Motor neurons in the escape response circuit of white shrimp (Litopenaeus setiferus)

Methods

Live white shrimp, Litopenaeus setiferus (Linnaeus, 1767), fished from waters around South Padre Island, Texas, were purchased from commercial seafood stores in Port Isabel, Texas and housed in aquaria. Individuals were anaesthetized by chilling on ice and dissected in physiological saline. The abdominal nerve cord was removed.

Neurons were backfilled (Pitman, Tweedle & Cohen, 1972; Quicke & Brace, 1979; Altman & Tyrer, 1980; Jones & Page, 1983). The nerve containing the neurons of interest was placed in a well of petroleum jelly containing 0.3 M solution of either nickel chloride or cobalt chloride, while the remaining tissue was bathed in physiological saline (mM: 410 NaCl, 12.7 KCl, 10.3 CaCl₂, 10 MgCl₂, and 14 Na₂SO₄, 10 tris[hydroxymethyl]aminomethane (Trizma Base); pH adjusted to pH 7.4). The tissue was incubated in a refrigerator for 7–18 h, precipitated with ammonium sulfide or dithiooxamide (a.k.a. rubeanic acid; this term is used hereafter), fixed in 4% formalin in saline, dehydrated with a progressive ethanol series (70% for 10 min, 90% for 10 min, 100% for 10 min, and 100% again for 5 min), and cleared in methyl salicylate. When precipitated with ammonium sulfide, neurons containing cobalt or nickel ions turn black or gray. When precipitated with rubeanic acid, neurons containing cobalt ions turn yellow, and those containing nickel ions turn blue (Quicke & Brace, 1979; Jones & Page, 1983). Neurons containing some mixture of the two ions turn an intermediate colour, ranging from dark orange to purplish-red (Quicke & Brace, 1979; Jones & Page, 1983).

The third nerve (N3) was filled 42 times in 30 abdominal ganglia of 14 individuals. The second nerve (N2) was filled 98 times in 61 ganglia of 23 individuals. Abdominal ganglia 1 through 5 were filled, although most fills were of the anterior four ganglia. Because backfills are often incomplete (Altman & Tyrer, 1980), the number of cells reported is the maximum number of cells seen across multiple individuals.

Cleared backfills were viewed on an Olympus CX41 microscope, and photographed using an attached Olympus C-5050Zoom digital camera. Images were assembled into final figures using Corel Photo-Paint 12. Some large images were stitched together from multiple photographs.

Results

Fast flexor motor neurons

The non-MoG fast flexor cell bodies are found in three clusters (Fig. 1), as in other decapods (Mittenthal & Wine, 1978). The flexor medial contralateral (FMC) cluster is contralateral and anterior of the filled N3 (in the terminology of Mittenthal and Wine, “posterior” refers to the position of the axon relative to the cell body). The flexor posterior ipsilateral (FPI) cluster is ipslateral and anterior of the filled N3. The flexor anterior contralateral (FAC) cluster of cell bodies is contralateral and posterior to the filled N3. As in other species (Mittenthal & Wine, 1978), there is segmental variation in the number of cell bodies in each ganglion, with the more posterior showing the greatest deviation (Table 1). White shrimp have one or two fewer cell bodies in each cluster than most other decapods examined to date (Table 2). The FMC cell bodies are more widely separated in L. setiferus than crayfish, with one anterior of the MoG cell body and near the midline, and the other more posterior and lateral of the MoG cell body. Although this separation means these two cells would not normally be described as being in a “cluster,” the FMC in other species is rarely a tight grouping of cell bodies. Fast flexor cell bodies are often pairs or singletons, depending on the number, somewhat separated from other cells in the cluster; e.g., Figs. 3A and 3B in Espinoza et al. (2006); see also Mittenthal & Wine (1978).

Table 1:

Number of fast flexor motor neurons in each abdominal ganglion of L. setiferus.

Abdominal ganglia(on)	FMC (Non-MoG)	MoG	FPI	FAC
A1–4	3	1	3	1
A5	3	1	2	0
A6	?	?	?	0

DOI: 10.7717/peerj.1112/table-1

Notes:

FMC

flexor medial contralateral

MoG

motor giant fast flexor motor neuron

FPI

flexor posterior ipsilateral

FAC

flexor anterior contralateral

Table 2:

Number of fast flexor motor neurons in abdominal ganglion 2 of different species.

Species	FMC (non MoG)	MoG	FPI	FAC
White shrimp (Litopenaeus setiferus)	2	1	3	1
Spiny lobster (Panulirus argus)a	3	0	4	3
Louisiana red swamp crayfish (Procambarus clarkii)b	3	1	4	3
American clawed lobster (Homarus americanus)e	3	1f	4	3
Squat lobster (Galathea strigosa)c	4	0	4	2
Squat lobster (Munida quadrispina)d	3	0	4	0

DOI: 10.7717/peerj.1112/table-2

Notes:

FMC

flexor medial contralateral

MoG

motor giant fast flexor motor neuron

FPI

flexor posterior ipsilateral

FAC

flexor anterior contralateral

a Espinoza et al., 2006.

b Mittenthal & Wine, 1978.

c Sillar & Heitler, 1985a.

d Wilson & Paul, 1987.

e Otsuka, Kravitz & Potter, 1967.

f See discussion in Mittenthal & Wine, 1978.

The MoG cell bodies in L. setiferus are extremely large, have a variegated appearance, irregular shape, and press closely together so that they look like one large mass covering much of the ventral surface of the abdominal ganglion (Fig. 2, Video S1). They are not two widely separate, bilateral, spherical cell bodies reported in caridean prawns (Holmes, 1942). In L. setiferus, each MoG is ∼300 µm across the ventral surface of the ganglion, and about 100 µm when viewed in the sagittal plane. Other fast flexor motor neurons in L. setiferus are ∼50–100 µm in diameter.

Two lines of evidence show the MoG axons are fused bilaterally. First, when the N3 of one side is filled with cobalt chloride and the other N3 is filled with nickel chloride and precipitated with rubeanic acid, the MoG cell bodies are dark red, clearly distinct from the blue and yellow of the other cells filled by a single nerve (Figs. 1, 2A and 2B), indicating they are filled with a mixture of both chemicals. Second, a single medial axon is visible in the cord between ganglia, which bifurcates and leaves both the left and right N3 (Fig. 3). Although the MoG cell bodies are pressed so close together to be sometimes indistinguishable as two cell bodies (e.g., Fig. 2B), two axons emerge from the MoG cell bodies (Fig. 2A), project dorsally a short distance within the ganglion before fusing as previously reported (Johnson, 1924; Holmes, 1942), continuing into the nerve cord as one axon until slightly anterior of the point where N3 exits the body, when it bifurcates and sends a branch both left and right (Figs. 3B and 3C).

The fast flexor motor axons appear slightly “haloed” (Fig. 3C) compared to the smooth axons seen filled though N2 (Fig. 4).

Discussion

The fast flexor motor giant neurons (MoGs) in Litopenaeus setiferus have a structure unlike that reported for any other decapod crustacean. Two aspects of the structure of the MoGs suggest that they may be syncytial cells formed by the fusion of many small neurons. First, they are larger than any other fast flexor motor neurons in this or other species, which are usually ∼100 µm in diameter (Otsuka, Kravitz & Potter, 1967; Mittenthal & Wine, 1978; Sillar & Heitler, 1985a; Wilson & Paul, 1987; Espinoza et al., 2006). Second, they are not spherical as most other neuron cell bodies are. Third, their variegated appearance suggests they have a different internal structure than other neurons. The hypothesized syncytial structure is reminiscent of the third-order giant neurons in squid stellate ganglia, which are also syncytial cells formed from many cell bodies, and also involved in an escape response (Young, 1936; Young, 1939). Annelid worms also have syncytial giant neurons (Nicol, 1948; Günther, 1975) that are involved in escape responses (Nicol, 1948; Günther, 1975; O’Gara, Vining & Drewes, 1982).

Although the size and hypothesized fusion in the MoG cell body is unexpected, it is consistent with the long-known fusion of the MoG axons in other caridean shrimp species (Johnson, 1924; Holmes, 1942; Friedlander & Levinthal, 1982). Given that there are genetic mechanisms to fuse the MoG axons during development (Friedlander & Levinthal, 1982), the same mechanisms could be used to fuse cell bodies. Reduction of fusion appears to be an evolutionary trend in the decapods, starting with hypothesized fusion of MoG cell bodies (this study) and axons in dendrobranchiates (this study; Xu & Terakawa, 1999), to fusion of the MoG axons only in carideans (Holmes, 1942), to no MoG fusion in reptantians (Mittenthal & Wine, 1978). A prediction of this fused cell body hypothesis is that the MoG cell bodies would contain multiple nuclei. Because backfills rarely reveal any subcellular structure, other techniques, such as thin sectioning and staining for higher resolution microscopy, will be needed to test this hypothesis.

The remaining fast flexor and fast extensor motor neurons appear to be found in homologous positions to the better studied reptantian species. In almost every case, there are fewer cell bodies in L. setiferus than in the homologous groups of neurons in most reptantian decapods. If other non-reptantian species have similarly small numbers of motor neurons, it would suggest that duplication of fast abdominal motor neurons occurred during decapod evolution. The loss or reduction of the massive MoG cell bodies may be correlated with the increased number of fast abdominal motor neurons in reptantians: the amount of ganglionic volume consumed by the MoGs may have constrained the addition of any new fast motor neurons.

The “haloed” appearance of the fast flexors, but not fast extensors, may be indicative of myelination. In Palaemon serratus, the MoG axons are myelinated in the periphery, and references to other axons of similar size being myelinated suggest the other fast flexor motor neurons are also myelinated (Holmes, 1942).

The smaller number of fast abdominal neurons might indicate that shrimp have less fine grained control over the fast flexor motor muscles than crayfish. In many reptantian crustaceans, the fast flexor muscles are used in two distinct forms of tailflipping. The MoGs and other fast flexor neurons are used in single stereotyped escape tailflips triggered by giant interneurons. The MoGs are not involved in repetitive variable tailflipping, which is controlled by an undescribed system of non-giant interneurons (Reichert, Wine & Hagiwara, 1981; Reichert & Wine, 1983; Sillar & Heitler, 1985b; Wilson & Paul, 1987; Faulkes, 2004); non-giant tailflipping would be generated by the remaining pool of fast flexor motor neurons. A larger pool of motor neurons may allow for some of the fine control necessary for such variability. Previously, I suggested that non-giant tailflipping originated at the base of the decapod clade (Faulkes, 2008), but I did not express this hypothesis as tentatively as it should have been. It is not known if non-reptantian decapods have variable, non-giant tailflipping behaviour like many reptantians do. Indeed, the myelination of the entire population of fast flexors and the axonal fusion of the MoGs point to a circuit specialized for explosive starts. It may be that tailflipping cannot occur without activity of the giant neurons in shrimp, and that a small pool of neurons is sufficient to generate shrimps’ more consistently explosive tailflips. Alternately, the variation in motor neuron number may be trivial and have little functional impact, because crustacean muscles generally have sparse polyneuronal innervation. In Munida quadrispina, the FAC cluster of fast flexor motor neurons was lost with no visible change in tailflipping behaviour (Wilson & Paul, 1987).

Supplemental Information