A. Nose Touch Avoidance

When animals collide nose-on with an object such as an eyelash they respond by initiating backward movement (Croll 1976; Kaplan and Horvitz 1993). Three classes of mechanosensory neurons (ASH, FLP, and OLQ) act in parallel to mediate this avoidance response (see Fig. 2A for ASH circuitry diagram). These nose touch neurons have cell bodies in the nerve ring ganglia and send long dendritic processes to the tip of the worm's nose. The response to nose touch is quantitative (normal animals respond in 90% of trials), and each sensory neuron class accounts for a fraction of the normal response, as follows: ASH, 45%; FLP, 29%; and OLQ, 5% (Kaplan and Horvitz 1993). The remaining responses (∼10%) are mediated by the ALM and AVM neurons, which sense anterior body touch (J. Kaplan, unpubl.). It is unclear what distinguishes the function of the three nose touch neurons. One attractive possibility is that these cells differ in their sensitivities and that the intensities of nose touch stimuli vary according to the violence of the collision. If this were the case, it would be expected that the most sensitive neuron (ASH) would account for the majority of responses and the less sensitive neurons (FLP and OLQ) would account for the remainder.

Figure 2

Mechanosensory circuits. (Ovals) Sensory neurons; (rectangles) interneurons; (inverted triangles) motor neurons. Cells that express glr-1 (more...)

The mechanosensory function of the ASH neurons is surprising because these cells were thought to act solely as chemosensory neurons. The ASH neurons are part of the amphid chemosensory organs, their sensory endings are exposed to the external environment, and they have two chemosensory functions mediating avoidance of osmotic and volatile repellents (Bargmann et al. 1990; Troemel et al. 1995; see Bargmann and Mori, this volume). Thus, the ASH neurons are polymodal sensory neurons. Several classes of chemosensory neurons respond to multiple chemical stimuli in C. elegans; however, ASH is unique among them in responding to such divergent stimuli. In this respect, ASH neurons are similar to vertebrate neurons that sense painful stimuli, which are called nociceptors (Besson and Chaouch 1987).

ASH and FLP provide direct synaptic input and OLQ provides indirect synaptic input (via RIC) to the AVA, AVB, and AVD interneurons (Fig. 2A) (White et al. 1986). Genetic evidence supports the function of these synaptic connections. The glr-1 gene encodes an AMPA-type glutamate receptor that is expressed in both the forward (AVB and PVC) and backing (AVA and AVD) interneurons (Hart et al. 1995; Maricq et al. 1995). glr-1 mutants are defective for nose touch avoidance, which suggests that the sensory transmitter is glutamate and that glr-1 encodes a subunit of the postsynaptic glutamate receptors at the nose touch neuron-to-interneuron synapses. Analysis of genetic mosaics revealed that GLR-1 receptors function in both AB.a and AB.p lineages, suggesting that GLR-1 receptors are required in both the forward and backing interneurons (which are derived from AB.p and AB.a, respectively) and that the forward interneurons also have a role in backward movement (Hart et al. 1995).

GLR-1 glutamate receptors appear to play a part in distinguishing between the two ASH sensory modalities. glr-1 mutants are defective for ASH-mediated touch sensitivity but are normal for ASH-mediated osmotic and volatile-repellent sensitivity (Hart et al. 1995; Maricq et al. 1995). Although they are specifically required for ASH-mediated touch sensitivity, GLR-1 receptors are expressed in synaptic targets of ASH. These results suggest that the ASH neurons produce different synaptic signals in response to the two stimuli. Consistent with this idea, ASH synaptic termini contain two distinct kinds of synaptic vesicles, clear and dense core vesicles, implying that ASH neurons utilize two distinct neurotransmitters (White et al. 1986). Separate signaling pathways for the two ASH sensory modalities are also supported by the phenotype of osm-10 mutants, which are defective for ASH-mediated osmotic avoidance but are normal for ASH-mediated nose touch sensitivity (A. Hart and J. Kaplan, unpubl.). Thus, differential signaling at the ASH-to-interneuron synapses may allow animals to distinguish between the ASH sensory modalities.

B. Foraging and Head Withdrawal

Touch also regulates movement of the animal's head, which is called foraging behavior. Foraging behavior consists of continuous, apparently exploratory, head movements. The patterns of head movements that occur during foraging are complex compared to the relatively simple sinusoidal movements that underlie locomotion. This difference is explained by the motor anatomy of the head. Head muscles are divided into eight radial symmetric sectors, and these are independently innervated by ten classes of motor neurons (Ware et al. 1975; White et al. 1986). As a consequence of this motor circuitry, worms can move their head through 360°.

When worms are touched on either the dorsal or ventral sides of their nose with an eyelash, they interrupt the normal pattern of foraging and undergo an aversive head-withdrawal reflex. This simple reflex is mediated by two classes of mechanosensory neurons (OLQ and IL1) and their synaptic targets, the RMD motor neurons (Fig. 2B) (Hart et al. 1995). Killing any of these cells, alone or in combination, diminishes the head withdrawal reflex. Here again, the sensory neurons have nonredundant functions, OLQ accounting for the majority of the normal responses. Interestingly, IL1, OLQ, and RMD also regulate spontaneous foraging movements (Hart et al. 1995). Laser operated animals lacking IL1 and OLQ forage abnormally slowly and make exaggerated dorsal and ventral nose turns. These results suggest that mechanosensory stimuli (touch or stretch) regulate both the rate and pattern of spontaneous foraging movements. The function of the sensory neuron-to-RMD synapses is supported by the phenotype of glr-1 mutants (Hart et al. 1995). The RMD neurons express GLR-1 glutamate receptors and glr-1 mutants are defective for the head withdrawal reflex.

C. Genetic Analysis of Nose Touch Avoidance and Foraging

The mechanosensory neurons that mediate nose touch avoidance and control foraging behavior have sensory endings that contain a single cilium (Ward et al. 1975; Albert et al. 1981; Perkins et al. 1986). Mutations in several genes (e.g., che-3 and osm-1 ) disrupt the ultrastructure of ciliated sensory endings (Lewis and Hodgkin 1977; Albert et al. 1981; Perkins et al. 1986). These cilium-structure mutants are all defective for nose touch avoidance and head withdrawal and have abnormally slow foraging behavior (Kaplan and Horvitz 1993 and unpubl.). These results suggest that the nose touch neurons act as sensory neurons (rather than interneurons) in these behaviors.

Mutations in two genes ( daf-11 and eat-4 ) disrupt the function of nose touch neurons and other ciliated sensory neurons. daf-11 mutants are defective for the nose touch avoidance response (J. Kaplan, unpubl.). The daf-11 gene encodes a membrane-bound guanylate cyclase (D. Birnby and J. Thomas, pers. comm.) that is also required for chemotaxis and for the regulation of dauer formation by ciliated chemosensory neurons. eat-4 mutants are defective for nose touch avoidance and head withdrawal and have an abnormal pattern of foraging movements (A. Hart and J. Kaplan, unpubl.). As discussed below (see also Avery and Thomas, this volume), several lines of evidence suggest that eat-4 mutants are generally defective for glutamatergic neurotransmission, again implicating glutamate as the sensory transmitter.

D. Response to Gentle Body Touch

1. The Neuronal Circuit

Animals respond to gentle touch (typically delivered with an eyelash) all along the length of their body (Chalfie and Sulston 1981; Chalfie et al. 1985). Anterior touch elicits backward movement, whereas posterior touch elicits forward movement. Gentle body touch is sensed by the mechanosensory neurons ALML/R, AVM, and PLML/R (Fig. 3A). These mechanosensory neurons were previously called the microtubule cells because their long axonal processes are filled with bundles of distinctive large-diameter microtubules (Fig. 3B) (Chalfie and Thomson 1979). Several observations support the conclusion that these touch cell processes, which run longitudinally along the body wall, function in mechanoreception (Chalfie and Sulston 1981; Chalfie et al. 1985). First, the touch cell processes lack synaptic specializations and hence are likely to be dendritic. Second, the touch cell processes are embedded in the hypodermis adjacent to the cuticle, a position expected to facilitate detection of mechanical stimuli. Third, the position of the processes along the body axis correlates with the sensory field of the touch cell. ALM and AVM have sensory receptor processes in the anterior half of the body, and ablation of these cells eliminates anterior touch sensitivity. Likewise, PLM has a posterior dendritic process, and ablation eliminates posterior touch sensitivity. The PVM neuron does not appear to have a critical role in touch sensitivity as it cannot mediate a touch response by itself. However, PVM is also considered a touch cell because it is ultrastructurally similar to the other touch cells, and its differentiation is controlled by the same genetic pathway (see below).

Figure 3

The touch receptor neurons. (A) Positions and general structure (White et al. 1986). Cell bodies of the embryonically derived ALM cells are situated laterally; (more...)

The touch cell processes enter the neuropile and form connections with many other neurons (Chalfie et al. 1985; White et al. 1986). ALM and AVM axons enter the nerve ring, whereas PLM and PVM axons enter the ventral nerve cord (Fig. 3A). The touch cells provide direct input to the interneurons that control locomotion; however, the pattern of these connections is highly asymmetric. The anterior touch cells ALM and AVM form gap junctions with the backward movement interneuron AVD, but they provide synaptic input to the forward interneurons (AVB and PVC). Conversely, PLMR provides input to the forward interneuron PVC via gap junctions, but it provides synaptic input to the backward interneurons AVA and AVD (Fig. 2C). Thus, the touch cells form gap junctions with agonist interneurons and chemical synapses with the antagonist interneurons. This reciprocal pattern of connectivities led to the proposal that the gap junctions are excitatory and the synaptic connections are inhibitory (Chalfie et al. 1985). The excitatory function of these gap junctions is supported by cell-killing experiments. Killing AVD in embryos eliminates anterior touch sensitivity in young larvae, whereas killing PVC eliminates posterior touch sensitivity. Analysis of the tap response (described below) provides some support for the inhibitory function of the touch cell-to-interneuron synapses (Wicks and Rankin 1995).

E. Circuitry for the Tap Response

As mentioned above, the anterior and posterior touch circuits are interconnected in two ways: (1) Touch cells make reciprocal connections to the opposing classes of interneurons (described above) and (2) the forward and backward interneurons form reciprocal connections (see Fig. 2C). These interconnections suggest that the anterior and posterior touch circuits functionally interact, perhaps allowing integration of opposing mechanosensory inputs. The capacity of these circuits for integration is beginning to be understood through the analysis of the tap response. Worms respond to a diffuse mechanical stimulus (a tap to the side of the dish on which they are resting) either by accelerating forward movement or by initiating backward movement (Chiba and Rankin 1990; Rankin 1991). Given that the stimulus is not spatially coherent and that the animal's response is variable, it was proposed that the tap response reflects the simultaneous activation of the anterior and posterior touch cells. The cellular basis of the tap response has been analyzed extensively (Wicks and Rankin 1995; see Jorgensen and Rankin, this volume). Since the tap response can be quantitated (as prevalence or magnitude of accelerations vs. reversals), relatively subtle effects of cell killing can be detected. As predicted, both the anterior (ALM and AVM) and posterior (PLM) touch cells, and their interneuron targets (AVD and PVC), contribute to the tap response, the anterior cells promoting reversals and the posterior cells promoting accelerations (Fig. 2E). Disabling either the anterior or posterior touch circuits results in exaggeration of the opposing response. For example, animals lacking PLM neurons respond to tap solely with reversals, and these reversals are of greater magnitude than those of unoperated controls. Similarly, animals lacking the PVC interneurons (which mediate accelerations) always respond to tap by reversing, although the magnitude of their reversals is indistinguishable from that of unoperated controls. These results show that the anterior and posterior touch circuits functionally inhibit each other. Furthermore, since killing PLM and PVC produced distinct phenotypes, these results also suggest that PLM neurons make functional connections other than the gap junctions with PVC. One attractive possibility is that PLM inhibits the magnitude of reversals via chemical synapses with the backing interneurons AVA and AVD.

Although the tap response can be thought of as a competition between the anterior and posterior touch circuits, there also are neurons that promote the activities of both circuits. The PVD and DVA neurons are presynaptic to both the forward and backing interneurons, and animals lacking these neurons respond to tap with diminished accelerations and reversals. Therefore, PVD and DVA appear to maintain the overall activity of the touch circuit. Since PVD neurons are mechanosensory (Way and Chalfie 1989), it is possible that the excitability of the touch circuit is modulated by mechanical stimuli. Thus, the tap response comprises the simultaneous activation of anterior and posterior touch cells, with the behavioral outcome being determined by the integration of these two antagonistic circuits (see Jorgensen and Rankin, this volume).

The tendency of animals to respond to tap with accelerations versus reversals varies over the course of development. Accelerations predominate in larvae and reversals predominate in adults (Chiba and Rankin 1990). Because this developmental switch occurs in young adults (i.e., 46−50 hours after hatching), it has been proposed that this change reflects the formation of functional connections by the AVM neuron. However, adults lacking AVM do not behave like larvae, as they respond to tap by reversing much more often than accelerating (Wicks and Rankin 1995). Therefore, this developmental switch must reflect more than the addition of AVM neurons to the circuit—perhaps the addition of the PVD neurons contributes to the developmental switch.

F. Circuitry for the Response to Harsh Touch

Although animals lacking functional touch cells are insensitive to touch with an eyelash, they remain sensitive to prodding with a platinum wire (typically responding by undergoing backward movement) (Chalfie and Sulston 1981). This result suggested that a separate mechanosensory circuit mediates sensitivity to harsh touch stimuli (Fig. 2D). The PVD neurons are thought to be harsh touch sensory neurons for several reasons. First, the PVD neurons have long undifferentiated processes that run along the lateral body wall, which could be mechanosensory (White et al. 1986; E. Hedgecock, pers. comm.). Second, the PVD neurons express genes involved in touch cell differentiation (e.g., mec-3 , see below), implying that they may also be mechanosensory (Way and Chalfie 1989). Third, killing the PVD neurons in animals that lack touch cell function eliminates harsh touch sensitivity (Way and Chalfie 1989). The locomotion interneurons AVA and PVC are direct synaptic targets of PVD. Mutants lacking GLR-1 glutamate receptors (which are expressed by the locomotory interneurons) are insensitive to harsh touch, which suggests that these synapses are functional and that glutamate is the PVD transmitter (Hart et al. 1995).