Bilateral asymmetry of the paired claws of the lobster Homarus americanus is determined during the fourth and fifth juvenile stages by differential reflex activity; the side with the greater activity becomes the crusher while the contralateral side becomes the cutter. Juvenile lobsters reared during this critical period with a substratum that could not be grasped or with reduced input from predominantly internal mechanoreceptors (proprioceptors) (achieved by cutting the dactyl and its chordotonal organ or by tenotomizing the claw opener or closer muscles) failed to develop a crusher claw and hence remained bilaterally symmetrical: they developed paired cutter claws. Therefore, the proprioceptive component of the reflex activity is implicated in bringing about the initial lateralization of the claw ganglion into a crusher and a cutter side.
Moreover, lobsters with a single claw reared without a substratum developed a crusher on the intact side only if the intact claw was exercised. In the unexercised condition, differences in reflex activity between the side with a claw and the side without one were insufficient for the development of a crusher claw on the intact side. A minimal amount of reflex activity is necessary for the development of a crusher. Lobsters reared with this minimal amount of activity in both claws developed asymmetrical claws rather than paired crusher claws. This means that initial lateralization of the claw ganglion into a crusher side, on a random basis, inhibited the opposite side from also becoming a crusher. This would explain why we failed to produce lobsters with paired crusher claws and why they were seldom found in the wild.
Many vertebrates and invertebrates are bilaterally symmetrical: the right side of the body is a mirror image of the left side. Often, however, structural or functional asymmetry is superimposed upon this basic plan. Some of the better known examples are handedness and cerebral lateralization in humans (Sperry, 1974), neural mechanisms underlying vocalization in song birds (Nottebohm, 1984), torsion of body parts in gastropod molluscs and claw laterality in crustaceans (Barnes, 1968). Amongst crustaceans, the differentiation of the paired claws into a major and a minor type provides some of the most flamboyant examples of bilateral asymmetry, e.g. male fiddler crabs have an enlarged and elaborate major or ‘fiddle’ claw, which is used in a characteristic waving pattern during sexual displays, and a diminutive minor claw, which is used in feeding (Crane, 1975). Alpheid snapping shrimps have a major ‘snapper’ claw, which makes a popping sound and ejects a jet of water during agonistic displays, and a minor ‘pincer’ claw, which is used for feeding and burrowing (Hazlett and Winn, 1962). The paired claws in the American lobster, Homarus americanas, are also bilaterally asymmetrical, consisting of a major and a minor claw (Herrick, 1895). The major or crusher claw is stout, molar-toothed and closes its dactyl against its pollex slowly (>100 ms) but with enough force to crack open shells of bivalves (Scrivener, 1971; Govind and Lang, 1974). The minor or cutter claw is slender, incisor-toothed and closes rapidly enough (<20ms) to catch free-swimming fish. These differences in closing behaviour are due principally to corresponding differences in the fibre composition of the paired closer muscles; the crusher muscle has 100% slow fibres while the cutter muscle has predominantly fast fibres with a small ventral band of slow fibres (Lang et al. 1977).
The development of asymmetry in the paired claws and their closer muscles has been extensively studied (reviewed by Govind, 1984). The paired claws are small, undifferentiated and symmetrical in larval and early juvenile lobsters (Herrick, 1895, 1911). Determination of cutter and crusher claws occurs some time during the juvenile fourth and fifth stages (Emmel, 1908) and the claws and closer muscles (Ogonowski et al. 1980) develop into their respective types in subsequent juvenile stages. Once claw laterality has been determined in these critical juvenile stages, it is fixed for life.
We have attempted to isolate the factors that determine claw laterality during the critical juvenile period by performing various manipulations. The presence of a graspable substratum during the fourth and fifth stages ensures the development of paired asymmetrical (cutter and crusher) claws while the lack of such a substratum leads to the development of paired symmetrical (cutter type) claws (Lang et al. 1978). This led to the hypothesis that the more active claw became the crusher and its less active counterpart the cutter. Restricting the use of a claw by immobilization did not prevent the claw from developing into a crusher, however, but curtailing activity of the muscles, by tenotomy of either the closer or opener muscles or by curtailing nerve activity by denervation, did effectively prevent the claw from becoming a crusher (Govind and Kent, 1982). These results demonstrated that the lack of nerve-mediated muscle tension prevented the claw from developing into a crusher. The corollary to this would be to promote the development of a crusher claw by imposing activity on a particular claw. Indeed, this proved to be the case when, in the absence of a graspable substratum, one of the paired claws was repeatedly stimulated to grip the bristles of a paintbrush (i.e. exercised): it developed into a crusher claw while its counterpart became a cutter (Govind and Pearce, 1986). However, when both claws were exercised, neither developed as a crusher; instead, both developed as cutters. These results showed that nerve activity by itself did not determine which claw became a crusher, but that the difference in reflex activity between the two sides predisposed the side with the greater activity to become a crusher and the opposite side to become a cutter. Since such bilateral differences are initially detected in the first thoracic ganglion serving the claws, presumably claw laterality is determined in the ganglion, i.e. the thoracic ganglion becomes lateralized into a crusher and cutter side and this lateralization is subsequently expressed at the periphery (Govind, 1989).
These experiments not only showed that differential reflex activity in the paired claws is an important factor in determining claw asymmetry but also that such activity is generated peripherally. Hence, the proximate factor in determining laterality is the sensory component and specifically that associated with movements of the claw, i.e. mechanoreceptors. A large variety of mechanoreceptors is found in crustaceans and these may be broadly classified into internal, cuticular and supracuticular (Bush and Laverack, 1982). The supracuticular receptors with end-organs or accessory structures projecting beyond the cuticle comprise setae, campaniform sensilla and articulated pegs, whereas the cuticular receptors comprise those located within the cuticle, either in the hypodermis or in connective tissue. In contrast to these external mechanoreceptors, located wholly within the exoskelton are internal mechanoreceptors or proprioceptors comprising muscle receptor organs, apodeme receptors, chordotonal organs and innervated strands. The occurrence and distribution of mechanoreceptors in crustacean claws is poorly understood. In the claws of the lobster Homarus americanus the morphology and distribution of setae has been described (Solon and Cobb, 1980) as have the axon number and composition of the chordotonal organ spanning the propus-dactyl joint, i.e. the PD chordotonal organ (Cooper and Govind, 1991). This organ contains the endings of movement- and position-sensitive cells embedded in an elastic strand, which spans the joint by attaching to the dactyl at one end and to the apodeme of the closer muscle at the opposite end. Recordings from axons of PD organs in crabs have shown that they are sensitive to length and tension changes in the elastic strand brought about by movements of the dactyl (Wiersma and Boettiger, 1959). By monitoring movements of the dactyl brought about passively or actively by muscle contraction, the PD organ provides a major source of proprioceptive input.
In the present experiments, we have attempted to determine whether external or internal mechanoreceptors are essential in determining bilateral asymmetry of the paired claws in developing lobsters and whether there is a minimal level of reflex activity required to bring about the effect. Results from these experiments aid in extending our current hypothesis for the determination of claw laterality in developing lobsters.
MATERIALS AND METHODS
Postembryonic development of the American lobster (Homarus americanus Milne-Edwards) consists of three larval stages (first, second and third) and several juvenile stages beginning with the fourth stage. Larval third-stage lobsters were obtained from the Massachusetts State Lobster Hatchery on Martha’s Vineyard and held communally in our rearing facilities at the Marine Biological Laboratory, Woods Hole. After the moult into the fourth stage, the animals were held individually in plastic trays (Lang, 1975) in order to monitor their moult cycle, allow experimental manipulation of the claws and prevent internecine encounters. Lobsters were not able to grip the smooth walls of these plastic trays but could easily grip small (approximate dimensions 6 mm × 3 mm × 1mm) pieces of broken oyster shells, 8–10 of which were provided as a substratum. When the juvenile lobsters had developed to the eighth or ninth stage, the claws had clearly differentiated into their final type, i.e. cutter or crusher, and the closer muscle had correspondingly established its fibre composition (Govind and Kent, 1982). At this time, we visually assessed the configuration of the paired claws in all lobsters and measured the fibre composition of the closer muscle in a few representative examples using histochemical techniques (Ogonowski et al. 1980). In all the experiments, the paired claws and their closer muscles were either of the symmetrical (cutter/cutter) type or of the asymmetrical (cutter/crusher) type (Fig. 1).
All manipulations of the claws were performed within 24 h after the animal had moulted. Claws were removed by a gentle pinch, which induced autotomy at a preformed fracture plane. Within a few days following claw removal, a blastema from which a limb bud later differentiates forms at this site. At the next moult, the limb bud emerges as a newly regenerated claw. Muscle tenotomies were performed under the microscope by cutting the tendon at its distal insertion onto the dactyl with a sharpened pin (Fig. 2). Proprioceptive input caused by movements of the dactyl was eliminated by cutting the dactyl where it articulates with the propus to ensure that the chordotonal organ spanning the propus-dactyl joint was sectioned (PD chordotonal organ in Fig. 2). Reflex activation of the claw, i.e. exercise, consisted of holding the lobster and gently stroking its claw with a small paintbrush so that the claw gripped the bristles several times in a 60s session. There were three such sessions each day separated by at least 5h. This treatment was maintained over the entire fourth and fifth stages. Reflex activation, therefore, consists of the opening and closing of the claw.
Differences in reflex activity between the paired claws in the critical fourth and fifth stages were instrumental in determining claw laterality in juvenile lobsters: the side with the greater activity became the crusher while the opposite side became the cutter (Govind and Pearce, 1986). Since these differences in reflex activity determine laterality, initially in the ganglion, the sensory component of the reflex activity is strongly implicated. Moreover, since the reflex activity was brought about by stroking the claw with a paintbrush, a mechanoreceptive sensory input is implicated. Such input could include a variety of mechanoreceptors, both internal (proprioceptors) and external (cuticular and supracuticular) (Bush and Laverack, 1982). Hence, we designed an experiment in which either predominantly external mechanoreceptors or both external and internal mechanoreceptors of the claw would be activated. One of the paired claws was stroked with a paintbrush for three daily 1 min bouts. The stroking was either vigorous or gentle. Vigorous stroking elicited closing and opening reflexes, thereby exercising the claw and presumably activating both external and internal mechanoreceptors. Gentle stroking did not elicit claw closing or opening, so presumably only external mechanoreceptors were stimulated (the possibility that there was some muscle contraction without visible movements of the dactyl cannot be ruled out entirely in these experiments). Exercise of the left claw (vigorous stroking) resulted in a majority of these animals developing a crusher on that side, whereas the majority of the controls in which the paired claws were untouched developed paired cutter claws (Table 1). Gentle stroking of the left claw (without exercise) also failed to cause the claws to differentiate a crusher. Paired cutter claws developed much as in the control condition.
The results of another experiment in which mainly external mechanoreceptors were stimulated are given in Table 2. Here a more indirect method was used to ensure that external mechanoreceptors, but not necessarily internal mechanoreceptors (proprioceptors), were stimulated by providing a substratum which could not be gripped by the claws. Several (6–8) solid plastic spheres approximately 3 mm in diameter were provided in each rearing tray during the fourth and fifth stages. The spheres were smooth and of a sufficiently large size that the claws were not able to grip them. For one group of lobsters, the plastic spheres were free-standing objects while for another group they were glued to the bottom of the tray in random positions. In both cases the majority of lobsters failed to develop a crusher and instead developed paired cutter claws. This was significantly different from the control condition in which lobsters reared with oyster chips developed asymmetrical claws. Oyster chips are small bits of broken oyster shells easily gripped by the juvenile lobsters.
The above experiments suggest that stimulation of external mechanoreceptors is not a sufficient condition for the determination of claw asymmetry and point to the possibility that input from internal mechanoreceptors is critical. Since one of the major sources of such proprioceptive input is the PD chordotonal organ, this receptor was eliminated in one of the paired claws in lobsters reared with a substratum (Table 3, series 1). This was done simply by cutting off the dactyl where it articulates with the propus. In this group, the majority of lobsters developed a crusher on the intact side while the treated side developed into a cutter. For comparison, in a control group of lobsters with paired intact claws the crusher appeared on either side. Therefore, sectioning of the PD organ together with removal of the dactyl appears to prevent the development of a crusher claw on the treated side. In an earlier experiment, cutting off most of the dactyl but leaving the PD organ intact did not prevent that claw from developing into a crusher (Govind and Kent, 1982). Presumably, input from the PD organ is crucial in the determination of asymmetry, although there may well be other receptors at this joint which could contribute to the effect.
A more stringent test for the role of proprioceptive input was to eliminate this input in a claw destined to become a crusher. This can be achieved by rearing lobsters with oyster chips and removing one of the claws; the remaining one always develops into a crusher (Emmel, 1908). Using this protocol, we reduced proprioceptive input in one of several ways in the remaining claw (Table 3, series 2). The dactyl was removed at its articulation with the propus so that the PD organ would be inactivated or the claw opener and closer muscles were tenotomized in separate experiments, since they bring about movements of the dactyl. Each of these treatments effectively prevented the development of a crusher claw, even though that claw was predisposed to become one. In all three experiments, the majority developed paired cutter claws. In controls, the majority developed a crusher claw on the intact side.
Minimal reflex activity
Our previous finding that bilateral differences in reflex activity between the paired claws were necessary in determining asymmetry suggested that perhaps any difference in activity may be sufficient, rather than the three daily bouts of exercise imposed on one of the claws (Govind and Pearce, 1986). A simple test of this possibility was to rear juvenile lobsters without oyster chips (as this ensured the development of paired cutter claws) and at the same time to remove one of the claws to ensure differences in activity between the two sides (Table 4, series 1). The results of this experiment were unequivocal: all but one lobster failed to develop a crusher claw; the remainder developed paired cutter claws. This was similar to the control condition in which lobsters were also reared without a substratum but had both claws intact.
Clearly, bilateral differences in reflex activity on the two sides in these lobsters with a claw missing were insufficient to trigger asymmetry. These results implicate a minimal level of activity for the determination of a crusher claw. Presumably, this minimal level was achieved when one of the paired claws was exercised in lobsters reared without a substratum (Govind and Pearce, 1986). Consequently, for our next experiment we reared juvenile lobsters without a substratum, removed the left claw and exercised the intact right claw (Table 4, series 2). Exercise entailed picking up the animal and stroking the claw with a paintbrush so that its bristles were gripped several times in a l min period. A second group of lobsters, in which the animals were handled but not exercised, served as controls. In the control animals, both the intact and regenerated claws became cutters. In contrast, a significant number of lobsters in which the intact claw was exercised developed a crusher on that side. Since the two groups of lobsters were reared under identical conditions, the exercise regime adopted in the present experiments exceeds a minimal level of reflex activity necessary for the determination of asymmetry.
Bilateral application of minimal activity
Our finding that a minimal level of reflex activity is necessary for the determination of a crusher claw led us to query whether it is possible to apply this minimal level on both sides and to develop paired crusher claws. To test this possibility we not only reared lobsters with oyster chips but exercised their claws as well. Rearing with oyster chips would ensure the development of paired asymmetrical claws with the crusher appearing either on the right or left side, while exercising would ensure that the other claw, the putative cutter, would also receive the minimal level for the development of a crusher. In the experiment in which only one of the paired claws was exercised (left one in Table 5) our intent was to stimulate minimal activity in the left putative cutter claw which would occur in 50% of the lobsters reared with oyster chips. In exercising both claws, our intent was to ensure that the putative cutter on either side experienced minimal reflex activity. In both these experiments lobsters developed bilateral asymmetry with an approximately equal distribution of the crusher on the right or left side (Table 5). In no case did a crusher develop on both sides. These results were similar to those of a control group in which the lobsters were handled but not exercised. While a minimal level of reflex activity is required for the determination of a crusher on one side, the imposition of this activity on the opposite side does not result in the development of another crusher. Presumably, the determination of a crusher on one side also ensures that the opposite side becomes a cutter.
Sensory components of reflex activity
Reflexive activity involving closing and opening of the dactyl during the critical juvenile stages is required for determining laterality in the paired claws of the lobster (Govind and Pearce, 1986). In the present experiments, we have attempted to isolate neural components within this activity which play a crucial role. Since laterality is initially determined in the ganglion based on reflexive activity of the claws, the simplest and most economical explanation is that sensory components of the reflex activity bring about this lateralization. Although this does not rule out the possibility of a role for the motor component, this would be a more complex method for determining laterality. Hence, the present experiments were designed to isolate components of the sensory activity and specifically to distinguish between external and internal mechanoreceptors. The evidence suggests that input from internal mechanoreceptors (proprioceptors) is necessary to determine laterality, particularly as its reduction suppressed crusher development in a claw otherwise predisposed to become a crusher. Presumably, input from largely internal mechanoreceptors from the paired claws serves to lateralize the first thoracic ganglion into a crusher and cutter side. Such lateralization is subsequently transmitted, via unknown pathways, to the periphery, resulting in the differentiation of the claws into crusher and cutter types.
How mechanosensory input may lateralize the ganglion is unknown, nor is the nature of such lateralization known. However, in male fiddler crabs with markedly asymmetrical claws, the hemiganglion on the side of the major claw is larger than its contralateral minor counterpart (Young and Govind, 1983). In alpheid snapping shrimps, which also have markedly asymmetrical claws, the somata of motoneurones innervating the closer muscle of the major claw are larger than those innervating its minor counterpart (Mellon et al. 1980). Moreover, there are many more axon profiles in the limb nerve to the major hemiganglion compared to its minor counterpart (Govind and Pearce, 1988). Structural asymmetries in the nervous system are also found amongst the vertebrates; for example, in the central nervous system of mammals (Galaburda, 1984) and singing birds (Nottebohm, 1984).
Minimal reflex activity
Having previously established that bilateral differences in reflex activity were essential in promoting claw laterality (Govind and Pearce, 1986), we now find that this reflex activity must be above some minimal level. This is because lobsters in which differences in reflex activity between the two sides are achieved by removing one of the claws still do not develop asymmetrical claws when raised without a substratum. In contrast, exercising one of the paired claws results in the development of asymmetry even though the lobsters are raised without a substratum. The threshold level for the development of claw asymmetry therefore lies somewhere between the levels represented by these two experiments. Although the precise level has not been determined in the present experiments, the concept of a threshold level has been uncovered.
When threshold activity levels were applied to both claws of a lobster, they did not result in the development of two crusher claws, but in a crusher and a cutter claw. This finding is not surprising, as threshold activity levels to both claws are likely to occur in the wild and yet lobsters with paired crusher claws are rarely found. In a survey of 2433 wild lobsters, Herrick (1895) found not a single individual with paired crusher claws; the overwhelming majority were asymmetrical and only three individuals had paired cutter claws. However, lobsters with paired crusher claws have been reported from the wild (Calman, 1906; Herrick, 1907). More recently, we were able to study two such individuals (Govind and Lang, 1979) in which, although the paired claws were of the crusher morphotype, one of the pair was capable of closing its claw with the speed of a cutter claw. Examination of the closer muscle of this unusual crusher claw revealed a substantial (40%) proportion of fast fibres; a normal crusher claw has 100% slow fibres (Govind, 1984). Thus, although masquerading as a crusher claw in external form, in behaviour and muscle composition it more closely resembled a cutter.
Hypothesis for determining claw laterality
These findings, together with our earlier studies, allow us to postulate how bilateral asymmetry of the paired claws is determined in developing lobsters. During the critical period represented by the fourth and fifth juvenile stages, when the previously planktonic larval stages begin to adopt a benthic habit (Botero and Atema, 1982; Cobb et al. 1989), initial use or contact of one claw with the substratum sets in motion an increasingly greater use of that claw. The greater neural input on that side lateralizes the hemiganglion to a crusher type and at the same time inhibits its opposite counterpart from also becoming a crusher. In this way bilateral asymmetry is assured. A useful analogy for determination of claw laterality would be a see-saw, which remains balanced in a horizontal plane when the forces are equal at both ends. For the lobster, this would represent the condition of paired cutter claws, which arises when lack of a graspable substratum limits activity on both sides to subthreshold levels, or when both claws are exercised, thereby equalizing neural input. Differential forces at the two ends of the see-saw will lower one end while elevating the other end. This is equivalent to differences in neural input promoting determination of a crusher at one end and a cutter at the other end. According to this model, the determination of paired crusher claws is nearly impossible unless the connecting rod between the two ends of the see-saw is broken. In other words, the paired hemiganglia would have to be functionally disconnected. Perhaps, in rare instances, partial disconnection may occur, resulting in those unusual lobsters in which the paired claws are both of the crusher type in structure but not in function (Govind and Lang, 1979).
A hypothesis for claw laterality in lobsters should also be able to account for the distribution of laterality in the wild. As mentioned above, the majority of lobsters have asymmetrical claws with an equal distribution of the crusher on the right or left side, a small fraction have symmetrical claws of the cutter type and none have the symmetrical condition of true crusher claws (Herrick, 1895). Since initial contact with the substratum or use of the claw would occur on a random basis, there is an equal probability of the crusher appearing on either side of the body.
The likelihood of encountering a graspable substratum is high as the changeover from a planktonic to a benthic habit takes place during the fourth and fifth juvenile stages (Botero and Atema, 1982; Cobb et al. 1989). Thus, most lobsters would develop a crusher and a cutter claw. A few may fail to develop a crusher claw and instead develop paired cutter claws because minimal activity levels were not encountered.
We thank Mike Syslow and Kevin Johnson of the State Lobster Hatchery, on Martha’s Vineyard, for generous supplies of larval lobsters, Robin Cooper for technical assistance and Harold Atwood for critical comments. Financial support was provided by the Natural Sciences and Engineering Research Council of Canada.