Figure 8 The spiking (a, b) and nonspiking (c) telson-uropod stretch receptors compared. Left column, ventral views of dissected tailfans to show position of the ATU muscle and the receptors in: a, M. quadrispina (Galatheidae); b, B. occidentalis (Albuneidae); and c, E. analoga (Hippidae). On the right side, the area of dorsal telson from which the ATU muscle fibers arise is shown; compare to position of AT muscle in crayfish (Figure 5b). Arrows point to dorsal attachment of receptor strand on inner surface of telson. Middle column, transverse sections of the receptor nerve midway between the receptor strand and the mixed nerve from the lateral telson which it joins (for nerve homologies, see Paul et al., 1985). Sample recordings of the response to stretch of the receptor strand: extracellular receptor nerve recordings from M. quadrispina and B. occidentalis; intra-cellular recording from one of the four nonspiking sensory neurons (NSRs) of E. analoga (lower trace: stretch stimulus). Right column, camera lucida drawings of abdominal ganglion 6 with silver-intensified, CoCl-backfilled neurons from sensory nerves in the three species. On the right side, fills from close to the receptor strand show only the sensory neurons filled; on the left side, backfilled farther from the receptor, show a few other neurons, including the pair of ATU motoneurons with caudal somata easily identified in all species (purple arrow) also filled. Adapted from Paul, D. H. and Wilson, L. J. 1994. Replacement of an inherited stretch receptor by a newly evolved stretch receptor in hippid sand crabs. J. Comp. Neurol. 350, 150-160.

of the uropod with respect to the body axis, i.e., the movement produced by the new RS muscle. Its sensory neurons are unusual both morphologically in having centrally located cell bodies and physiologically in that they are nonspiking (NSR, nonspiking stretch receptor). Nonspiking neurons transmit only graded potentials and are incapable of generating action potentials (Figure 8c; Paul and Bruner, 1999). TUSRs evolved twice The discovery of TUSRs in albuneid sand crabs and squat lobsters in a roughly similar position as the nonspiking TUSR in hippids should not have been a surprise, because, although these animals retain the ancestral behavior of tailflipping, their uropod articulation allows greater freedom of movement than that in crayfish (Paul et al., 1985). Like the hippid TUSR, the albu-neid and galatheid TUSRs arise from the dorsal telson and insert on the ventral-medial rim of the uropod propodite. In each of the three groups, the receptor flanks the major uropod depressor muscles, the anterior telson-uropodalis muscle (ATU), which is considerably larger in these Anomala than its homologue in crayfish (Figures 5 and 8). The hippid receptor strand, however, arises on the lateral side of the ATU, whereas the receptor strand in the two tailflipping anomalans arises adjacent to the medial-anterior face of this muscle. Also, the sensory neurons of the latter are spiking (generate action potentials).

No comparable proprioceptor monitoring movements of the basal joint of the uropod has been found in any macruran or paguroid. This suggests that the greater freedom of movement of the uro-pods relative to the rest of the tail in sand crabs and galatheids was not only accompanied by alterations in neuromusculature, as discussed above, but made desirable the evolution of a proprioceptor that could monitor whole limb movement and, therefore, mediate stabilizing reflexes in uropod motoneurons. In fact, all three TUSRs are ideally positioned to sense elevation of the propodite, and in vitro experiments show that they mediate similar stabilizing (negative feedback) reflexes (Paul and Wilson, 1994); in vivo, however, the roles of the TUSRs are more complex (see Section

The most unusual features of all three TUSRs are the morphology of their sensory neurons (they are monopolar) and the central location of the sensory somata (within the sixth abdominal ganglion) (Figure 8; Maitland et al., 1982; Paul and Wilson, 1994). This contrasts with typical arthropod mechanosensory neurons, including most proprio-ceptors, which are bipolar or multipolar and have peripherally located cell bodies. The TUSR sensory neurons are similar to those of stretch receptors associated with crayfish swimmerets and crayfish and brachyuran crab walking legs and all resemble motoneurons in general structure (Bush, 1976).

Given the phyletic relationships of these animals (Figure 2) and the comparative data on their neuromusculature (Figure 5; Paul et al., 1985), it was reasonable to have assumed that spiking TUSRs were the precursor of the NSR in hippids. Why, then, are the hippid and the albuneid/galatheid receptors not homologues?

The clues are the different positions in the sixth abdominal ganglion of the sensory neuron somata and the orientations of the primary neurites relative to common landmarks, such as conserved (homologous) motoneurons and axon tracts in A6. The central features of the galatheid and albu-neid spiking sensory neurons are almost identical, more similar even than their peripheral morphologies, but they are strikingly different from the central positions and morphology of the nonspik-ing sensory neurons, NSRs, in hippids (Figure 8, right column). These central differences stand out against the background of conserved ganglionic structure and lead to the conclusion that TUSRs evolved (at least) twice, once in the common ancestor of squat lobsters and sand crabs and once in hippids (Figures 2 and 6). The extraordinary nonspiking TUSR of hippid sand crabs Why replace the spiking telson-uropod receptor that appeared in the common ancestor of sand crabs and squat lobsters with a nonspiking receptor across the same joint that mediates similar resistance reflexes (Figures 2 and 8)? The answer may be that the demise of the spiking receptor was a byproduct of the elongation of the telson in the hippid lineage (Figure 5). This shifted to longitudinal the orientation of the muscles, thereby giving them the mechanical leverage to execute large excursions of the uropods during swimming and digging, when both the load on the uropods and the variability of that load are likely to be greater. A more interesting question is why are the neurons of the hippid receptor nonspiking? What is the adaptive value of graded rather than action potential signaling in proprioceptors? Nonspiking cells are common in places where conduction distances are short; also, the speed of information transfer is higher than for action potential transmission (as in visual and auditory centers), but the former is untrue for the NSRs and the latter is an unlikely requirement for stretch receptors responding to whole limb movement. Combining neurobehavioral data on the motor system and what is known about how the NSRs mediate their diverse functions (next section) with data from the much better understood sensory and motor systems of crayfish suggests that these new neurons were wired into ancestral circuitry serving the tailfan. Nonspiking neurons are prominent in this and other neural networks (references in Paul, 2004). The evolution of the NSRs' physiological properties might have been determined by their being interconnected with nonspiking neurons in the central pattern generator (CPG) circuitry for the uropods; that is, proprioceptive feedback transmitted by graded-potential neurons could have been favored for smoother integration by a network using graded potentials to perform its neurobehavioral functions (Paul and Bruner, 1999; Paul, 2004; see Section Whatever its evolutionary cause, having nonspiking membranes may have been influential in the evolution of these sensory neurons'

extraordinary peripheral morphology (Figure 9), which is thought to explain the unusual combination in single neurons of high sensitivity and ability to respond over the entire range of tension experienced by the receptor strand (Figure 9; Wilson and Paul, 1990; Paul and Wilson, 1994).

Relaxed Stretched

Figure 9 The NSRs of E. analoga (Hippidae) have distinctive central (a1, b1) and peripheral (b2, b3, d) morphology. a1 and a2, Silver-intensified backfills showing two of the four sensory neurons (a1) and the two relatively large ATU motoneurons (a2; see Figure 8c), each in one hemiganglion; note the general similarity between the sensory and motor neuron structures. b1, An NSR filled with Lucifer yellow by microelectrode injection. b2, The peripheral dendrites of the NSRs terminate linearly along the elastic receptor strand (uropod end toward top of page) in which their dendritic tips extend longitudinally for a short distance. b3, A peripheral end of a Lucifer yellow-filled NSR dendrite penetrating the elastic receptor strand (edges of strand marked by white lines). c, Cartoon cut-away view from the right side of the tailfan with uropod in rest position (see Figure 1), showing the receptor strand (red line) to which the NSRs attach, which is stretched during the uropod return stroke (arrow). The inset shows a sample electromyogram pattern recorded in freely swimming E. analoga. The motor patterns during swimming and treading water differ; the treading water motor pattern is dependent on the presence of intact NSRs, whereas the swimming pattern is unchanged after bilateral receptor ablation (Paul, 1976). d, The ultrastructure of the NSR dendrites within the elastic strand may explain these sensory neurons' unusual combination of features: (1) high sensitivity to stretch of the receptor strand and (2) ability to respond over the full range of tensions experienced by the whole stretch receptor. d1, The dendrite of each NSR gives rise to approximately 21 000, 0.1 mm-diameter tips oriented longitudinally within the receptor strand. d2, Cross-sectional profiles of these tips are larger and less numerous in stretched than relaxed receptors. e, Model to illustrate how differential compressibility of the extracellular matrix surrounding the dendritic tips could account for the change in tip profiles shown in (d2) by producing hydrostatic forces that would cause concomitant constriction of the distal portions and expansion of the proximal portions of the tips (and open stretch-activated ion channels); LS, longitudinal section; TS, transverse section profiles of tips. The number of tips contributing to the receptor potential recorded outside of the strand would increase with increasing stretch amplitude (full range fractionation by each sensory neuron). Below, diagram showing the difference in shape and dimensions of the elastic receptor strand between the relaxed and fully stretched states. d, Adapted from Wilson, L. J. and Paul, D. H. 1990. Functional morphology of the telson-uropod stretch receptor in the sand crab Emerita analoga. J. Comp. Neurol. 296, 343-358.

Was this article helpful?

0 0

Post a comment