A Neuronal Role for A Crustacean Red Pigment Concentrating Hormone-Like Peptide: Neuromodulation of the Pyloric Rhythm in the Crab, Cancer Borealis

Many neuropeptides were first identified because of their presence in peripheral tissues and organs and their action upon them, and subsequently were found in the nervous system where they are thought to function as transmitters and modulators (Krieger, Brownstein & Martin, 1983). The hormones that control pigment translocation in crustaceans have been extensively studied for most of this century (Perkins & Snook, 1931; Josefsson, 1983; Rao, 1985). In particular, red pigment concentrating hormone (RPCH) was identified originally by its ability to concentrate red pigment granules within the erythrophores of many different crustaceans, including the crabs Uca pugilator, Cardisoma carnifex and Ocypode pallidula (Cooke, Haylett & Weatherby, 1977; Josefsson, 1983; Rao, 1985). More recently, RPCH has been implicated in the concentration of pigments in the crustacean ommatidia (Kulkarni & Fingerman, 1986).

The amino acid sequence of RPCH (pGlu-Leu-Asn-Phe-Ser-Pro-Gly-TrpNH₂) was determined originally from the shrimp Pandulus borealis (Fernlund & Josefsson, 1972). An apparently identical RPCH has been identified in the shrimp Leander adspersus (Carlsen, Christensen & Josefsson, 1976) and in the crab Cardisoma carnifex (Newcomb, 1983). The RPCH-like peptide from the crab Carcinus maenas appears similar but not identical to the shrimp peptide (Manger-ich, Keller & Dircksen, 1986). Interestingly, the sequence of RPCH is quite similar to those of the adipokinetic hormones (AKHs) in insects, although RPCH and the AKHs are thought to serve different physiological functions. In addition to RPCH, the RPCH/AKH peptide family includes at least seven structurally distinct peptides (Jaffe et al. 1986; Schaffer, 1986; Gade & Rinehart, 1987). Two of these (AKH IIS and AKH IIL) differ from RPCH by only a single amino acid. Like RPCH and the AKH Ils, MI (CCI) and MII (CC2) are also octapeptides. The other three identified members of this peptide family are slightly longer: M-AKH is a nonapeptide while AKH I and HTF II are decapeptides. The adipokinetic hormones are known to increase lipid mobilization in insect muscle during flight (Stone, Mordue, Batley & Morris, 1976), enhance hyperglycaemic activity and effect cardioacceleration in insects (Baumann & Gersch, 1982; Scarborough et al. 1984) and increase myoactivity in insect skeletal muscle (O’Shea, Witten & Schaffer, 1984). An RPCH/AKH-like peptide with cardioexcitatory activity has also been isolated from the clam Mercenaria mercenaria (Greenberg et al. 1985).

In crustaceans, a major source of hormonally released RPCH is the sinus gland, a neurohaemal organ in the crustacean eyestalk (Cooke et al. 1977 ; Cooke & Haylett, 1984; Rao, 1985). Several regions of the crustacean central nervous system also contain RPCH, or an RPCH-like peptide, including the brain, thoracic ganglia and the commissural ganglia (Rao, 1985; Mangerich et al. 1986). Recently, the distributions of RPCH-like immunoreactivity (RPLI) and AKH-like immunoreactivity have been studied in several crustacean species (Madsen, Herman & Elde, 1985; Mangerich et al. 1986; Schooneveld, van Herp & van Minnen, 1987). These studies show that the central nervous systems of crustaceans contain a peptide identical, or similar, to RPCH as well as an AKH-like peptide (Madsen et al. 1985; Mangerich et al. 1986; Schooneveld et al. 1987), and prompted the suggestion that an RPCH-like peptide also functions as a neurotransmitter or neuromodulator (Mangerich et al. 1986; Schooneveld et al. 1987). We now provide physiological evidence that this is the case.

The crustacean stomatogastric ganglion (STG) contains only 30 neurones, and produces two different motor patterns that control the movements of the animal’s stomach (Selverston & Moulins, 1987). Previous work has shown that a large number of amines and peptides are present in inputs to the STG (Marder, 1984, 1987) and that these substances can influence the motor pattens produced by the STG (Marder, 1984, 1987; Marder, Hooper & Eisen, 1987b; Harris-Warrick & Flamm, 1986). Using an antiserum against RPCH (Madsen et al. 1985) we find RPLI in input fibres to the STG and in fine processes within the STG neuropile. Moreover, we find that exogenously applied RPCH has strong excitatory effects on the pyloric rhythm produced by the STG. Thus RPCH or a similar peptide is likely to be a neurotransmitter or modulator in the stomatogastric nervous system. It appears, therefore, that RPCH can play both a systemic hormonal role in the control of pigment translocation and a neurotransmitter/modulator role in the nervous system. Some of these data have been presented in abstract form (Nusbaum & Marder, 1986).

Cancer borealis were obtained from local (Boston) commercial fishermen, and held in artificial seawater aquaria until used. 95 male animals weighing between 100 and 500g were used (68 for immunocytochemistry; 27 for electrophysiology).

C. borealis physiological saline was as reported by Marder, Hooper & Siwicki (1986). All peptide-containing solutions were made directly before use at the desired concentrations from stock solutions that were kept frozen in small samples. Peptides were obtained from Peninsula or Bachem. Several of the stock solutions were prepared by first dissolving the peptide in 5-8% dimethylsulphoxide (DMSO). Control experiments showed that the trace amounts of DMSO applied to the physiological preparations had no effect on activity.

Whole-mount immunocytochemistry on the stomatogastric nervous system was performed by the method of Beltz & Kravitz (1983) as described previously (Marder et al. 1986; Marder, Calabrese, Nusbaum & Trimmer, 1987a). The primary antiserum, a rabbit anti-RPCH antibody raised against RPCH coupled to thyroglobulin, was a gift from the Elde laboratory (Madsen et al. 1985). Incubations in the primary antiserum were made at 1:200 in triton-azide-containing buffer (Marder et al. 1986), containing 10% non-immune goat serum, for 24h at 4°C. The secondary antiserum was Cappell rhodamine-labelled goat anti-rabbit used at 1:25. Preincubation blocks were performed by incubating the primary antibody with the peptide for 2h at room temperature before applying the mixture to the preparation. Lucifer Yellow backfills were carried out as described by Marder et al. (1987a). Preparations were viewed using a Zeiss IM35 fluorescence microscope equipped with rhodamine and fluorescein filters. Colchicine-treated animals were injected with 15 mg ml⁻¹ colchicine in physiological saline. Animals were sacrificed 2-4 days later, and tissues were processed as usual.

Physiological experiments used routine methods for the STG. Extracellular recordings were made with Vaseline-insulated stainless steel pin electrodes. Intra-cellular recordings were made with 20—50 MΩ microelectrodes filled with potassium acetate. All experiments were carried out using a continuously flowing superfusion system (6-10 ml min⁻¹) cooled to 12—14°C that allowed rapid changing of the bath solution. Peptides were introduced into the bath without interrupting the flow of the solution. Bath volume was 15—20ml.

The 30 somata of the crab STG are about 50-100μm in diameter and encircle a flat, central neuropile in which synaptic contacts are made. When the STG was processed by whole-mount immunocytochemistry (Beltz & Kravitz, 1983) using antibody raised against RPCH (Madsen et al. 1985), RPLI was observed exclusively in fine fibres and swellings throughout the neuropile (Fig. 1A). No somata in the STG were stained, nor was any RPLI seen in the motor nerves of the STG. Several fibres were seen in the stomatogastric nerve (STN), the nerve that brings inputs to the STG from the rest of the nervous system.

The specificity of the RPLI stain was studied using the ability of the antibody to label STGs subsequent to preincubation with different peptides. Preincubation of the antiserum with 10⁻⁶moll⁻¹ RPCH abolished staining (Fig. 1B), as did preincubation with 10⁻⁶mol 1⁻¹ AKH IIL (Fig. 1C) and 10⁻⁶mol 1⁻¹ AKH IIS (not shown). In contrast, a much higher concentration of AKH I (10⁻⁴ mol 1⁻¹) (Fig. 1D) was relatively ineffective in blocking the staining seen with antiserum to RPCH. The AKH peptides M I and Mil blocked staining at 10⁻³ mol 1⁻¹. RPLI staining was not blocked by preincubation with thyroglobulin (Imgml⁻¹), to which RPCH was coupled in generating the primary antibody. RPLI was also not blocked by preincubation with proctolin, FMRFamide or substance P (10⁻⁴mol 1⁻¹). These peptides were chosen because previous work has shown that these peptides or immunologically related ones are also present in the neuropile of the STG (Marder et al. 1986, 1987a ; Goldberg, Nusbaum & Marder, 1986).

Some of the fibres that project into the STG come from neurones with their somata either in the single oesophageal ganglion (OG) or the paired commissural ganglia (CGs) (Selverston & Moulins, 1987). Therefore, we determined the distribution of RPLI in the OG and CGs, and then attempted to determine the location of the source of the RPLI projections into the STG neuropile. For these latter experiments we combined Lucifer Yellow backfills of individual nerves with subsequent immunolabelling of the same tissue using the antiserum against RPCH, visualized with a rhodamine-conjugated secondary antibody (see Marder et al. 1987a).

Two of the approximately 14 OG neurones showed RPLI. The Lucifer Yellow backfill-RPLI double-label technique showed that the two RPLI neurones in the OG send axons towards the brain via the inferior ventricular nerve (IVN), rather than into the STN towards the STG (Fig. 2A,B). Interestingly, these same two neurones also showed FMRFamide-like immunoreactivity (Marder et al. V)81a’, Callaway, Masinovsky & Graubard, 1987) and SCPb-like immunoreactivity (Callaway et al. 1987).

The CGs contain about 400 neurones each. The CGs showed extensive neuropilar staining, and 5-10 somata showed RPLI (Fig. 2C). These somata were visualized most effectively after colchicine treatment, which precluded combining Lucifer Yellow backfills with immunocytochemistry for these cells. Thus, although these neurones are good candidates for the source of the RPLI in the STG neuropile, we have been unable to demonstrate this directly. The commissures connect the brain to the suboesophageal ganglion. RPLI was found in individual fibres within the commissures (Fig. 2D). Some of these fibres were seen to ramify within the CGs, which are outpocketings of the commissures. The distribution of RPLI throughout the complete stomatogastric nervous system and the position of the more anterior ganglia are shown in Fig. 2E.

The pyloric rhythm of the stomatogastric ganglion consists of a repeating, alternating pattern of dilation and constriction of the pyloric region of the stomach (Selverston & Moulins, 1987). This rhythm can be easily recorded with extracellular electrodes on the motor nerves, as can be seen in the first trace of Fig. 3A. This is a recording from the lateral ventricular nerve (LVN), and shows the sequential activity of the lateral pyloric (LP) neurone (largest unit in the recording), pyloric (PY) neurone (smallest unit in the recording) and pyloric dilator (PD) neurone (medium-sized units in the recording).

When the STG is isolated from the OG and CGs some preparations continue to produce pyloric rhythms, while others become quiescent. Fig. 3 illustrates the effects of bath application of increasing concentrations of RPCH on a robust pyloric rhythm in an isolated STG. In this case the pyloric cycle frequency was 0·87 ±0·02 Hz (mean ± S.D.) in the control condition. RPCH applications produced a dose-dependent increase in the frequency of the pyloric rhythm, to 1·21 ±0·02Hz in 10⁻⁶moll⁻¹ RPCH (these differences were statistically significant at the P<0·001 level by two-tailed t-test). Additionally, increasing RPCH concentrations produced increases in the number of LP neurone action potentials per burst and extended the percentage of each cycle taken up by the LP neurone burst. In the example shown in Fig. 3 the LP neurone fired 7·8 ± 0·4 action potentials per burst in control saline and 16·8 ± 0·4 action potentials per burst in 10⁻⁶mol 1⁻¹ RPCH. Again, these differences were significant at the P< 0·001 level.

RPCH activated rhythmic pyloric cycling when it was applied to quiescent preparations, as shown in recordings from a PD neurone and the LP neurone (Fig. 4). In control saline the PD neurone was silent, whereas the LP neurone fired tonically. After application of 10⁻⁷moll⁻¹ RPCH, rhythmic activity commenced, resulting in repeating alternations of bursts of action potentials in the PD and LP neurones.

The threshold for the activation of quiescent preparations by RPCH was around 5xl0⁻⁹moll⁻¹. Increases in the concentration of RPCH from threshold to 10⁻⁶mol 1⁻¹ decreased the latency to onset and increased the frequency of the pyloric rhythm (Fig. 5).

The excitatory effects of RPCH on pyloric frequency are summarized in Fig. 6, a plot of the pyloric rhythms recorded in 16 preparations in various RPCH concentrations. The mean control pyloric frequency was 0·25 ±0·36 Hz (mean±S.D., N =42), and the pyloric frequency in 10⁻⁶moll⁻¹ RPCH was 1·07 ± 0·23 Hz (N = 22). The mean number of LP spikes per burst increased from 3·19 ±4·1 in saline to 15·8 ±3·8 in l0⁻⁶molL¹ RPCH (P<0·001). RPCH increased the number of spikes per burst in another component of the pyloric network, the IC neurone, and caused a slight but consistent decrease in the number of PD spikes per burst.

In many preparations RPCH and the AKHs mimic the physiological actions of each other. Therefore, we compared the physiological effects of the AKHs with those of RPCH in six experiments. In the representative recordings shown in Fig. 7, the preparation was virtually silent before peptide application, and returned to quiescence after each wash. RPCH and the AKHs, at 10⁻⁷moll⁻¹, were considerably more potent than were M I and M II. In the same preparation as that shown in Fig. 7, RPCH was the only peptide to evoke a regular repeating pyloric rhythm when applied at 10⁻⁸moll⁻¹.

There are a number of substances that can initiate rhythmic pyloric activity in the STG, including dopamine, the muscarinic agonist pilocarpine, serotonin, octo-pamine, proctolin and an FMRFamide-like peptide (Raper, 1979; Anderson, 1980; Marder & Paupardin-Tritsch, 1978; Marder & Eisen, 1984; Flamm & Harris-Warrick, 1986a, b; Hooper & Marder, 1984, 1987). Several of these, including dopamine and pilocarpine (Raper, 1979; Anderson, 1980; Harris-Warrick & Flamm, 1986), can produce and maintain rhythmic non-impulse-mediated alternations between antagonists in the presence of tetrodotoxin (TTX).

We have found in three experiments that the ability of RPCH to evoke rhythmic alternating patterns of activity is also maintained in the presence of TTX. In the experiment shown in Fig. 8, the preparation was silent after the application of 10⁻⁶moll⁻¹ TTX. Soon after 10⁻⁶moll⁻¹ RPCH had been applied to the bath, rhythmic alternations of membrane potential ensued in the LP and PD neurones.

The frequency of this rhythmic activity, however, was considerably slower than the pyloric frequency that occurs in similar concentrations of RPCH in normal saline. The alternating LP and PD activity persisted as long as RPCH remained in the bath.

We have shown that RPCH and the AKHs can produce dramatic physiological actions on the stomatogastric ganglion. These data, together with the finding that RPCH or a closely related peptide is present in input fibres to the STG, provide strong evidence that RPCH is likely to have transmitter/modulator functions in the crustacean nervous system, in addition to its long-known hormonal roles. Thus, RPCH joins a growing list of compounds, such as serotonin, octopamine, dopamine and proctolin, that are thought to be released both into the general circulation from neurohaemal organs in arthropods and to be neurally released into ganglionic neuropiles (Beltz & Kravitz, 1986).

RPCH also joins a growing list of neuromodulatory inputs to the STG that modify the pyloric rhythm (Marder, 1987). Despite the large number of known modulatory inputs to the STG, it appears that each of these produces different effects on the motor patterns of the STG (Marder, 1984, 1987; Marder et al.1987b). In this context it is interesting to compare the effects of RPCH with those of the previously described modulatory inputs to the STG. RPCH is most similar in its actions to proctolin (Hooper & Marder, 1984; Marder et al. 1986). RPCH and proctolin can both initiate rhythmic activity in quiescent preparations. Additionally, both sub-stances produce increases in the number of LP neurone spikes per burst. However, in C. borealis proctolin also increases the number of PD neurone spikes per burst (M. P. Nusbaum & E. Marder, unpublished results), while RPCH produces a slight decrease in the number of PD neurone spikes per burst. Thus, again, each additional neuromodulatory substance appears to evoke a slightly different variation of the motor patterns produced by the STG.

It is premature to speculate on the nature of the synaptic contacts made by the RPLI-containing terminals in the neuropile of the STG. It is possible that the RPLI fibres make defined synaptic contacts on a subset of the STG neurones. Alterna-tively, the RPLI-containing terminals may release peptide generally into the neuropile, as a local neurohormone (Jan & Jan, 1982; Sigvardt, Rothman, Brown & Mayeri, 1986). The source of the RPLI-containing fibres that project into the STG is also not yet clear. Possibly these fibres descend from CG somata. It is also possible that they come from somata in the brain or thoracic ganglia via some of the RPLI fibres seen in the commissures..

RPLI is co-localized in two OG somata that also contain an FMRFamide-like peptide (Marder el al. 1987a) and an SCPb-like peptide (Callaway et al. 1987). These neurones might contain a large peptide that is recognized by several antisera, or they might synthesize and release several different neurally active peptides. This issue must wait for resolution until the protein precursors for these peptides are isolated and purified.

The results of our preabsorption controls suggest that the antiserum of Madsen et al. (1985), raised against RPCH, recognizes the amidated tryptophan (-TrpNH2) at the-COOH terminus of RPCH. Of the RPCH/AKH peptides that were tested, the one that was least able to preabsorb the antiserum against RPCH was AKH I. AKH I was also the only tested member of this group that has an amidated threonine instead of an amidated tryptophan at the-COOH terminus (see Fig. 7). At the-NH₂ terminus, the sequences of the first four amino acids are identical in RPCH and AKH I. The fact that M I and M II are 10 times less effective than RPCH and the AKH IIs in preabsorption controls suggests further that the tested antiserum prefers-Gly-TrpNH₂ to-Asn-TrpNH₂.

For several reasons, our data are consistent with the RPLI in the stomatogastric nervous system being either ‘true’ RPCH or a very closely related molecule, such as AKH IIS or AKH IIL. First, RPCH, AKH IIS and AKH IIL appear to be similarly effective in activating the pyloric rhythm. Second, although we obtained staining with anti-RPCH antiserum, we saw no AKH-like staining with several different anti-AKH antibodies (M. P. Nusbaum & E. Marder, unpublished results). Third, the peptides that resemble RPCH most closely were most effective in blocking staining. AKH I, while being a potent physiological agonist, was ineffective in blocking RPLI. Fourth, although RPCH has not yet been isolated and sequenced from Cancer borealis, it has been characterized biochemically from several different crustaceans (Fernlund & Josefsson, 1972; Carlsen et al. 1976; Newcomb, 1983; Mangerich et al. 1986), and in each case the peptides are closely related. Fifth, extracts of Cancer borealis nervous systems showed RPCH-like bioactivity (K. R. Rao, personal communication). Sixth, injection of purified RPCH produced concentration of the red pigments in Cancer borealis, suggesting that the RPCH-like hormonal activity in C. borealis is produced by a peptide similar in structure to authentic RPCH (M. P. Nusbaum & E. Marder, unpublished results).

The sinus gland appears to be a major source of hormonally released RPCH. In the crayfish, the RPCH release from the sinus gland is circadian, and is maximal at night (Arechiga, Cortes, Garcia & Rodriguez-Sosa, 1985). For crabs with calcified exoskeletons, the role of RPCH in the control of body pigment is not obvious. However, RPCH also causes pigment concentration in the ommatidia, thus enhancing visual sensitivity (Kulkarni & Fingerman, 1986). This enhanced visual sensitivity might increase foraging success, insofar as many crustacean species are nocturnally active and forage at night. If RPCH is released into the crab haemolymph at high enough concentrations to affect the STG (which is located within the dorsal aorta), then hormonally released RPCH might act to coordinate the animal’s feeding behaviour. Alternatively, haemolymph RPCH levels may remain below the threshold for effects on the STG, but may act as a primer for neurally released RPCH by lowering the threshold for neural action, as has been suggested for the effects of serotonin on swimming in the leech (Willard, 1981; Nusbaum & Kristan, 1986).

We thank the Elde laboratory for the kind supply of the RPCH antibody. We thank Mr Jorge Golowasch for reading an earlier version of this manuscript and Ms Carolyn Sherff for technical assistance. Mr Michael O’Neil gave us invaluable photographic assistance. Research supported by NRSA NS-07446 to MPN and NS-17813.