Ecdysis of Decapod Crustaceans is Associated With a Dramatic Release of Crustacean Cardioactive Peptide into the Haemolymph

Moulting is a prerequisite for development and growth in crustaceans and other arthropods. Among the different moult stages, the stage of ecdysis comprising the actual exuviation process is the shortest stage, but the most crucial step in the life cycle of any of these invertebrates, because death is inevitable if it cannot be completed successfully. The ecdysis of a freshwater crayfish and some aspects of the stereotyped behaviour involved were described in detail at the beginning of the eighteenth century (Réaumur, 1712, 1718a,b) and later refined to some extent for crustaceans including crayfish (Huxley, 1880; Aiken, 1968; Aiken and Waddy, 1987), crabs (Vitzou, 1882; Drach, 1939; Guyselman, 1953; Passano, 1960), prawns (Jefferies, 1964), lobsters (Travis, 1954; Aiken, 1980; Waddy et al., 1995) and isopods, the latter showing a complex biphasic ecdysis (Carlisle, 1956). Most authors have divided the procedures leading to the shedding of the old exoskeleton into a passive phase, mainly characterised by significant water uptake, followed by an active phase of exuviation behaviour. Because of the short duration of this process, only one author has so far tried to subdivide this important stage of the crustacean moult cycle in the crayfish Orconectes virilis but merely on the basis of moult mortalities (Aiken, 1968).

Several reports have dealt with the physiological events that accompany crustacean ecdysis. In crabs and lobsters, ecdysis involves profound changes in the rate of absorption of several ions, especially Mg²⁺ and Ca²⁺ (Robertson, 1960; Wheatly, 1985), connected with the rapid water uptake initiated in late premoult stage D₄ (Baumberger and Olmsted, 1928; Dandrifosse, 1966; Dall and Smith, 1978; Mykles, 1980; Neufeld and Cameron, 1994). A transient increase in hydrostatic haemolymph pressure (deFur et al., 1985), a decrease in the rate of oxygen consumption and an increase in the rate of carbon dioxide and lactate production also occur at this time (Lewis and Haefner, 1976; Mangum et al., 1985b; Mangum, 1992; Clemens et al., 1999). There is only one report of the electrical activities of the heart, stomach and skeletal muscles in a spiny lobster (Kuramoto, 1993). However, surprisingly little is known about the mechanisms controlling the initiation and maintenance of the crustacean ecdysis process and the stereotyped behaviour patterns involved.

The induction of moult and a successful ecdysis in crayfish has been suggested to be dependent upon photophase length, upon the balanced concentrations of a moulting hormone (ecdysteroids) and a moult-inhibiting hormone (MIH) (Aiken, 1969) and upon water temperature (Westin and Gydemo, 1986). A currently accepted model of moult control in crustaceans involves the inhibitory neuroendocrine regulation of Y-organ ecdysteroid biosynthesis in intermoult by MIHs from the X-organ/sinus gland (XOSG) neurosecretory system of the eyestalk. These neuropeptides are members of a large peptide family also including crustacean hyperglycaemic hormones (CHHs) and gonad/vitellogenesis-inhibiting hormones (GIHs/VIHs) (for reviews, see Keller, 1992; Chang et al., 1993; Webster, 1998). However, almost nothing is known about the endocrine regulatory principles governing the ecdysis process. The first preliminary evidence for an involvement of the neuropeptide crustacean cardioactive peptide (CCAP) in crustacean ecdysis was reported by Johnen et al. (1995), who demonstrated that CCAP levels in the ventral nervous system of the isopod Oniscus asellus are elevated shortly before and during ecdysis. Furthermore, recent investigations have demonstrated a large transient increase in levels of a CHH in the haemolymph of shore crabs, Carcinus maenas, associated with ecdysis. This CHH is identical to that known from the XOSG system, but is released from hitherto unknown gut paraneurons only shortly before, during and after ecdysis and obviously drives water uptake during this process (Chung et al., 1999).

Compared with crustaceans, ecdysis has been investigated for more than 30 years in insects, resulting in a much deeper understanding of the endocrine regulatory principles. These involve an eclosion hormone (EH), ecdysis-triggering hormone (ETH) and crustacean cardioactive peptide (CCAP), studied principally in the sphingiid moth Manduca sexta and its developmental stages (for reviews, see Hesterlee and Morton, 1996; Truman et al., 1998). A central role for CCAP in initiating ecdysis motor programmes within the central nervous system under the influence of ETH and EH has recently been demonstrated (Gammie and Truman, 1997a,b, 1999). Although in M. sexta, during all developmental stages, some CCAP neurons in the ventral nerve cord project into peripheral neurohaemal release sites (Davis et al., 1993), almost nothing is known about the putative neurohormonal roles of CCAP and/or other cardioactive peptides (CAPs) during ecdysis with the exception of a proposed role for CAPs in heartbeat control during wing-spreading behaviour (Tublitz and Truman, 1985b).

In this paper, we describe the natural ecdysis process of two decapod crustacean species, the shore crab Carcinus maenas and the crayfish Orconectes limosus, in greater detail than before and introduce a novel nomenclature for subdivisions of the stage of ecdysis (stage E; Drach, 1939; Drach and Tchernigovtzeff, 1967) in the decapod moult cycle on the basis of morphological and behavioural criteria applied to living specimens. This was necessary since a previously proposed substaging of E (Aiken, 1968) could not be considered as a useful basis for our studies because it merely reflects results obtained artificially after manipulations during premoult. As has been discussed previously (Rao et al., 1973), these stages therefore represent the results of unnatural changes leading to distinct forms of death in stage E obviously caused by certain developmental deficiencies attained during premoult, rather than describing any of the natural processes during ecdysis in living animals. Furthermore, we show using a novel and evaluated assay system that crustacean ecdysis is invariably associated with the release of large quantities of CCAP into the haemolymph during ecdysis and that this release can be set into the behavioural context of active ecdysis behaviour. We show that CCAP is released from the crayfish ventral nerve cord, which is known to contain large areas of perineural release sites for CCAP. These are distributed along the entire dorsal perineural sheath of the ventral nerve cord, including the segmental nerve roots, and originate from identified serially homologous CCAP-immunoreactive neurosecretory neurons (Trube et al., 1994).

Mature adults of medium-sized (carapace width 5–6 cm) shore crabs, Carcinus maenas L., of all stages, especially those in the premoult stages (so-called ‘peeler’ crabs), were collected in July and August (1997 and 1998) from the shore off the Isle of Anglesey, UK. The animals were maintained in individual culture tanks connected to a recirculating seawater system at 12–15 °C under a light:dark photoperiod of 16 h:8 h.

Adult specimens of the crayfish Orconectes limosus Rafinesque, 1871 were caught from the river Havel in Berlin and obtained from a local fisherman. Animals were maintained under running tap water in large aquaria at a water temperature of 12–15 °C and were fed pelleted cat food twice a week. These stock aquaria received a light:dark photoperiod of 12 h:12 h. Mature premoult crayfish with a rostrum–telson length of 7–9 cm were transferred into individual culture tanks under a light:dark photoperiod of 16 h:8 h at room temperature (18–22 °C).

The moult stages of premoult crabs were determined by investigation of the setal development of the basipodite of the second maxilla, by morphological criteria, such as carapace and epimeral suture rigidities, and by behavioural variables. Moult staging of premoult crayfish was carried out using a setal index according to Aiken and Waddy (1987). Some further morphological variables are introduced in the present study (see below). In general, the moult stages referred to the staging system first introduced by Drach (1939) and further modified for different crustaceans by several authors (for a review, see Aiken and Waddy, 1987). In the present report, a substaging system of stage E has been defined by the use of morphological and especially behavioural variables (see Results) and following the three criteria of Drach (1939) for the usefulness of a such staging system; i.e. permitting rapid stage recognition of the living animal, applicability to animals of different sizes and to different species of decapod crustaceans, and accounting for transformations without placing undue emphasis upon ecdysial staging based upon time intervals.

Crab ecdysis was documented on colour slide film. In the case of crayfish ecdysis, single video frames were taken with a Kappa no. CF 15/2 video camera connected to a Panasonic NV-HS1000 video recorder. The images were digitised using a Miro Video DC 30 frame grabber card in a personal computer and analysed using image-processing software (Adobe Premiere 4.2, Corel Photopaint 7.0).

Haemolymph was purified by a method developed to give the highest recoveries and the most consistent results. Samples (200–300 μl) of haemolymph were taken from the ventral abdominal sinus (Orconectes limosus) or the hypobranchial sinus (Carcinus maenas) and snap-frozen in liquid nitrogen. Prior to purification, samples were gently thawed and centrifuged (2500 g, 5 min, 4 °C). A defined volume of the supernatant was then slowly (approximately 1 ml min⁻¹) applied to a Sep-Pak C₁₈ cartridge (Waters) preconditioned with 10 ml of 60 % acetonitrile and equilibrated with 10 ml of water, both at a flow rate of 10 ml min⁻¹. After loading of the sample, the cartridge was washed with 10 ml of water followed by elution of the peptide fraction with 3 ml of 60 % acetonitrile, again at a slow flow rate (1 ml min⁻¹); the first 0.5 ml (void volume) was discarded. The eluate was collected into a Minisorb tube (NUNC Gibco), dried in a vacuum centrifuge (Speed Vac, Savant) and redissolved in 125–200 μl of assay buffer (PBST) containing (in mmol l⁻¹): NaCl, 138; Na₂HPO₄, 11.25; KCl, 2.68; KH₂PO₄, 1.76; Tween 20, 0.1 % v/v; pH 7.4).

The recovery of CCAP after Sep-Pak purification was determined by adding an 125I-radiolabelled synthetic CCAP analogue PYCNAFC-NH₂ (Y2-CCAP; see below) to the haemolymph samples. Since the hormone contents of the haemolymph varied over a wide range at different moult stages, every sample was assayed in a dilution series to find the best fit within the linear range of the assay.

A highly sensitive competitive enzyme immunoassay (EIA) has been developed for CCAP by modifying a principle described previously (Kingan, 1989). Synthetic CCAP was coupled to human serum albumin (HSA, Sigma) using 1-ethyl-3,3′-diaminopropyl-carbodiimide (EDC, Sigma) at a molar ratio of CCAP:HSA:EDC of 50:1:560 according to Skowsky and Fisher (1972). This CCAP–HSA conjugate was diluted 1:1500 in phosphate buffer (PB, 0.1 mol l⁻¹, pH 8.0) and coated overnight (4 °C, humid atmosphere) onto 96-well microtitre plates (NUNC, Maxisorp F16). After three washing steps in PB (0.01 mol l⁻¹, pH 8.0), the plates were blocked with normal goat serum (NGS; 1 % in PB, 0.1 mol l⁻¹, pH 8.0) for 2 h at room temperature. The blocking buffer was poured out, and the plates were either used immediately or stored for up to 2 weeks at 4 °C. A standard curve was prepared in assay buffer (PBST) in the range 0–400 fmol of CCAP per well (50 μl). Thereafter, 50 μl per well of a primary antiserum against CCAP (code 1TB; Dircksen and Keller, 1988) diluted in PBST was added to standards and samples at a final antiserum concentration of 1:3000, and the plates were incubated overnight at 4 °C in a sealed humid chamber. After three washes in PBST, 100 μl per well of an enzyme-labelled secondary antibody (sheep anti-rabbit IgG–peroxidase conjugate, Boehringer) was added at a concentration of 1:3500 diluted in PBST, and the plates were incubated for 1 h at room temperature in a humid chamber. Finally, after three washes in PBST, 100 μl of the substrate solution containing 0.2 μl ml⁻¹ 30 % hydrogen peroxide and 0.4 mg ml⁻¹ 2,2′-azino-bis-3-ethyl-benzthiazoline-6-sulphonic acid (Sigma) in citrate buffer, pH 4.0, was added, and the mixture was incubated for 1 h at room temperature. The optical density of the reaction products was then measured with an EIA reader (MK II Multiscan Plus, Flow Laboratories) at 405 nm. The concentrations of standards and samples were evaluated using a connected computing system (software version EIA3, Flow Laboratories). For statistical evaluations, a computerised program was used (Microsoft Excel version 5.0).

The presence of CCAP in the haemolymph and in extracts of the pericardial organs of crabs in different moult stages, dissected under ice-chilled crab saline (Webster, 1986), was identified by HPLC analysis and subsequent CCAP–EIA analysis of every fraction. Haemolymph collected from intermoult animals (stage C₄, 20 ml) was directly compared with haemolymph from animals that were in the active part of ecdysis (stage E₃₀–E₁₀₀, 20 ml). The samples were collected from the hypobranchial sinus and immediately frozen in liquid nitrogen. Prior to HPLC purification, the haemolymph was gently thawed and centrifuged for 5 min at 2700 g (at 4 °C) to remove remaining cells and debris. The supernatant was prepurified using Sep-Pak cartridges (2 ml per cartridge) preconditioned with 10 ml of isopropanol (100 %) followed by equilibration with 10 ml of water and eluted with 3 ml of 40 % isopropanol. The eluates were pooled and dried in a vacuum centrifuge, and the resulting pellet was redissolved in 200 μl of 2 mol l⁻¹ acetic acid prior to chromatography under the following conditions. A linear gradient of 30 % to 80 % solvent B [solvent A: 0.11 % trifluoroacetic acid (TFA); solvent B: 0.1 % TFA in 60 % acetonitrile] over 60 min was used to elute a phenyl column (Waters μBondapak, 4.6 mm×250 mm) at a flow rate of 1 ml min⁻¹. Fractions of 1 ml were collected automatically, dried, redissolved in PBST and assayed for CCAP immunoreactivity in a dilution series. A standard of 10 pmol of synthetic CCAP in 2 mol l⁻¹ acetic acid was chromatographed immediately after the sample under the same conditions.

Haemolymph from crayfish that had just completed ecdysis (stage E₁₀₀) was compared with haemolymph from intermoult animals (stage C₄) using gel filtration, and every fraction was assayed for CCAP immunoreactivity by CCAP–EIA. Haemolymph was collected from the ventral abdominal sinus and immediately frozen in liquid nitrogen. Prior to gel filtration, the sample was thawed, centrifuged (2000 g, 5 min, 4 °C) and the supernatant diluted 2:1 with an anticoagulant buffer (PBS containing 1 % glycine ethyl ester). The diluted haemolymph was chromatographed at 4 °C on a 90 cm×0.6 cm column packed with Sephadex G-50 superfine (Pharmacia) using 0.1 mol l⁻¹ PBS, pH 7.4, as eluant. Fractions of 400 μl were collected automatically at intervals of 15 min, and their optical density was measured at 220 nm. Subsequently, every fraction was assayed in duplicate for CCAP immunoreactivity.

Small quantities (approximately 0.25 nmol) of synthetic Y2-CCAP were iodinated with 9.25 MBq of Na¹²⁵I (Amersham) using chloramine-T (Bolton, 1989). The purification of the radiolabelled peptide was carried out with a Sep-Pak cartridge previously conditioned with isopropanol and equilibrated with water. The labelled peptide was eluted with 3 ml of 40 % isopropanol. A specific activity of approximately 26 TBq mmol⁻¹ was obtained.

To determine the half-life of the hormone within the haemolymph, crabs were injected with radiolabelled Y²-CCAP in crab saline (Webster, 1986) into the coxopodite of the third leg, and haemolymph samples (100 μl) were taken from the hypobranchial sinus at different intervals (2, 5, 10, 20 and 30 min). After having taken the last haemolymph sample, the animals were anaesthetised on ice, killed by cutting the connectives between the cerebral and thoracic ganglia, and the gills and heart were dissected out. Haemolymph samples, gills and heart tissue were examined for radiolabelled Y²-CCAP using a gamma scintillation counter (Wallac 1470).

Entire ventral nerve cords without brains were carefully dissected in ice-cold standard saline (in mmol l⁻¹: NaCl, 205; KCl, 5.4; CaCl₂, 15.3; MgCl₂, 2.5; maleic acid, 5.0; Trizma base, 5.0; pH 7.4; modified after Van Harreveld, 1936). The ganglia were partially desheathed on the ventral side and incubated in tissue-culture plates (Linbro) at room temperature in a volume of 1 ml of different salines for 5 or 10 min. The standard procedure included several washing steps in Ca²⁺-free washing saline (WS; standard saline without Ca²⁺) interrupted by control steps in two different control salines [control saline 1 (CS1): free of Ca²⁺, 10×KCl (54 mmol l⁻¹); control saline 2 (CS2): standard saline]. The nervous system was depolarized with releasing saline (RS) in which the K⁺ content had been increased tenfold in the presence of Ca²⁺ (54 mmol l⁻¹ KCl, 15.3 mmol l⁻¹ CaCl₂). Incubations with control or releasing saline were always preceded and followed by at least two steps in washing saline. High-K⁺ salines (CS1, RS) were depleted of the same amount of Na⁺ (NaCl), and Ca²⁺-free salines (WS, CS1) were supplemented with the same amount of Mg²⁺ (MgCl₂) to maintain the same osmolarity.

All samples were frozen immediately after the incubation. Since Ca²⁺ obviously interferes with the phosphate buffer system in the CCAP–EIA, all samples had to be depleted of Ca²⁺. This was achieved either by Ca²⁺ precipitation using ice-cold phosphate buffer, followed by centrifugation (12 000 g, 5 min, 4 °C), or by Sep-Pak purification. The conditioning, equilibration and elution of the Sep-Pak cartridges were carried out as described above for the haemolymph samples. The eluate was dried in a vacuum centrifuge, redissolved in assay buffer (PBST, pH 7.4) and tested for released CCAP in a dilution series.

The ecdysis (E) stage of the green shore crab is differentiated into substages by the use of overt morphological and behavioural variables that are characteristic of the respective stages and are also applicable to the American crayfish, as summarised in Fig. 1. The proposed nomenclature (E₀–E₁₀₀) indicates the percentage of the whole ecdysis process. These percentages cannot strictly be correlated with the temporal pattern of ecdysis since the duration of a single behavioural event can vary considerably among the specimens. However, we observed that the sequences of appearance of distinct morphological characteristics and behavioural elements were absolutely constant among all specimens examined (N>15 for both species). Two major parts of the ecdysis process can be distinguished, the passive and the active parts. Whereas the passive phase (E₀–E₃₀) is mainly characterised by extensive water uptake resulting in body swelling and by pronounced inactivity of the animals, the active phase (E₃₀–E₁₀₀) comprises the final active and stereotyped exuviation behaviour patterns that include several rhythmic muscle movements in a fixed sequence with regard to their appearance in different parts of the body. This substaging allowed reproducible and substage-refined haemolymph sampling to unravel the temporal patterns of CCAP titres during ecdysis.

The duration of a complete ecdysis (passive and active) of crabs varies from 45 to 90 min, because the animal can prolong or even interrupt the passive phase, e.g. when disturbed by a potential predator, or, as is occasionally seen in the case of male crabs, when mating behaviour is preferred in the presence of a freshly moulted female of stage A or B. The entire ecdysis usually takes approximately 60 min, during which the passive phase lasts 45–50 min and the active phase takes 10–15 min in males and 5–10 min in females. The transition between the last premoult stage D₄ and stage E₀ is characterised by the widening from back to front of the pleural suture, which has opened to a narrow gap (finally up to approximately 1 mm) several hours before the ecdysis as a result of the beginning of water uptake in stage D₄ (Fig. 1). At a final gap size of 2–3 mm (stage E₅), the thoracoabdominal intertergal membrane ruptures, and continuous lifting of the old carapace starts in response to more extensive water uptake by the animal (Figs 1, 2A,B). Whilst this lifting of the back of the carapace continues beyond an angle of approximately 30 ° to the longitudinal axis, active ecdysis starts with the so-called ‘wiggling’ behaviour (Fig. 1) when the lateralmost spines of the new dark green carapace become clearly visible and emerge below the old carapace (Fig. 2B; stage E₃₀). The wiggling movements affect all appendages of the body for a very short period (approximately 20 s) and obviously assist the general loosening of the whole exoskeleton. Within a few minutes of the wiggling movement, the active shedding process of the head appendages clearly starts with unique and rhythmic horizontal spreading movements of the antennules, which halt their previously regular perpendicular flicking movements at this time. Head shedding consists of the shedding of the antennae, the eyestalks and the mouthparts, including the scaphognathites, thus depriving the animal of sensory inputs and of an oxygen supply from the water during these stages (Fig. 2B,C; E₃₀–E₅₀). This is almost directly followed by shedding of the abdomen, which covers the stages E₅₀–E₈₀ and is characterised by peristaltic contractions of the abdominal body musculature (Fig. 2C,D). Only after the third abdominal segment is withdrawn (Fig. 2C) does the animal begin characteristic contractions of the thoracic coxopodites and basipodites. These contractions, which always alternate between the two sides (duration approximately 5–6 s per contraction), mark the beginning of pereiopodal shedding, which is performed during stages E₇₀–E₁₀₀ (Fig. 2D–F). The alternating coxopodite contractions are performed regardless of whether the corresponding leg exists. Abdominal shedding is terminated by stage E₈₀ when pereiopodal shedding exhibits the strongest muscle movements (Fig. 2E). The entire ecdysis is finished when the chelae are completely removed from the old exuvia (stage E₁₀₀; Fig. 2F). Immediately after exuviation, the animal continues to take up water and expands to its final size within the first 2–3 h after ecdysis during stage A₁.

The behavioural pattern during ecdysis of the crayfish is similar to that of the crab with only minor differences. The onset of passive ecdysis is difficult to determine because water uptake commences in stage D₄ and continues throughout passive ecdysis but is less prominent than in the crab. However, bulging of the thoracoabdominal intertergal membrane marks the point of transition from stage D₄ to passive ecdysis, which has definitely started when the thoracoabdominal membrane is ruptured (Fig. 1; stage E₅). The animal finally rolls onto its side (stage E₁₀; Figs 1, 3A) and continues with the elevation of the posterior margin of the carapace, thereby revealing the homogeneously dark brownish new carapace (Fig. 3B). The onset of active ecdysis starts with shedding of the head appendages, which becomes especially clear when the eyestalk is suddenly retracted from the eyecup of the exuvia (Fig. 3B,C). Active ecdysis is usually performed over 10–18 min, a time range similar to that of active crab ecdysis. Within the next few minutes, after the head appendages have been withdrawn, the carapace is lifted to an angle of approximately 90 ° to the longitudinal axis (Fig. 3C). This is then directly followed by abdominal shedding, which is characterised by peristaltic contractions of the abdominal segments (Figs 1, 3C–F; stages E₅₀–E₁₀₀). When abdominal shedding is approximately half completed, shedding of the pereiopods starts and is similar to the behaviour seen during crab ecdysis, with extensive contractions of the coxopodites and basipodites of the legs (Fig. 3D). Furthermore, these are accompanied by peristaltic vermiform movements of the meropodites within the exosketal tube of the old leg (particularly visible through a ‘window’ of decalcified cuticle in the meropodites of the first leg). Shortly after the onset of pereiopodal shedding, scaphognathite beating becomes visible through the transparent branchiostegites (Fig. 1; stage E₇₅). In contrast to the situation in the crab, abdominal withdrawal takes longer and lasts until the animal finally frees itself from the exuvia using two or three tailflips, which are initiated as soon as the last abdominal appendages, the uropods and the telson are withdrawn (Figs 1, 3F; stage E₁₀₀). The animals perform these final tailflips regardless of whether pereiopodal shedding is complete. This can consequently lead to the loss of appendages, most often of the chelipeds.

The competitive EIA allowed the detection of CCAP (Fig. 4) at a limit below 5 fmol of CCAP per well (50 pmol l⁻¹), and the standard curve was routinely linear between 10 and 100 fmol of CCAP per well with an ED₅₀ of 32 fmol of CCAP per well. Peptide recovery after Sep-Pak (C₁₈) purification was approximately 85 % (estimated using radiolabelled Y²-CCAP), and samples of purified haemolymph showed curves parallel to those of authentic CCAP after logit/log transformations of dilution curves (Fig. 4, inset). Furthermore, the CCAP-immunoreactive fractions of purified samples co-eluted with the synthetic hormone after HPLC fractionation (Fig. 5A,B). This fact, combined with the parallel curves, indicated that the immunoreactive material was CCAP.

The first evidence of a large increase in the haemolymph titre of CCAP during ecdysis of the crab came from HPLC separations and subsequent assays of every fraction for CCAP immunoreactivity of intermoult (stage C₄) haemolymph in comparison with ecdysis haemolymph collected during active ecdysis at stages E₈₀–E₁₀₀ (Fig. 5A,B). For both stages, 20 ml of pooled haemolymph was analysed, each collected from a group of 10 medium-sized crabs (5–6 cm carapace width) regardless of sex. The average intermoult titre of approximately 30 fmol CCAP ml⁻¹ haemolymph (3×10⁻¹¹mol l⁻¹) in stage C₄ contrasted with a titre of at least 1000 fmol CCAP ml⁻¹ haemolymph (10⁻⁹ mol l⁻¹) during active ecdysis (Fig. 5A,B). Similar haemolymph concentrations of CCAP were found during the active stage of ecdysis of destalked animals, which had been operated upon in early premoult (stage D₂).

The temporal pattern of CCAP haemolymph content of crabs in different premoult and postmoult stages and especially during the ecdysis substages was evaluated by analysing the CCAP titre in the haemolymph of individual animals (Fig. 6).

The timing of haemolymph sampling was based upon the morphological and behavioural criteria described above. With the onset of active ecdysis (stage E₃₀), the CCAP titre increased progressively and reached a maximum of approximately 1.4 pmol ml⁻¹ haemolymph at the end of active ecdysis in stage E₁₀₀. Directly after ecdysis, the titres decreased to less than 30 % of the ecdysis maximum within 30 min (Fig. 6; stage A₁/0.5) and reached almost basal levels within the next 3 h during stage A₁ (Fig. 6; stage A₁/3).

Pericardial organs were extracted from medium-sized male and female crabs (carapace width 5–6 cm) at different moult stages. The CCAP contents of these organs are 2–3 times higher in premoult stage D₄ than in intermoult (C₄), ecdysial (E₈₀–E₁₀₀) and postmoult (A,B) stages (Fig. 7). The CCAP contents in stage D₄ amounted to 36 pmol per pair of pericardial organs and, after ecdysial release, 15 pmol per pair of pericardial organs was found upon completion of stage E. Thus, one pair of pericardial organs releases more than 20 pmol of CCAP during the whole ecdysis process.

Injections into the circulatory system of intermoult crabs (N=2; stage C₄) of 5×10⁵ cts min⁻¹ of [¹²⁵I]Y2-CCAP as a radiolabelled analogue of CCAP diluted in 100 μl of standard crab saline provided a first approach to estimating the half-life of CCAP in the haemolymph in vivo. Since it takes at least 2 min after injection until a substance is evenly distributed throughout the haemolymph (Webster, 1996), samples were taken 2, 5, 10, 15 and 30 min after the injection, during which time we observed an approximately exponential decline of sample radioactivity. At the end of the experiment, background radioactivity amounting to approximately 1.5×10⁴ cts min⁻¹ ml⁻¹ remained in the haemolymph. After subtraction of this background, a half-life for [¹²⁵I]Y²-CCAP of approximately 6 min was calculated (Fig. 8). Rapid dissection of the gills and the heart, immediately after taking the last haemolymph sample, showed that these organs bound less than 10 % of the injected peptide tracer (data not shown).

In experiments similar to those described for crabs, fractions obtained after gel filtration of haemolymph (1 ml per run) from single crayfish specimens in intermoult and ecdysis stages were analysed for CCAP content by EIA. In the haemolymph of an animal that had just finished ecdysis (see Fig. 3F; stage E₁₀₀), an enormous concentration of CCAP of almost 11 pmol ml⁻¹ (1.1×10⁻⁸ mol l⁻¹) was found in contrast to less than 100 fmol ml⁻¹ (10⁻¹⁰ mol l⁻¹) from an intermoult (stage C₄) animal (compare fractions 68–72 in Fig. 9). A minute amount of CCAP immunoreactivity was detected in the high-molecular-mass fraction of ecdysis haemolymph, which has not yet been analysed further (fractions 23, 24; Fig. 9).

Assays of haemolymph CCAP titres of 10 crayfish in different moult stages undergoing natural ecdysis (Fig. 10) resulted in an increase of more than 15-fold in CCAP content (range of maximum contents 5.5–11 pmol ml⁻¹) during active ecdysis (E) compared with the levels of premoult (D₄) or late postmoult (A₂–B₂) stages. After ecdysis, the hormone titres decreased rapidly within the first 3 h to approximately 30 % of the peak value observed at active ecdysis and reached almost basal intermoult levels within the next 24 h (stage B₁). All samples of stage E were taken from animals during the last 5 min of active ecdysis (stages E₈₀–E₁₀₀; Fig. 6). Thus, the actual haemolymph CCAP levels at ecdysis are approximately five times higher in the crayfish than in the crab. Furthermore, a clear-cut increase in haemolymph CCAP concentration was detected in destalked animals during ecdysis. However, the maximum haemolymph titres in these animals reach approximately 2 pmol ml⁻¹, which is approximately five times lower than the titres found in the haemolymph of intact animals undergoing ecdysis.

Several in vitro experiments using release-inducing high-K⁺ saline (releasing saline; RS) applied to isolated crayfish ventral nerve cords without brains showed that a large CCAP release can be evoked repeatedly. Whereas control salines (CS1, CS2) produced no significant release of CCAP above normal background levels (0.1–0.2 pmol), each incubation with releasing saline led to a release of approximately 3 pmol of CCAP. The system was not exhausted even after six successive incubation cycles (Fig. 11), since this large amount of peptide is only a small proportion of the content of a ventral nerve cord, which usually contains 70–80 pmol of CCAP. The hormone was typically detected in the releasing fraction and in two subsequent washing fractions, probably because of the relatively slow exchange of ions between the central parts of the ventral nervous system and the incubation saline. Notably, the use of washing saline or normal standard saline in the washing steps after high-K⁺ stimulation (RS) led to essentially the same results. We have not been able to detect any striking difference between the in vitro CCAP release capabilities of intermoult (C₄) and premoult (D₄) ventral nerve cords with regard to the quantities of hormone found after incubation in releasing saline.

The novel substaging using percentages of crab and crayfish ecdysis, designed following the criteria for a useful staging method suggested by Drach (1939), has allowed the substage-specific analysis of the temporal patterns of CCAP content in the haemolymph of two decapod species during natural ecdysis. This substaging system appears to be flexible, leaving space for further refinements in the future using details that we may have overlooked or which may become apparent only after partial dissection of the integument of the animal, e.g. after removal of parts of branchiostegites or limbs. The massive increase in haemolymph CCAP concentration coinciding in both species with active ecdysis was followed by a rapid decline during the early postmoult stages to normal intermoult levels, which strongly suggests that the peptide participates in the regulation of active ecdysis.

Our approach of subdividing the passive and active phases of ecdysis in species from two different decapod crustacean orders, Brachyura and Astacura, has revealed remarkable similarities with regard to the behavioural elements required for successful ecdysis (Figs 1–3). This is largely in accordance with observations described earlier for several other decapod species of crab, crayfish and lobster (Vitzou, 1882; Drach, 1939; Passano, 1960; Aiken and Waddy, 1987; Waddy et al., 1995). Most aspects in earlier detailed descriptions, especially of the ecdysis of Carcinus maenas by Vitzou (1882), match exactly, although this author has overlooked the vital step of water uptake which, as has recently been established in this species, is probably driven by CHH released into the haemolymph from intestinal paraneurons (Chung et al., 1999). This water uptake leads to the visible opening of the crab carapace at predefined pleural sutures or to bulging of the intertergal membranes posterior to the crayfish cephalothorax, providing a definitive sign of the beginning of ecdysis, as previously noted by Réaumur (1712). Although the use of these signs still makes it somewhat difficult to assign a clear and distinct starting point for ecdysis, since water uptake by drinking (Neufeld and Cameron, 1994; Chung et al., 1999) is not an overt behaviour, we consider these signs to be unequivocal indicators for the beginning ecdysis. Interestingly, even after their thoracoabdominal membranes have ruptured (thereby providing access of environmental water to exuvial fluids in the apolysis space), neither species necessarily accelerates its ecdysis process; the animals may continue ‘apparently normal’ activities for an unpredictable period, especially under certain circumstances, e.g. if a potential predator is visible. Therefore, a ‘point of no return’ during crustacean ecdysis appears to be attained only later, i.e. directly after active ecdysis has started with its distinct overt wiggling behaviour followed by the remarkable horizontal movements of the antennules in the case of the crab, first noted in the present study, or the shedding of the eyestalks in the case of the crayfish. These behavioural elements definitively mark the onset of the stereotyped behaviour patterns that consist of a shedding sequence of head appendages, abdomen and pereiopods. The movements of the pereiopods, in particular, are characterised by alternating unilateral contractions, suggesting the existence of a basic motor program controlling decapod ecdysis, if not ecdysis of crustaceans in general. Once active exuviation is initiated, ecdysis is either completed or the animal dies.

Vitzou (1882) has highlighted some differences between brachyuran and astacuran crustaceans, namely the existence, in crabs, of an epimeral suture line preformed for rupture by water uptake. He further stressed that in astacurans, contrary to the situation in the brachyurans, the extrication of the head, cephalothorax and walking legs precedes that of the abdomen, as has been described in lobsters (Aiken, 1980; Waddy et al., 1995). However, in Orconectes limosus, we always observed that freeing of the abdomen occurs earlier or perhaps concomitantly with withdrawal of the chelipeds and other thoracic appendages. The same sequence has been reported for another crayfish (O. virilis), which died during these ecdysis steps, but does not form part of the naturally observed sequence (Aiken, 1968); thus, this sequence may be unique to the natural ecdysis of O. limosus. In fact, the final withdrawal of the chelipeds is always accompanied by brisk movements of the tail, an observation that also differs from that described for O. virilis (Aiken, 1968). These tailflips resemble those known to govern crayfish escape behaviour (Wine and Krasne, 1982). They may be initiated by sensory inputs from the uropods, because the crayfish tries to escape from its exuvia as soon as the last abdominal segment is withdrawn, regardless of whether shedding of the pereiopods is complete. Thus, for stereotyped ecdysis behaviour, preformed neural circuits are obviously recruited to serve the different and specialised behavioural repertoire for ecdysis. Pereiopod shedding by the crab is characterised by alternating (by side) and peristaltic contractions of the legs, in particular the coxopodites and basipodites, a movement that, according to Drach’s earlier observations during ecdysis of the spider crab Maja squinado, resembles normal walking but may differ in that probably only the extensor muscles and not the flexor muscles of these leg compartments are being used (Drach, 1939). However, we do not favour the restrictive view of Drach (1939) that only these extensor muscles are involved in these movements, because we have regularly seen peristaltic-like muscle movements in other parts of the crayfish chelipeds during active ecdysis. Since in C. maenas the coxopodite contractions are performed regardless of whether the corresponding leg exists (e.g. after previous autotomy), it seems obvious that an underlying ecdysis motor program, once initiated, is independent of peripheral sensory inputs and probably cannot be interrupted. Furthermore, these alternating contractions correspond in terms of their temporal pattern to periodic changes in the hydrostatic pressure pulses, which are of much lower frequency than the heart rate during active ecdysis, as has been described for the blue crab Callinectes sapidus (deFur et al., 1985). These observations support the view that the haemolymph functions as a hydrostatic skeleton during ecdysis when the hard exoskeleton is shed and the new still-uncalcified skeleton is too weak to serve as a buttress or a stable tendon insertion point for proper muscle action (Drach, 1939; Mangum, 1992). Hardening of the shell in shore crabs during postmoult may take at least 3–16 days depending on the size of the crab, the temperature, the pH and the Ca2+ concentration in the water (Broekhuyzen, 1936).

The ecdysis of brachyuran and astacuran decapods is associated with a massive release of the neuropeptide CCAP into the haemolymph (Figs 6, 10). This event is restricted to the active ecdysis process, suggesting an important role for CCAP during complex ecdysis behaviour. Interestingly, while CHH titres in crab haemolymph show a similar temporal pattern during active ecdysis with a peak concentration at the end of ecdysis, CHH appears earlier in passive ecdysis, although at elevated but more moderate concentrations compared with premoult situations. During ecdysis, CHH certainly plays a role in the control of water uptake (Chung et al., 1999) and possibly also in energy resourcing during late active ecdysis (see below). The massive increase in CCAP titre beginning in E₃₀ with a peak concentration at E₁₀₀, which drops during the first few hours of the postmoult phase (A₁), suggests that this peptide is necessary for the successful termination of the active part of ecdysis only. A remarkably similar temporal pattern of changes in CCAP content in the ventral nerve cord has been described previously in the isopod Oniscus asellus during ecdysis (Johnen et al., 1995). However, the increase in CCAP concentrations in late premoult stage D₄ suggested an accumulation of the peptide in the isopod ventral nerve cord prior to a presumed peak of release during posterior ecdysis. Levels of CCAP remain high after anterior ecdysis, approximately 1 day later. Peaks of immunoreactivity in CCAP neurons in the ventral nerve cord correspond to the posterior and anterior exuviation phases (H. Dircksen and C. Johnen, unpublished observations). Our observations concerning the transient change in the CCAP content of the crab pericardial organ extracts allow a similar interpretation, i.e. that CCAP is accumulated in the pericardial organs during stage D₄ to meet the high demands for release of the hormone during ecdysis.

Specimens of both species lacking eyestalks show the same transient elevation of CCAP titre during active ecdysis as intact animals, which suggests that the main source of humoral CCAP is probably the neurohaemal pericardial organs, as has been recognized previously in crabs (Dircksen and Keller, 1988; Stangier et al., 1988), or the ventral nerve cord (Trube et al., 1994) and the pericardial organs in crayfish, and is not under the control of classical eyestalk neuroendocrine systems. One pair of pericardial organs from an intermoult crab contains approximately 20 pmol of CCAP, which is in accord with earlier estimations (Stangier et al., 1988), and a single crayfish ventral nerve cord contains four times as much. Thus, the stored CCAP in the central nervous system and pericardial organs can entirely account for the extraordinarily high levels found in ecdysis haemolymph, even taking into account the rapid hormone turnover at this time, as shown by the estimations of CCAP half-life. The difference between the fivefold higher peak haemolymph concentrations of CCAP in crayfish compared with those in the crab is not understood at present. Moreover, whether there is a causative relationship between this accumulation of CCAP and the premoult peak of ecdysteroid titres will have to be a matter for future studies.

Ecdysis obviously demands a complex regulation of cardiovascular functions affecting heartbeat performance and the redirection of haemolymph flow, since prebranchial hydrostatic haemolymph pressures caused by the excessive water uptake are known to rise steeply (approximately fivefold) from stage D₄ onwards, reaching a maximum in stage E followed by a rapid decline in stages A₁ and A₂ (deFur et al., 1985; Mangum, 1992). It is known that CCAP is cardioacceleratory and slightly inotropic (Stangier and Keller, 1991; Wilkens, 1995) and elevates several variables of heartbeat performance in the crab such as cardiac output, stroke work, power output and intraventricular pulse pressure (Wilkens and McMahon, 1992; Wilkens and Mercier, 1993). It is also involved in the redirection of haemolymph flow into different regions of the body (McGaw et al., 1994; Wilkens et al., 1996) that are known to be associated with varying blood pressures during ecdysis (Drach, 1939; Passano, 1960; deFur et al., 1985). Such redirection of flow is a compensatory response to lowered oxygen contents of the blood and to abdominal movements in crayfish, lobsters and crabs, probably because the scaphognathite is less efficient during ecdysis (Kuramoto and Ebara, 1984; McMahon, 1992; Reiber et al., 1992). It is likely, therefore, that CCAP participates in the control of haemolymph pressure and flow.

Transient anoxic states during ecdysis, as have been demonstrated for exuviating blue crabs C. sapidus (Lewis and Haefner, 1976) and lobsters (Clemens et al., 1999), may be continued during a short postmoult period of anaerobic metabolism (Mangum et al., 1985b; Mangum, 1992). In blue crabs, this occurs concomitantly with a significant decrease in ventilation rate at ecdysis (deFur et al., 1985; Mangum et al., 1985a), although we have regularly seen overtly beating scaphognathites from stage E₇₅ onwards. Since CCAP haemolymph titres are still moderately high shortly after ecdysis, the peptide could be involved in a stimulation of scaphognathite ventilatory activity, thus matching earlier observations at the end of ecdysis (Passano, 1960), and/or in pyloric motoneuron rhythmicity within the stomatogastric neural network, both of which are well-established effects of CCAP in intermoult animals (Stangier and Keller, 1991; Weimann et al., 1997). This view would be in accordance with events in blue crabs and lobsters shortly after ecdysis. For instance, Mangum et al. (1985a,b) observed increased haemolymph and increased rates of O₂ uptake by several muscles of blue crabs during postmoult stage A₁, while lactate levels were still high. In lobsters, the lack of a relationship between persisting hypoxic conditions and the normal (compared with the slower ecdysial) pyloric network rhythmicity directly (approximately 10 min) after ecdysis has suggested that ‘unknown modulatory factors’ were influencing the neuronal activities at the end of ecdysis (Clemens et al., 1999). One of these factors may well be CCAP, which can both elicit and modulate motoneuronal activities in members of different arthropod groups (Mulloney et al., 1997; Weimann et al., 1997; Gammie and Truman, 1997a) and shows direct myotropic actions on several visceral muscle systems such as those of the crustacean hindgut, the locust hindgut and oviduct and even the skeletal muscles of the metathoracic extensor tibiae in this insect (for reviews, see Dircksen, 1994, 1998). Consequently, a potentiation of muscular force at ecdysis may be considered a major function of humoral CCAP during and after the active crustacean ecdysis process, particularly since several limb muscles experience a reduction in mass of up to approximately 50 % during premoult compared with their original size and mass during intermoult (Mykles and Skinner, 1981; Mykles, 1997). A factor potentiating muscle power would thus be of great help, if not indispensable, during active ecdysis, which is probably the most power-demanding behaviour in the crustacean life cycle. Furthermore, this would call for the participation of CHH, which shows its highest titres at this time (Chung et al., 1999), in providing additional energy by mobilising glucose from glycogen stores.

Taken together, these results are indicative of a similar importance for the peptide CCAP during ecdysis in several malacostracan crustaceans and support the roles proposed earlier for CCAP as a highly conserved peptide in at least two putatively homologous CCAP neurons occurring in almost all different arthropod groups (Dircksen, 1998). Although, at present, we have no electrophysiological or immunocytochemical data on the behaviour of such CCAP neurons during ecdysis of crustaceans, their activity might be expected to resemble that of the well-documented CCAP-immunoreactive cell 27 (NS-L₁) and IN 704 groups in Manduca sexta (for a review, see Truman et al., 1998). These neurons exhibit large transient increases in immunoreactivity for CCAP and the second messenger cyclic GMP during the active parts of several ecdyses (Ewer et al., 1994, 1997; Ewer and Truman, 1996, 1997; Truman et al., 1996) and are the terminal executives of Manduca sexta ETH and EH action controlling the ecdysis motor programme in the insect ventral nerve cord (Gammie and Truman, 1997a,b, 1999). These neurons are the homologues of the type-1 and type-2 CCAP neurons that occur in several arthropod groups (Dircksen et al., 1991; Davis et al., 1993; Dircksen, 1994, 1998; Truman et al., 1998). The humoral release of CCAP, described here for the first time, is probably related only to the neurosecretory type-1 cells, which project into central and peripheral neurohaemal organs in crab and crayfish (Dircksen and Keller, 1988; Trube et al., 1994). Preliminary evidence from iontophoretic and backfilling studies (A. Trube, G. Plachta and H. Dircksen, unpublished results) suggests that these two neuron types are electrically coupled and thus must act simultaneously during arthropod ecdysis, thereby linking a central CCAP release onto motoneuronal networks executing the ecdysis motor programme to a parallel peripheral release that controls and modulates scaphognathite beating, systemic haemolymph pressure and flow alterations and probably several muscle systems. The humoral aspects seem to be in agreement with earlier reports on the involvement of cardioactive factors in the haemolymph of M. sexta during adult ecdysis and wing-spreading behaviour (Tublitz and Truman, 1985a,b), one of which is CCAP (Cheung et al., 1992) most likely released from the cell 27 group via perivisceral organs.

The discovery of an association between CCAP and crustacean ecdysis represents the first step in unravelling the most critical part (in terms of neuroendocrine and motor control) of the crustacean moult cycle, the ecdysis. Much has still to be learnt about the control of CCAP release and the putative target organs of this neurohormone at ecdysis. The lack of CCAP release in standard saline incubations (CS2; Fig. 11) in our crayfish ventral nerve cord in vitro system showed that there is no spontaneous release from CCAP neurons. Thus, the existence of some ecdysis-inducing factors might be expected, but there is no evidence for the existence of any Manduca ETH- or EH-like compounds in crustaceans. Our in vitro CCAP release assay in combination with the highly sensitive EIA provide a rationale for the elucidation of other neuroendocrine factors acting on CCAP neurons, thereby regulating the complex physiological and behavioural components of crustacean ecdysis.

We are grateful to Mr A. Tweedale, Bangor, for help with the supply of crabs. This work was supported by grants from the German Academic Exchange Service (DAAD-ARC program) to H.D. and a Natural Environment Research Council grant to S.G.W.