Giant neurons and the crayfish tail flip

Crayfish escape from the strike of a predator by means of a rapidly executed tail flip, produced by flexing and re-extending the whole abdomen. The abdomen is able to act as an effective locomotory organ because the last two (sixth and seventh) abdominal segments are modified to form the tail fan (Fig. 3.1a). A single flip of the tail fan is capable of moving the animal several centimetres through the water. The power for this movement is provided by the fast flexor muscles, which occupy much of the space within each abdominal segment. These are called fast muscles because they produce rapid twitch contractions, in contrast to a set of much smaller, slow muscles that produce graded postural movements of the abdomen. In addition, each abdominal segment contains fast and slow extensor muscles, and these are also much less substantial than the flexors.

The innervation of these muscles was first studied by Keis Wiersma (1947), who showed that the giant neurons are involved in the control of these muscles during a tail flip. About ten large motor neurons innervate the flexor muscles on each side of each abdominal segment. One of these is an exceptionally large motor neuron, called the motor giant, which sends an axon branch to every fast flexor muscle fibre. Another is an inhibitory motor neuron that also innervates every muscle fibre. The remaining motor neurons are simply known as fast flexor neurons, and each of these innervates only a localised group of fibres within the fast flexor muscles. During a tail flip, the motor giant and the fast flexor motor neurons are excited via two pairs of large interneurons, the lateral and medial giant interneurons (Fig. 3.1b, c). The control exerted by these two giant interneurons is so strong that a single spike in either a lateral or a medial giant interneuron is sufficient to trigger a tail flip.

Both of these giant interneurons extend along much of the length of the central nervous system but they differ considerably in structure. The medial giants have their cell bodies and dendrites in the brain, where they receive sensory input, and their axons extend down to the last abdominal ganglion. In contrast, the lateral giants are segmentally repeated structures, formed from separate cells linked end to end. Each segment contains a cell body, dendrites and a length of axon that abuts against the corresponding axon in the next segment. Where the axons abut, there is a segmental synapse between two successive lateral giants (see Fig. 3.3a). The lateral giants receive input only in the abdominal segments. At first, it was thought that both kinds of giant interneurons initiate the same behaviour pattern because a spike in either of them produces a rapid flexion of the abdomen. However, a more detailed series of studies, using a combination of neuro-physiological and high-speed filming techniques, showed that the lateral and medial giants initiate different patterns of behaviour (Fig. 3.2).

Activation of the medial giants elicits contraction of the fast flexor muscles in all abdominal segments, and this produces a uniform curling of the abdomen that propels the animal straight backwards (Fig. 3.2a). Activation of the lateral giants elicits flexor contraction in the anterior segments of the abdomen but not in the posterior segments; the latter remain

Figure 3.2 The different kinds of tail flip produced by the medial and lateral giant interneurons. The precise pattern of movement is correlated with activity in the giants by filming animals with electrodes implanted (as in Figure 3.1a). Tracings from these high-speed films show (a) a medial giant flip elicited by a tap on the head, and (b) a lateral giant flip elicited by a tap on the abdomen. (c) Diagram showing the pattern of synaptic connections between the two giants (horizontal lines) and the motor giant neurons (vertical lines) in the abdominal segments. Direct synaptic connections are represented by filled circles and the absence of synaptic connections is indicated by an asterisk. (From Wine, 1984.)

Figure 3.2 The different kinds of tail flip produced by the medial and lateral giant interneurons. The precise pattern of movement is correlated with activity in the giants by filming animals with electrodes implanted (as in Figure 3.1a). Tracings from these high-speed films show (a) a medial giant flip elicited by a tap on the head, and (b) a lateral giant flip elicited by a tap on the abdomen. (c) Diagram showing the pattern of synaptic connections between the two giants (horizontal lines) and the motor giant neurons (vertical lines) in the abdominal segments. Direct synaptic connections are represented by filled circles and the absence of synaptic connections is indicated by an asterisk. (From Wine, 1984.)

straight and so cause the thrust to be directed mainly downwards, thereby pitching the animal forwards (Fig. 3.2b). These two kinds of movements are well adapted to the different sorts of stimuli that excite the two types of giant interneuron. The lateral giants are triggered only by sudden mechanical stimuli that originate posteriorly, such as a sharp tap to the abdomen, and a lateral giant flip appropriately carries the animal in an anterior direction. Similarly, the medial giants are triggered only by stimuli applied to the head region and a medial giant flip carries the animal in a backward direction. In this way, each kind of tail flip removes the animal from the source of the stimulus.

The differences between the two kinds of tail flip can be explained by differences in the synaptic connections between the giant interneurons and the motor giant neurons in the abdominal segments (Fig. 3.2c). At the

Crayfish Anatomy

Figure 3.3 Electrical synapses between neurons involved in crayfish startle behaviour. (a) Drawing of an abdominal ganglion to show the relative positions of the lateral giant and giant motor neurons, the synapse between them, and the arrangement for recording from each side of this synapse. The segmental synapse between successive lateral giants is also shown. (b) Simultaneous intracellular recordings from the lateral giant and motor giant neurons close to the synapse, demonstrating the negligible delay due to electrical transmission. (c) Intracellular recording from the lateral giant at a point close to the synapse. Following electrical stimulation of the appropriate sensory neurons, a compound EPSP and a spike are recorded in the lateral giant with a very small delay. The two components of the EPSP (a and |3) are produced by separate pathways from the sensory neurons, as described in the text. (a and b modified after Furshpan & Potter, 1959; c redrawn after Krasne, 1969.)

Figure 3.3 Electrical synapses between neurons involved in crayfish startle behaviour. (a) Drawing of an abdominal ganglion to show the relative positions of the lateral giant and giant motor neurons, the synapse between them, and the arrangement for recording from each side of this synapse. The segmental synapse between successive lateral giants is also shown. (b) Simultaneous intracellular recordings from the lateral giant and motor giant neurons close to the synapse, demonstrating the negligible delay due to electrical transmission. (c) Intracellular recording from the lateral giant at a point close to the synapse. Following electrical stimulation of the appropriate sensory neurons, a compound EPSP and a spike are recorded in the lateral giant with a very small delay. The two components of the EPSP (a and |3) are produced by separate pathways from the sensory neurons, as described in the text. (a and b modified after Furshpan & Potter, 1959; c redrawn after Krasne, 1969.)

point of synaptic contact, the motor giant branches over the surface of the interneuron's axon in a characteristic manner, which is readily recognised when the motor axons are filled with intracellular dye (Fig. 3.3 a). In the anterior segments of the abdomen, both lateral and medial giants receive these synaptic branches from the motor giants, but in the posterior segments the synaptic branches to the lateral giant are clearly missing, whereas those to the medial giant are present. This distribution of synapses has been confirmed by testing for postsynaptic responses by recording with a microelectrode: responses to medial giant activity can be obtained in all abdominal segments, but responses to lateral giant activity are only obtained in the anterior segments. In the thorax, the situation is reversed. The medial giant makes no output connections to motor giants, but the lateral giant does connect in the more posterior segments (Heitler & Fraser, 1993). Contraction of flexor muscles in the posterior part of the thorax helps bend the body into the jackknife shape that propels it rapidly forwards. Hence, the consistent difference in the pattern of abdominal flexion is brought about by differences in the synaptic connections between the respective controlling interneurons and a shared motor output system.

Another behavioural difference, noticed early in the study of crayfish startle behaviour, is that sometimes only a single tail flip occurs, while at other times an apparently similar stimulus evokes a whole series of tail flips in rapid succession. The latter are produced by alternating flexions and extensions of the abdomen, repeated at a frequency of 10 to 20 Hz, and this behaviour is termed escape swimming. Studying animals with implanted electrodes reveals that escape swimming does not involve the giants, which are active only before the first tail flip. The fast flexor muscles of the abdomen are used in swimming, and they are controlled by the fast flexor motor neurons but not by the motor giants and the giants. Escape swimming can be triggered on its own, without an initial giant-mediated flip, by stimuli that are weaker or have a slower onset compared with those which trigger the giant interneurons. Although a sharp stimulus can evoke a single giant-mediated flip without subsequent swimming, it is more usual for the giant-mediated flip to be followed by non-giant swimming.

The tail flips generated in escape swimming fall into three relatively stereotyped classes: linear flips, which resemble medial giant flips; pitching flips, which resemble lateral giant flips; and twisting flips, which tend to rotate the animal about its longitudinal axis as well as having a backward component. During swimming, these classes of tail flips are arranged in sequences that result in the animal being propelled backwards away from the original stimulus. Following a medial giant response, swimming involves only linear flips, which simply continue the backward movement away from the threat at the head end. A lateral giant response is followed by one or two pitching flips, which turn the animal in a complete somersault so that it lands on its back with its head facing the stimulus. Then two or three twisting flips turn the animal dorsal side up again, and finally a series of linear flips carry it backwards away from the stimulus (Cooke & Macmillan, 1985).

In addition, the swimming system is able to act upon directional information in the stimulus that is ignored by the giant system. The lateral giants, for example, always generate a bilaterally symmetrical response that carries the animal straight forward, regardless of whether the stimulus comes directly from behind or from one side. However, the path followed by subsequent swimming has a lateral component that steers the animal away from a stimulus delivered to one side of the abdomen (Reichert & Wine, 1983). All these results show that escape swimming is well adapted to exploit the initial advantage gained by a giant-mediated tail flip. The following account focuses on the lateral giant response and its relation to subsequent swimming because this system has been studied in most detail.

+1 0

Post a comment