Discovery of a neuropeptide that acts as an autotomy-promoting factor

One of the most remarkable adaptations to survive attacks from predators is to detach an appendage—a process known as autotomy. This occurs in a variety of animals, including lizards (tail), crabs (legs), and starﬁsh (arms). There has been extensive investigation of the evolution, ecology, and biomechanical impact of autotomy, 1-3 but little is known about neural mechanisms controlling autotomy in animals. However, evidence for the existence of a peptide that acts as an autotomy-promoting factor in starﬁsh has been reported. 4 While investigating in vivo effects of a sulfakinin/cholecystokinin-type neuropeptide (ArSK/CCK1) in the star-ﬁsh Asterias rubens, 5,6 we observed that this peptide triggered arm autotomy in some animals. Furthermore, when injection of ArSK/CCK1 was combined with mechanical clamping of an arm, autotomy of the clamped arm occurred in 85% of animals tested, with 46% also autotomizing one or more other arms. In contrast, no autotomy


RESULTS
Advances in transcriptome/genome sequencing have facilitated discovery of a huge variety of neuropeptides in the starfish A. rubens. 5We recently reported the identification of two neuropeptides in A. rubens derived from the same precursor protein that are related to sulfakinin (SK)-type neuropeptides in insects and cholecystokinin (CCK)-type neuropeptides in vertebrates and that we therefore refer to as ArSK/CCK1 and ArSK/CCK2 (Figure 1).SK/CCK-type neuropeptides act as inhibitory regulators of feeding behavior in insects and vertebrates and, accordingly, we discovered that ArSK/CCK1 and ArSK/CCK2 trigger cardiac stomach contraction and retraction and inhibit feeding behavior in A. rubens. 6However, when testing the in vivo pharmacological effects of ArSK/CCK1 we have also observed that it induces arm autotomy in some animals.Thirty-three animals were injected with 10 mL of 10 À3 M ArSK/CCK1 and in five of these animals, arm autotomy was observed following injection of the neuropeptide.In contrast, arm autotomy was not observed in animals injected with 10 mL of 10 À3 M ArSK/CCK2 (n = 25) or in control animals injected with 10 mL of water (n = 54) (Figure 2A).Among starfish injected with ArSK/CCK1, in two animals autotomy of one arm was observed, in one animal autotomy of three arms was observed, and in two animals autotomy of four arms was observed.Examples of the latter are illustrated in Figures 2B-2F and described henceforth.Within a 10 min period after injection of ArSK/CCK1, constriction of the body wall was observed at the junction of the arms with the central disc region and at a position consistent with that of the tourniquet muscle in the autotomy plane (Figure 2B).Following this, arm autotomy began to occur, with breakage of the body wall first occurring on both sides of the ambulacrum and with autotomizing arms being twisted into an abnormal position (Figures 2C-2E).Eventually the body wall of the autotomizing arms became completely separated from the central disc and the only structures connecting those arms with the central disc were the pyloric ducts that link the pyloric stomach in the central disc with the pyloric caeca in the arms (Figure 2F).Finally, the pyloric ducts of the autotomizing arms broke and only one arm remained attached to the central disc.
To facilitate investigation of a potential physiological role of ArSK/CCK1 as a regulator of autotomy in A. rubens, we performed experiments to identify mechanical stimuli that trigger arm autotomy in this species.Our aim was to use a method that mimics the pressure exerted on arms when starfish are attacked by predators; for example, when the arm of a starfish is clamped in the beak of a seagull. 9To accomplish this, a Mohr tubing clip was used to clamp one arm and then the animal was suspended by this clip, but with its other arms in contact with seawater contained in a glass bowl below.Arms were clamped with the Mohr tubing clip in one of three positions: (1), proximal (P) to the junction of the arm with the central disc, coinciding with the position of the autotomy plane; (2), mid-way (M) along the length of an arm (Figure 3A); or (3), distal (D) from the central disc and proximal to the arm tip.In animals with arm clamping proximal to the central disc region, autotomy of the clamped arm was observed in 80% of animals tested (16 out of 20).In contrast, when the arm was clamped mid-way along its length or proximal to the arm tip, autotomy of the clamped arm was observed in only 10% (2 out of 20) and 4.2% (1 out of 24) of animals tested, respectively (Figure 3B).The time taken for arm autotomy to occur was also measured and this ranged from as little as 103 s in an animal that was clamped proximal to the central disc to as much as 1,333 s in one starfish that exhibited arm autotomy with clamping proximal to the arm tip.The mean time taken for arm autotomy was 303.4 s (n = 16) and 118.5 s (n = 2) in starfish clamped proximal to the central disc or mid-way along the length of an arm, respectively (Figure 3C).Following autotomy, we also examined the precise location where autotomy had occurred in animals and found that it occurred at a position corresponding to the junction between the second and third, third and fourth, or fourth and fifth ambulacral segments (Figure 3D), consistent with previously reported findings for autotomy in this species. 7aving discovered that clamping mid-way along the length of an arm induced autotomy in only 10% of starfish tested, we then investigated a potential autotomy-promoting effect of ArSK/ CCK1 by combining this clamping method with injection of ArSK/CCK1 (10 mL of 10 À3 M).In parallel, experiments were performed in the same way to test the effect of ArSK/CCK2.Control experiments were performed where clamping mid-way along the length of an arm was combined with injection of 10 mL water, and in these experiments arm autotomy was not observed in any of the animals tested.In striking contrast, in animals where clamping mid-way along the length of the arm was combined with injection of ArSK/CCK1, autotomy of the clamped arm was observed in 84.6% of animals (11 out of 13; Figure 3E).Furthermore, in five animals one (n = 1), two (n = 2), or three (n = 2) of the other arms were also autotomized (Figure 3F).In animals injected with ArSK/CCK2 (10 mL of 10 À3 M), autotomy of the clamped arm occurred in 27.3% of animals (3 out of 11 animals; Figure 3I) and in one of these animals, two of the other arms were also autotomized (Figure 3J).Analysis of the site of autotomy revealed that in both ArSK/CCK1-treated and ArSK/CCK2-treated animals, arm autotomy occurred between the first and second, second and third, or third and fourth ambulacral segments (Figures 3H and  3L).In conclusion, these experiments revealed that both ArSK/ CCK1 and ArSK/CCK2 promote autotomy in A. rubens, but with ArSK/CCK1 being more effective than ArSK/CCK2.
Having observed the effect of ArSK/CCK1 in promoting autotomy in A. rubens, we investigated the physiological relevance of this effect by using immunohistochemistry to examine the expression of this neuropeptide.We have reported previously the expression of ArSK/CCK1 in A. rubens using affinitypurified antibodies to this neuropeptide. 6However, this study did not examine ArSK/CCK1 expression in the relatively small region at the base of arms where autotomy occurs-the autotomy plane.The body wall of A. rubens contains an endoskeleton comprising a network of calcite ossicles, which are linked by interossicular muscles that mediate changes in body posture.1][12] Arm autotomy in starfish requires irreversible breakage of both the muscular and collagenous components of the body wall that interlink endoskeletal ossicles in the autotomy plane.Furthermore, analysis of the structure of the autotomy plane has revealed two specialized features that are not seen in other regions of the arms.First, in the dermis of the dorsal and lateral parts of the body wall there is an expanded, sparsely fibrous region, which contains widely separated bundles of collagen fibrils and scattered microfibrils.Second, proximal to this region there is a circumferential band of muscle, which is referred to as the tourniquet muscle because it mediates local constriction of the arm during and after autotomy. 7,8To investigate ArSK/CCK1 expression in the autotomy plane of A. rubens, two series of adjacent transverse sections of arms in the region of the autotomy plane were prepared.Then one series of sections was stained using Masson's trichrome method to enable analysis of tissue histology, and the second series of sections was processed for immunohistochemistry using affinity-purified antibodies to ArSK/CCK1.Examination of trichrome-stained sections enabled identification of the tourniquet Sulfakinin/cholecystokinin-type precursor protein (ArSK/CCKP) has an N-terminal signal peptide (blue) and two SK/CCK-type neuropeptide sequences (red), which are bounded by putative dibasic or tetrabasic cleavage sites (green).The C-terminal glycine of each neuropeptide sequence is shown in orange to indicate that this residue is a potential substrate for post-translational conversion to an amide group.Use of mass spectrometry to determine mature structures of the neuropeptides ArSK/ CCK1 and ArSK/CCK2 derived from ArSK/CCKP reveals three post-translational modifications.In both peptides, the C terminus is amidated and the tyrosine residue is sulfated.Furthermore, an N-terminal glutamine residue is converted to pyroglutamate in ArSK/CCK1.Sequence and structural data are from Semmens et al. 5 and Tinoco et al. 6 muscle, which is located beneath the epidermis of the body wall in the aboral region of the arm proximal to the position in the arm where the gonoducts are located (Figures 4A-4C and 4E).Furthermore, analysis of adjacent sections revealed the presence of ArSK/CCK1-immunoreactive nerve fibers amongst the muscle fibers of the tourniquet muscle (Figures 4D and 4F).

DISCUSSION
Here, we report the discovery of a neuropeptide, ArSK/CCK1, that acts an autotomy-promoting factor in the starfish A. rubens.Evidence for the occurrence of autotomy-promoting factors in starfish was first reported by Chaet, who observed that a substance derived from the coelomic lining of the body wall in scalded specimens of A. forbesi induces autotomy when injected into untreated starfish. 13A more detailed (A) A control animal that was injected with 10 mL of distilled water; no autotomy was observed during a 5 h period of monitoring after injection or when examined 24 h after injection in this animal or in 53 other animals injected with water.(B-D) Test animal (no. 3) after injection of ArSK/ CCK1.By $10 min after injection of ArSK/CCK1, indentation of the body wall at the junctions of the arms with the central disc region can be seen at a position that corresponds with the location of the tourniquet muscle (B; arrowheads).The first signs of autotomy were observed at $30 min after injection, with one of the arms twisted and partially detached from the central disc region (C; white arrowhead).By 2 h after injection, four of the five arms were twisted and partially detached from the central disc (D; black asterisks) and only one arm remained untwisted and completely attached to the central disc (D; white asterisk).By 2 h and 40 min after injection of ArSK/CCK1, four of the five arms were completely autotomized and a single arm remained attached to the central disc (image not shown).(E and F) Test animal (no.4) after injection of ArSK/ CCK1; by 1 h and 40 min after injection of ArSK/ CCK1, autotomy of four arms is ongoing, with the black arrowheads in (E) labeling regions of the body wall on either side of the ambulacrum where breakage has occurred.By 2 h and 15 min after injection of ArSK/CCK1, autotomy of three arms is completed (F; white arrowhead showing an autotomized arm), one arm remains attached to the central disc only by the pyloric duct, which links the pyloric stomach in the disc with pyloric caeca in the arm, (F; black arrowhead), and only one arm remains completely attached to the central disc (F; white asterisks).investigation of the starfish Pycnopodia helianthoides revealed that coelomic fluid collected from autotomizing starfish was found to induce autotomy in untreated starfish, whereas coelomic fluid from untreated starfish had no such effect.Thus, the existence of an autotomy-promoting factor in P. helianthoides was postulated.Investigation of the source of the autotomy-promoting factor revealed that it is derived from the body wall and released into the coelomic fluid.Furthermore, biochemical characterization and partial purification of the autotomy-promoting factor revealed it to be a heat-labile substance(s) whose activity is reduced by or destroyed by proteolytic enzymes and that has an estimated molecular mass of $1,200 Da.Accordingly, it was concluded that the autotomypromoting factor in P. helianthoides is a peptide or peptides derived from the body wall, but the molecular identity of the autotomy-promoting factor was not determined. 4Interestingly, the molecular mass of ArSK/CCK1 in A. rubens is 1,613 Da and therefore it is possible that ArSK/CCK1 is in fact an ortholog of the autotomy-promoting factor that was detected in P. helianthoides.
ArSK/CCK1 is one of two SK/CCK-type neuropeptides in A. rubens, with a second peptide (ArSK/CCK2) being derived from the same precursor protein as ArSK/CCK1 (Figure 1).Therefore, it is likely that ArSK/CCK1 and ArSK/CCK2 are coreleased from the same nerve processes in A. rubens.When first testing the in vivo effects of ArSK/CCK2, arm autotomy was never observed.However, when injection of ArSK/CCK2 was combined with mechanical clamping of an arm, autotomy of the clamped arm was observed in 27% of animals injected with ArSK/CCK2, which contrasts with autotomy occurring in 85% of animals injected with ArSK/CCK1.This indicates that ArSK/CCK2 is less important/effective than ArSK/CCK1 as a regulator of arm autotomy in A. rubens.Our previous studies have revealed that both ArSK/CCK1 and ArSK/CCK2 act as ligands for a single SK/CCK-type G-protein coupled receptor in A. rubens (ArSK/CCKR).Furthermore, the peptides exhibit similar potency as ligands for this receptor, with ArSK/CCK1 having an EC 50 of 0.25 nM and ArSK/CCK2 having an EC 50 of 0.12 nM. 6However, these EC 50 values were determined for ArSK/CCKR when expressed heterologously in a mammalian cell line.Interestingly, consistent with the differences in the effectiveness of these peptides as regulators of autotomy reported here, ArSK/CCK2 was found to be less effective than ArSK/ CCK1 in causing contraction of in vitro preparations of tube feet (locomotory organs) from A. rubens; and ArSK/CCK2 was also found to be less effective than ArSK/CCK1 in acting as an inhibitory regulator of feeding behavior in A. rubens. 6Differences (A and B) Transverse sections of arms at the position of the autotomy plane in A. rubens, which were stained using Masson's trichrome method to reveal the histology of the aboral body wall.The tourniquet muscle (tm) is located between the external epithelium and calcite ossicles that form the bulk of the body wall.Other anatomical features that can be seen here are the immature gonad (gon), gonoduct (gd), the pyloric duct (pd), which links the pyloric stomach located in the central disc region with the pyloric caeca located in the arm, coelomic epithelium (ce), papulae (pa), and pedicellariae (pe).(C-F) High magnification images of sections showing the tourniquet muscle stained using the Masson's trichrome method (C and E; muscle fibers are red; collagen fibers are blue) and adjacent sections immunostained using affinity-purified antibodies to ArSK/CCK1, with the boundaries of the muscle indicated by the black dots (D) and (F).ArSK/CCK1-immunoreactive nerve fibers can be seen within the muscle tissue (white arrowheads), with the approximate position of these fibers indicated in the trichrome-stained sections (C) and (E) by white asterisks.in the effectiveness of ArSK/CCK1 and ArSK/CCK2 as regulators of arm autotomy, feeding behavior, and tube foot contractility could be due to differences in the half-life of these peptides resulting from variability in their susceptibility to degradative peptidases.It is noteworthy in this respect that during its biosynthesis, ArSK/CCK1 is post-translationally modified to acquire an N-terminal pyroglutamate, whereas the N terminus of ArSK/CCK2 is formed by the amino group of a glycine residue (Figure 1).Pyroglutamylation confers protection of neuropeptides against degradation by aminopeptidases 14,15 and therefore the presence of an N-terminal pyroglutamate in ArSK/CCK1 probably explains why it is more effective than ArSK/CCK2 when tested in vivo.
Histological analysis of arm tissue during the process of autotomy in A. rubens has provided insights into the mechanisms of autotomy.Thus, arm autotomy is thought to involve the following processes in the autotomy plane at the base of each arm.First, there is contraction of the tourniquet muscle that causes constriction of the arm in the region of the autotomy plane and this is accompanied by irreversible softening and breakage of the collagenous tissue in dorsal (aboral) and lateral regions of the body wall.Second, there is irreversible breakage of muscles/ligaments that link ossicles in the ventral (oral or ambulacral) region of body wall. 7Informed by these findings, we investigated the expression of ArSK/CCK1 in the autotomy plane of A. rubens, employing use of immunohistochemistry with specific antibodies to this neuropeptide.Importantly, this revealed ArSK/CCK1-immunoreactive nerve fibers in the tourniquet muscle.In a previous study, we reported that ArSK/CCK1 acts as muscle contractant in A. rubens, causing contraction of in vitro preparations of three neuromuscular organs in this species: the cardiac stomach, the body wall-associated apical muscle, and tube feet (locomotory organs). 6Therefore, a parsimonious explanation for the arm-autotomy-promoting effect of ArSK/CCK1 in A. rubens is that it causes contraction of the tourniquet muscle.Evidence in support of this hypothesis can be found in our observation that constriction of the arm in the autotomy plane occurred in animals that autotomized arms following injection of ArSK/CCK1.However, constriction of the arm in the autotomy plane was also observed in animals injected with ArSK/CCK1 that did not autotomize an arm.Therefore, the autotomy-promoting effect of ArSK/CCK1 cannot be solely attributed to its proposed contracting effect on the tourniquet muscle.As discussed above, for arm autotomy to occur there must be irreversible softening and breakage of tissues in the region of the autotomy plane.Accordingly, ArSK/CCK1 may also be involved in regulation of tissue softening and/or breakage.
From an integrative perspective, it is of interest to consider the autotomy-promoting effect of SK/CCK-type neuropeptides in A. rubens in the context of other actions of these neuropeptides in this species and in other taxa.Neuropeptides are often pleiotropic, and it may be simplistic to think that their diverse actions are necessarily linked at the level of the whole animal.Nevertheless, it is noteworthy that SK/CCK-type neuropeptides, in addition to their well-established roles in inhibitory regulation of feeding behavior, are also involved in anxiety-, panic-and stress-related behavior. 16,17Therefore, our discovery that SK/CCK-type neuropeptides both inhibit feeding-related processes 6 and act as autotomy-promoting factors in A. rubens may be evidence of an integrative role in mediating adaptive responses to stressors.Hence, facilitated by release of SK/CCKtype neuropeptides, both the termination of feeding and arm autotomy in A. rubens could be components of a coordinated response to predation and other stressors. 18,19However, SK/ CCK-type neuropeptides are unlikely to be solely responsible for mediating neural control of arm autotomy in A. rubens because injection of ArSK/CCK1 in vivo only induced autotomy in a few animals.Furthermore, because ArSK/CCK1 and ArSK/ CCK2 induced arm autotomy (85% and 27%, respectively) when combined with a mechanical stimulus that did not trigger autotomy in control animals (water injected), these findings indicate that mechanical stimulation triggers the release of other substances that act synergistically with SK/CCK-type neuropeptides to trigger autotomy in A. rubens.Further studies will be needed to identify what may be a quite complex ''cocktail'' of signaling molecules that act in concert with SK/CCK-type neuropeptides to mediate neural control of autotomy in starfish.
In conclusion, this study reports the molecular identification of a neuropeptide (ArSK/CCK1) that acts as a regulator of autotomy in the starfish A. rubens.Discovery of this role of ArSK/CCK1 was facilitated by previous studies that have investigated the mechanisms of autotomy in A. rubens and the use of this species as a model system for investigation of the evolution and comparative physiology of neuropeptide signaling systems. 6,7,20,21Recognizing that autotomy is a key adaptation for survival that has evolved in several animal taxa, the findings of this study provide a seminal insight into the neural mechanisms that control this remarkable biological process.

METHOD DETAILS
Testing in vivo effects of SK/CCK-type neuropeptides on behavior in A. rubens As reported previously, 6 the effects of the SK/CCK-type neuropeptides ArSK/CCK1 and ArSK/CCK2 on starfish behavior were investigated.Following a period of starvation for twenty-seven days, specimens of A. rubens were injected with 10 mL of 10 À3 M ArSK/ CCK1 (n = 33, with mean diameter of 11.8 ± 0.4 cm) or with 10 mL of ArSK/CCK2 10 À3 M (n = 25, with mean diameter 12.8 ± 0.3 cm).Control experiments were performed for both neuropeptides where animals were injected with 10 mL of distilled water: ArSK/CCK1 control group (n = 29, with mean diameter of 11.5 ± 0.5 cm) and ArSK/CCK2 control group (n = 25, with mean diameter of 13 ± 0.3 cm).The behavior of the starfish was observed continuously for 3 h after injection, at 15-min intervals for a further period of 2 h and then finally at 24 h after injection.Representative images of feeding and autotomy behavior were taken using a Canon EOS 700D camera.

Investigation of mechanically induced autotomy in A. rubens
Because injection of ArSK/CCK1 was found to induce arm autotomy in some specimens of A. rubens, we also investigated mechanically induced arm autotomy in this species.A Mohr tubing clip was used to clamp one arm of each starfish tested and then the animal was suspended by this clip, but with its other arms in contact with seawater contained in a glass bowl below.Sixty-four starfish (mean diameter 11.4 ± 0.2 cm) were used for this experiment, all of which had been starved for one week in order to normalize their physiological status.Arms were clamped with the Mohr tubing clip in one of three positions: i).proximal (P) to the junction of the arm with the central disc, coinciding with the position of the autotomy plane (n = 20), ii).mid-way (M) along the length of an arm (n = 20) or iii).distal (D) from the central disc and proximal to the arm tip (n = 24).Each experiment was video recorded using a GoPro Hero 6 camera for 1 h or until the animal had autotomized the clamped arm.If autotomy occurred, starfish were observed under the microscope to determine the location of the breakage point along the length of the arm by counting the number of rows of tube feet remaining following autotomy.The time taken for autotomy of the clamped arm to occur (i.e., from injection to when the arm broke and the rest of the body fell into the bowl of seawater below) and the total number of autotomized arms per animal were also recorded.
Investigation of effects of SK/CCK-type neuropeptides when combined with arm clamping in A. rubens Our investigation of mechanically induced autotomy in A. rubens revealed that clamping of an arm with a Mohr tubing clip only induced autotomy reliably when the clip was positioned proximal to the disc (80% of tests).In contrast, when the clip was positioned mid-way along the length of the arm or close to the arm tip, autotomy only occurred in a few animals (10% or 4.2%, respectively).Informed by these findings, we investigated if the proportion of animals exhibiting autotomy would be increased if clamping midway along the length of an arm was combined with injection of ArSK/CCK1 or ArSK/CCK2.For this experiment, a total of fifty-two starfish (mean diameter 11 ± 0.2 cm) were first starved for one week in order to normalize their physiological status.Then, the animals were divided into a control group (injected with 10 mL of distilled water), with sixteen control animals used for the ArSK/CCK1 experiment and twelve control animals used for the ArSK/CCK2 experiment, and a test group (injected with 10 mL of ArSK/CCK1 or ArSK/ CCK2 at a concentration 10 À3 M), with thirteen animals used for the ArSK/CCK1 experiment and eleven animals used for the ArSK/ CCK2 experiment.The rationale for injecting 10 mL of peptide at a concentration of 10 À3 M was based on the approximate volume of coelomic fluid in starfish of the size used in these experiments ($10 mL) so as to achieve a final concentration in the coelom of $10 À6 M, which is a high but physiologically relevant concentration for neuropeptides.As described previously, 6 injection of ArSK/CCK1 or ArSK/CCK2 was performed by inserting the needle of a 50 mL syringe through the aboral body wall into the perivisceral coelom of an arm, proximal to its junction with the central disc region.Injection was consistently administered into the arm located in an adjacent anti-clockwise position with respect to the madreporite.Then, the same arm was clamped mid-way along its length using a Mohr tubing clip and the animal was suspended by this clip, but with its other arms in contact with seawater contained in a glass bowl below.The animal was video recorded using a GoPro Hero6 camera for a period of 35-40 min.If autotomy occurred, the time taken for autotomy of the clamped arm to occur was recorded (i.e., from injection to when the arm broke and the rest of the body fell into the bowl of seawater below) and starfish were observed under the microscope to determine the location of the breakage point along the length of the arm by counting the number of rows of tube feet remaining following autotomy.If autotomy of unclamped arms occurred during the experiment, this was also noted.Furthermore, starfish were also observed for 4 h after the experiment to observe if any arm autotomy occurred during this period.
Immunohistochemical localization of ArSK/SSK1 in the autotomy plane of A. rubens Specimens of A. rubens (n = 3) were fixed and sectioned, as described previously. 6However, for this study we focused specifically on immunohistochemical analysis of the arm region proximal to the central disc, which comprises the autotomy plane.Antibodies to ArSK/CCK1 were affinity-purified from a rabbit antiserum (RRID: AB_2877176) using AminoLink Plus Immnobilisation Kit (Thermo Fisher Scientific) and then eluted with trimethylamine (TEA; Sigma Aldrich).The eluted antibodies were diluted 1:15 in 5% normal goat serum (NGS; Sigma Aldrich)/phosphate-buffered saline containing 0.1% Tween 20 (PBST) and applied to sections.Following a series of washes in PBST, indirect immunohistochemical detection was carried out using Peroxidase-AffiniPure Goat Anti-Rabbit IgG (H + L) conjugated to Horseradish Peroxidase (RRID: AB_2313567; Jackson ImmunoResearch, West Grove, PA) diluted 1:1000 in 2% NGS/PBST.Bound antibodies were revealed using a solution containing 0.015% hydrogen peroxide, 0.05% diaminobenzidine (VWR Chemicals, Leicestershire, UK) and 0.05% nickel chloride (Sigma-Aldrich, Gillingham, UK) in PBS.Photographs of

Figure 1 .
Figure 1.Sequence of the A. rubens sulfakinin/cholecystokinin-type precursor protein and structures of neuropeptides derived from this precursor-ArSK/CCK1 and ArSK/ CCK2

Figure 3 .
Figure 3. SK/CCK-type neuropeptides promote mechanically induced autotomy in A. rubens (A) One arm of each starfish tested was clamped using a Mohr tubing clip (white arrowhead) and a retort stand clamp (black arrowhead) was used to support the Mohr tubing clip so that the starfish was suspended over a glass bowl containing seawater.The clip was positioned proximal to the disc (n = 20), mid-way along the length of the arm (n = 20; as shown in this photograph), or at the arm-tip distal from the disc (n = 24).(B) Clamping of an arm proximal (P) to the central disc triggers autotomy in 80% of animals tested, whereas clamping mid-way (M) along the length of an arm or at the arm-tip distal (D) from the central disc only triggers autotomy in 10% and 4.2% of animals, respectively.(C) The mean time taken for autotomy of the clamped arm to occur was 303.4 s (n = 16), 118.5 s (n = 2), and 1,333 s (n = 1) for animals clamped in the P, M, and D positions, respectively.(D) Proportion of animals in which arm autotomy occurred between the second and third (2-3), third and fourth (3-4), and fourth and fifth (4-5) ambulacral segments of the clamped arm.(E) When starfish injected with 10 mL water (control) are clamped M along the length of an arm (Water + M), no arm autotomy is observed.However, when starfish injected with ArSK/CCK1 (10 mL 10 À3 M) are clamped M along the length of an arm (ArSK/CCK1 + M), autotomy of the clamped arm occurs in 84.6% of animals.(F) Clamped animals injected with ArSK/CCK1 (ArSK/CCK1 + M) autotomized one (54.5%),two (9.1%), three (18.2%),or four arms (18.2%).(G) The mean time taken for autotomy of the clamped arm to occur was 440.8 s in ArSK/CCK1injected animals (ArSK/CCK1 + M). (H) Proportions of ArSK/CCK1-injected animals (ArSK/CCK1 + M) in which arm autotomy occurred between the first and second (1-2), second and third (2-3), and third and fourth (3-4) ambulacral segments of the clamped arm.(I) When starfish injected with 10 mL water (control) are clamped M along the length of an arm (Water + M), no arm autotomy is observed.However, when starfish injected with ArSK/CCK2 (10 mL 10 À3 M) are clamped M along the length of an arm (ArSK/CCK2 + M), autotomy of the clamped arm occurs in 27.3% of animals.(J) Clamped animals injected with ArSK/CCK2 (ArSK/CCK2 + M) autotomized one (66.7%)or three arms (33.3%).(K) The mean time taken for autotomy of the clamped arm to occur was 285 s in ArSK/CCK2-injected animals (ArSK/CCK2 + M). (L) Proportions of ArSK/CCK2-injected animals (ArSK/CCK2 + M) in which arm autotomy occurred between the first and second (1-2) and second and third (2-3) ambulacral segments of the clamped arm.Data are expressed as percentages or means ± s.e.m.Examples of experiments in (E) are shown in Videos S1 (ArSK/CCK1 + M) and S2 (water + M).Examples of experiments in (I) are shown in Videos S3 (ArSK/ CCK2 + M) and S4 (water + M).