositol trisphosphate (IP₃)¹ and/or Ca²⁺ in visual excitation have not yet been established (1-3). The involvement of phosphatidylinositol 4,5-bisphosphate (PIP₂) hydrolysis by phospholipase C (PLC) in the generation of IP₃ in photoreceptors of Limulus (4, 5), squid (6-8), Drosophila (9-12), and Hermissenda (13) is supported by several lines of biochemical and biophysical evidence. In particular, the photoresponse in Drosophila was found to be absolutely dependent on PLC activity. Until recently, it was thought that IP₃ is the second messenger involved in the photoresponse in invertebrates because a brief injection of IP₃ mimicked the quantal response to light (4, 5) and IP₃-induced Ca²⁺ release resulted in membrane depolarization in the Limulus ventral photoreceptor (14) by activating Ca²⁺/Na⁺ exchange (15, 16).

More recent evidence, however, indicates that the mechanism is more complex. For example, many contradictions arise if Ca²⁺/Na⁺ exchange is assumed to be the only mechanism involved in phototransduction: 1) the light response precedes the rise in Ca²⁺ (1), 2) excitation occurs in the absence of an IP₃-induced increase in intracellular Ca²⁺ in Limulus (17), 3) abrupt Ca²⁺ elevation using caged Ca²⁺ fails to activate any channels in Drosophila photoreceptor membrane (18), and 4) the Drosophila mutant deficient in IP₃ receptors still produces the native photoresponse (19). Furthermore, cGMP induces an inward current in photoreceptor cells in Limulus (20) and Drosophila (21) and activates channels directly in excised patches of molluscan microvilli membrane (1, 22). It is still uncertain, however, whether intracellular cGMP concentration is altered by light (23-25). Adding to the complexity of the system is the possibility of second messenger cross-talk between IP₃, Ca²⁺, and cyclic nucleotides to activate intracellular chemical cascades (26).

Hermissenda provide an ideal animal model in which to examine invertebrate visual transduction mechanisms. The photoreceptors are relatively large (50 µm in diameter) and are well characterized electrophysiologically (27-35). They possess two types of photoreceptors: A and B (28, 29). Photoreceptor B responds to a bright flash with a complex potential change involving an initial depolarization, a hyperpolarization, and a depolarizing tail, corresponding to a rapid and transient Na⁺ conductance increase, a slower increase in K⁺ permeability, and a delayed decrease in K⁺ permeability, respectively (32). The initial Na⁺ current appears to be the primary phototransduction event because 1) following light adaptation, light flashes produce only the hyperpolarization component (32) and 2) the hyperpolarization is Ca²⁺-dependent (33) and Hermissenda photoreceptors respond to single photons even in TTX containing, low Ca²⁺ high Mg²⁺ solution (34). Our previous work also indicates that the photoresponse may be generated through a TTX-resistant Na⁺ channel (36).

The present study focuses on investigating the second messengers underlying the initial Na⁺ current component of the photoresponse. Recently, channels directly gated by IP₃ have been described in other cell lines (37, 38), including Na⁺ channels in rat megakaryocytes (39). This raises the possibility that the native photoresponse in Hermissenda is generated by an IP₃-gated Na⁺ channel. To further characterize the role of IP₃ and the possible role of cGMP, cAMP, and Ca²⁺ in invertebrate phototransduction, the effects of injection of these second messengers were examined electrophysiologically in Type B photoreceptors of Hermissenda.

	EXPERIMENTAL PROCEDURES

Top Abstract Introduction Procedures Results Discussion References

Electrophysiological Assay-- Hermissenda crassicornis obtained from the Sea Life Supply (Sand City, CA) were maintained in artificial sea water (ASW) at 14 °C. Type B photoreceptors were dissected in ASW (430 mM NaCl, 10 mM KCl, 50 mM MgCl₂, 10 mM CaCl₂, and 10 mM Tris-HCl, pH 7.4). Dissection and intracellular recording were performed as described previously (13). cGMP, cAMP, and IP₃ (Sigma) were dissolved in intracellular solution (67 mM sodium acetate, 400 mM potassium gluconate, 10 mM MgCl₂, 10 mM HEPES, 1.5 mM CaCl₂, 1.5 mM EGTA, pH 7.4) at a concentration of 1 mM. Isobutylmethylxanthine (IBMX) was dissolved in Me₂SO and diluted with intracellular solution to 1 mM. The intracellular solution alone or with Me₂SO had no effect on the photoresponse. A microelectrode containing the drug was inserted into the photoreceptor cell, and the drug was injected using a pressure injector (Medical System, Greenvale, NY) at 30.0 p.s.i. for 200 ms. In some experiments the drug was iontophoresed into the photoreceptor for at least 3 min with DC current of 2.5 nA, after 10 min of dark adaptation. A photoresponse was elicited every 90 s during the entire experimental period to keep the adaptation level constant. Unattenuated light of 34.0 µW/cm² at 550 nm delivered from a halogen lamp source was administered to the underside of the preparation.

Assay of PIP₂ Breakdown and PLC Activity-- Isolated Hermissenda eyes were incubated in 10 µl of ASW containing 10 µCi of [³H]inositol (Amersham International) in darkness at 23 °C for 16 h. Then one eye of each pair was exposed to light for 30 s. The incubation was stopped by adding chloroform/methanol/12 N HCl (200:100:0.75), and PIP₂ standard was added as a carrier and to aid in the identification of the lipids. Phospholipids were extracted and separated by thin layer chromatography as described previously (9). Lipids were visualized with I₂ vapor, and appropriate bands were scraped and counted. PIP₂-PLC activity was assayed using the particulate fraction obtained from the Hermissenda circumesophageal nervous system as described previously (13).

	RESULTS

Top Abstract Introduction Procedures Results Discussion References

Comparison between the Photoresponse and the IP₃-induced Response-- The IP₃-evoked response was highly dependent on the distance between the injection site and the rhabdomeric membrane, the light-sensitive region (Fig. 1a). Immediately after IP₃ was injected into a microvilli of the Type B photoreceptor close to the rhabdomeric membrane, a large transient response was observed (Fig. 1a). No response was observed after injection approximately 10 µm away from the photosensitive region. The latency of the photoresponse decreased with increasing light intensity (Fig. 1b). With low light intensity, the rate of rise and the peak amplitude of the photoresponse were low. With increased light intensity, a more complex response was observed, consisting of an initial depolarization, a hyperpolarization, and a depolarizing tail, consistent with previous reports (32). The first component was because of a rapid and transient Na⁺ conductance increase, the hyperpolarization was because of a slow increase in K⁺ permeability, and the depolarizing tail was because of a slow decrease in K⁺ permeability (33). The Na⁺ current was a graded response, consistent with previous observations in Hermissenda (36) and photoreceptors of other species (40, 41). The spontaneous firing activity superimposed on the photoresponse is known to arise from the distal portion of the axon and is not involved in the visual transduction process (31). Many characteristics of the response to IP₃ injection were similar to the response to a relatively weak light flash, including the latency and the generation of action potentials (Fig. 1c). Simultaneous light flash and IP₃ injection were additive (Fig. 1d). Furthermore, both responses were completely inhibited upon removal of extracellular Na⁺ (Fig. 1d). Similar to the native photoresponse, the IP₃-evoked response was not dependent on the extracellular Ca²⁺ concentration (data not shown). Successive injection of IP₃ resulted in a broadening and a decrease in the amplitude of the first peak and a higher frequency of spontaneous firing. This response mimicked the photoresponse to flickering light stimulation (Fig. 1e). Flickering light stimulation elicited an increase in the Na⁺ conductance and a decrease in the hyperpolarization.

View larger version (27K): [in this window] [in a new window]   Fig. 1.   Light- and IP3-induced responses from the Type B photoreceptor of Hermissenda. a, schematic drawing of an eye composed of five photoreceptors with the positions of the electrodes. The rhabdomeric membranes are indicated by the black portions. The lightly shaded portion indicates the Type B photoreceptor. IP3 injections were applied via electrodes placed in microvilli of the Type B photoreceptor at position 1, close to the rhabdomeric membrane, and position 2, approximately 10 µm away. The traces are electrical recordings of the response of the photoreceptor to IP3 injection. b, photoresponses to increasing intensities of light (unattenuated 200-ms light flash of 34 µW/cm2 at 550 nm) from 6 log to 1 log units. c, responses of the same cell to a light flash and an IP3 injection. The intensity of light was adjusted to obtain a response amplitude comparable with that of the response to the IP3 injection. The duration of both the light flash and the IP3 injection was 200 ms; the former was controlled by an electromagnetic shutter, and the latter was controlled by a pressure injector with a monitor of the pressure at the end of the electrode. d, the IP3-induced response (dotted line) and the response to light were additive. The additive response (continuous line) was obtained by a simultaneous IP3 injection and light flash, of which the intensity was attenuated to produce a small photoresponse. Note that the IP3-induced response was abolished in Na+-free ASW. e, the response to repetitive injection of IP3 mimicked the response to flickering light (bottom panel).

Effects of Ca²⁺ on the Photoresponse-- To examine whether a transient increase in intracellular Ca²⁺ activated the visual transduction channel, Ca²⁺ (1 mM) was injected into the photosensitive region of the Type B photoreceptor. In contrast to results obtained in Limulus (14), injection of Ca²⁺ alone did not induce any significant response (Fig. 2a), but when a light flash was applied to the photoreceptor preinjected with Ca²⁺, the amplitude of the initial peak of the photoresponse and the hyperpolarization were potentiated (Fig. 2b). This response was similar to the response to an intense light flash shown in Fig. lb.

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Ca2+ facilitated the native photoresponse. a, Ca2+ injection alone (1 mM for 200 ms) did not produce a significant response. b, the light-induced response amplitude increased 1.5 times compared with control, and a significant after-hyperpolarization (arrow) appeared after the injection of Ca2+. The resting membrane level is indicated by a dotted line. c, the effect of 1 mM BAPTA on the photoresponse. Photoresponse before BAPTA injection (top panel), 3 min later (middle panel), and 6 min later (bottom panel). The initial depolarizing photoresponse gradually diminished and then reversed to a hyperpolarization.

View larger version (26K): [in this window] [in a new window]   Fig. 3.   Modulation of the native photoresponse to aminoglycosides, polyamines, and IBMX. a and b, responses were suppressed by the injection of neomycin (a, 1 mM) and spermine (b, 1 mM). c, IBMX reduced the amplitude of the initial depolarization of the photoresponse. The reduced response completely recovered 540 s after IBMX injection. d, IBMX reduced the light-induced inward current measured using two microelectrode voltage clamps (36) at a holding potential of 60 mV. The numbers on the left represent the time in seconds after the injection. The bottom panel shows the recovery of the response at 18, 10, 9, and 9 min from a to d, respectively.

Effects of Cyclic Nucleotides on the Photoresponse-- Although a significant change in the concentration of cyclic nucleotides in invertebrate photoreceptors during light irradiation has not been observed (25, 43), we tested the possibility that these cyclic nucleotides are involved in visual transduction (44). Intracellular pressure injection of 1 mM cAMP generated a train of action potentials without a significant change in resting membrane potential (Fig. 4a). Consistent bursts of action potentials were observed in response to successive injections of cAMP (Fig. 4b). Successive bursts were not significantly different from the first burst, indicating that the effects of the injection were transient, unlike the effects of IP₃, Ca²⁺, or BAPTA injection. The responses to cAMP injections were abolished in Na⁺-free ASW, similar to the response to IP₃ injection (Fig. 4b). The response to cAMP in Hermissenda also differed from that in Limulus (44), suggesting further that the phototransduction mechanism in Hermissenda is different from the ventral eye in Limulus.

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Responses induced by 1 mM cAMP injection. a, comparison between the light- and cAMP-induced response. cAMP injection evoked a train of action potentials without a change in resting membrane potential. b, repetitive injection of cAMP-induced responses that did not change significantly with successive injections.

View larger version (9K): [in this window] [in a new window]   Fig. 5.   cGMP-induced responses. a, cGMP injection evoked a transient depolarization without any generation of action potentials. The numbers on the left represent the distance in microns that the injection electrode was advanced from the penetration site, position 2 of Fig. 1. b, comparison of the responses to cGMP injection and light flash. The latency of the response to cGMP was much shorter than that of the response to light.

View larger version (15K): [in this window] [in a new window]   Fig. 6.   Effects of injected cGMP and extracellular Na+ and Ca2+ concentration on Hermissenda photoresponses. a, injection of cGMP reduced the photoresponse in cells maintained in normal ASW containing 10 mM Ca2+. b, the inhibitory effect of cGMP was abolished in low Ca2+ (1 mM) ASW. c, the cGMP-evoked response was unchanged in Na+-free ASW. d, inhibition of the cGMP-induced response in Ca2+-free ASW (10 mM CaCl2 was replaced with equimolar CdCl2). e, the photoresponse was reduced in Na+-free ASW. f, the photoresponse was unchanged in Ca2+-free ASW.

	DISCUSSION

Top Abstract Introduction Procedures Results Discussion References

The results of the present study demonstrate for the first time that IP₃ may mediate invertebrate visual transduction, particularly in Hermissenda. Although cAMP, cGMP, and Ca²⁺ have been proposed to activate several types of channels, injection of these second messenger candidates did not reproduce the native photoresponse. In contrast, injections of IP₃ produced responses that were very similar to the native photoresponse, whereas Ca²⁺ acted as a modulator of the photoresponse (46).

The hypothesis that IP₃ may be the second messenger involved in visual transduction in Hermissenda is supported by the results of the present study and by results obtained previously: 1) light flashes result in greater increases in intracellular IP₃ than cyclic nucleotides (25) and 2) inhibition of PLC activity and reduction of phosphoinositide turnover result in a reduction in the amplitude of the native photoresponse (13). Although the primary role of IP₃ has been considered to be the release of Ca²⁺ from intracellular stores, several IP₃-gated ion channels have been recently characterized in olfactory cilia (37, 38) and rat megakaryocytes (39), raising the possibility that IP₃ directly activates a transduction channel in Hermissenda.

The possibility that IP₃ may be a second messenger in Drosophila visual transduction is supported by observations by Inoue et al. (10) that a reduction of PIP₂-PLC activity is closely correlated with electrophysiological abnormality in the Drosophila phototransduction mutant norpA (no receptor potential A) eye. Furthermore, the norpA gene has been shown to encode PLC (11, 47). Yoshioka et al. (43) have also reported previously that IBMX inhibits phosphoinositide turnover and that the photoresponse is dependent on PIP₂ hydrolysis by PLC in squid photoreceptor. The photoresponse in Limulus has also been shown to be diminished by the injection of the PLC inhibitor neomycin without affecting the response to IP₃ (14) and by polyamine injection (17). Therefore, we propose that IP₃ produced from PIP₂ hydrolysis via PLC activation is an essential component of invertebrate visual transduction.

The question remains about whether IP₃ acts directly or via Ca²⁺ in Hermissenda visual transduction. In the present study, injection of Ca²⁺ into the photoreceptor in Hermissenda induced no response in darkness, suggesting that Ca²⁺ is not directly involved. Furthermore, excitation occurs even in the absence of an IP₃-induced increase in intracellular Ca²⁺ in Limulus (17) and in Drosophila mutants lacking IP₃ receptors in photoreceptor cells (19). It has also been reported that the photoresponse precedes the rise of intracellular Ca²⁺ in Limulus (48, 49). Recently, Hardie (18) reported that a transient elevation of Ca²⁺ using a caged Ca²⁺ compound did not directly excite Drosophila photoreceptors. These results indicate that Ca²⁺ is unlikely to be directly involved in phototransduction in Drosophila, Hermissenda, and Limulus but may act as a modulator of the photoresponse.

Several previous studies indicate that Ca²⁺ excites photoreceptors (50-52) and that the injection of a Ca²⁺ buffer reduced the photoresponse in Limulus (50). Also, in the present study, reduction of intracellular Ca²⁺ by a Ca²⁺ chelator resulted in a reduction of the photoresponse; however, the mechanism underlying this response is presently unclear. It may be because of reduced activation of PLC and a subsequent decrease in IP₃; however, a decrease in protein kinase C activity cannot be ruled out. According to Thio and Sontheimer (53), phorbol esters reduced the peak of the TTX-sensitive Na⁺ current by 25-60% and potentiated the TTX-resistant Na⁺ current by 60-150%. In the present study, the Ca²⁺-induced increase in the photoresponse was 60-80% of control, which is consistent with the phorbol ester-induced increase in the TTX-resistant Na⁺ current.

Involvement of cyclic nucleotides in visual transduction is also unlikely. The cAMP-evoked response was very different from the light-evoked response; injection of cAMP induced short bursts of Na⁺-dependent action potentials. Repeated injection evoked trains of similar bursts, indicating that these action potentials may be mediated by fast-inactivating cAMP-sensitive Na⁺ channels.

Interestingly, cGMP-evoked responses were similar to light-evoked responses, but the following differences were observed: the amplitude of the cGMP-evoked responses depended on the distance from the rhabdomeric membrane, but the latency of the response was constant for all trials; the rising phase of the responses to cGMP was faster than that of responses to light; the cGMP-evoked response depended on external Ca²⁺, whereas the light-evoked response depended on external Na⁺; and the response induced by long term cGMP injection and the light-evoked response from light-adapted cells were very similar. Taken together, these findings indicate that neither cAMP nor cGMP act as second messengers in invertebrate visual transduction. Thus the results of the present study suggest that the transduction channels of Drosophila and Hermissenda eye may be IP₃-gated Na⁺ channels. Cloning of IP₃-gated Na⁺ channels may help clarify the mechanisms underlying invertebrate visual transduction.