Graded Potentials and Spiking in Single Units of the Oval Organ, a Mechanoreceptor in the Lobsterventilatory System Iii. Sensory Habituation to Repetitive Stimulation

In vivo the oval organ is subjected continually to a cycle of stretch and relaxation as the scaphognathite beats. This paper describes the diminution of response which occurs in vitro when the isolated oval organ is stimulated repetitively with pull stimuli imitating the natural situation. Both components of the response are affected, but the decline in spiking is more pronounced than that of the graded potential.

A decline in response during a train of identical stimuli has been noted in studies on other mechanoreceptors (Catton, 1961; Hunt & Ottoson, 1976; Bush & Roberts, 1971), but has not been adequately named. A term is needed which differentiates the phenomenon both from classical adaptation which occurs during a single maintained stimulus, and from postsynaptic responses to incoming repetitive presynaptic events. ‘Sensory habituation’ is proposed, and defined as a diminution of the peripheral responses of primary sensory units to trains of repetitive, identical stimuli.

Preparative procedures, mechanical stimulation and microelectrode methods were similar to those described in the first paper of this series (Pasztor & Bush, 1983). In brief, controlled stretch stimuli were delivered to the oval organ from a loudspeaker-driven puller, via a hook inserted into the ventral attachment of the connective tissue array. In these experiments, trains of repetitive trapezoidal or sinusoidal pulls were used, with pull frequencies imitating natural respiratory rhythms (0·5–2·5 Hz). Intracellular recordings taken with 3 M-KCl-filled glass micropipettes were either filmed or printed out on a Brush recorder on line, or taped for subsequent analysis. Recording loci were 3–6 mm central to the confluence of the sensory dendrites in afferent fibres measuring 12–15 mm from oval organ to the suboesophageal ganglion.

When single stretch stimuli of constant parameters were delivered to an oval organ at intervals of 1 min or more, the responses were remarkably consistent, both in graded potential magnitude and shape, and in spiking pattern. If the stimulus repetition rate was increased, however, the responses to the second and subsequent presentations declined. The example in Fig. 1 shows Y fibre ‘s responses to a series of trapezoidal pulls delivered at 0·66 Hz. Initially there were 19 spikes attaining a maximum firing frequency of 38 Hz. At the second stimulus there were six spikes less, and lower firing frequencies throughout the response. Finally, after 10 identical stimuli the response stabilized at seven spikes, with a maximum frequency of 19 Hz. In all trials, rest periods of at least 55 s were required to restore the full response.

When a lower amplitude or shorter duration stimulus was used, which only elicited a brief burst in the fully rested preparation, as in the top two panels of Fig. 2A, B, spiking often ceased altogether during repetitive presentations, leaving all, or a large proportion, of the responses without impulses. Only the graded analogue component of the response was consistently retained, although this too showed some evidence of decline.

By superimposing a series of responses to repetitive stimuli, as in Fig. 3B, it can be seen that sensory habituation involves several distinct effects which are not all ex pressed equally in each series. (1) A decline in the rate of rise in the graded potential (also seen in Fig. 2). (2) A decline in amplitude of the graded potential (especially noticeable in X). (3) A decline in the rate of growth of the active process leading to spike initiation. (4) Finally, there may be a rise in spiking threshold. Together, these effects contribute not only to the diminution in numbers and frequency of spikes mentioned above, but in increasing delay in the appearance of the first spike and the abolition of the initial dynamic burst.

In these responses, recorded several millimetres proximal to the confluence of the sensory dendrites, it is difficult to ascertain whether sensory habituation is mainly a function of the transducer membrane or of the spike initiating zone. Fig. 4A shows a series of responses of a TTX-treated Y fibre to a train of identical sinusoidal stimuli at 2 Hz. The second response is 1 mV less than the first, but the decline is not nearly so marked nor progressive as in the same fibre before TTX-treatment. In this fibre therefore, it appears that sensory habituation is mainly, but not exclusively, a resul tant of changes in the TTX-sensitive ion channels.

Additional support is given to this notion by the responses to depolarizing current shown in Fig. 4C. Since the current was injected 6·8 mm from the confluence of the sensory dendrites, it is unlikely that the transducer membrane contributed to the response recorded. The first current injection brought the spike initiation zone rapidly to the threshold for active responses, and it responded with a sharply rising pre-potential leading to impulses. But in response to subsequent identical depolariz ing pulses delivered at approximately 1·5-s intervals, the pre-potential slope became more and more gradual, and spike initiation was delayed and finally abolished.

The X and Z fibre responses shown in Fig. 3 were recorded concurrently in res ponse to a 1-Hz trapezoidal pull. As sensory habituation developed in fibre Z, spiking was not only reduced in number and frequency, but delayed, and the peak of the receptor potential shifted to the end of the pull. Since the X response declined only in amplitude, not timing, the pair of responses underwent a significant phase delay. This can be discerned very clearly in the superimposed responses.

Fig. 5 gives several excerpts (recorded at a slower paper speed) from a long record ing of the same preparation as in Fig. 3. At the beginning of the series the resting length (l₀) of the oval organ strands was adjusted to ensure an immediate onset of depolarization at the start of a stretch stimulus. In this preparation habituation led the cessation of spiking in both X and Z in response to 0·5-mm pulls presented at 1 Hz The last sample was taken from a later portion of the same stimulus train after l₀ had been increased by 0·2 mm. Now the habituated responses of X retained one spike per response and a larger graded potential. Z responses were also larger and retained a higher ratio of spiking to non-spiking responses. Thus, even though habituation leads to diminished responses during repetitive stimulation, the fibres are still capable of encoding stimulus parameters such as extent of stretch.

In the habituated state, small transient perturbations in stretch, introduced manu ally, produced highly significant modulation of the responses. Small increments in background stretch temporarily restored spiking and amplitude to the responses, while small decrements abolished spiking and reduced the graded potentials markedly (Fig. 5).

During gill ventilation in Crustacea, long term recordings from muscles and nerves of the spontaneously active scaphognathite have shown that beating maintains con stant parameters over many cycles (Pasztor, 1968; Pilkington & Simmers, 1973; Young, 1975; Best, 1982; Simmers & Bush, 1983a). When these conditions were mimicked experimentally, by presenting repetitive pulls to the isolated oval organ, the responses of its three afferent fibres habituated. Using moderate amplitude, sinusoidal stretch stimuli, spiking diminished and often disappeared, leaving the afferent fibres responding solely in the non-impulsive mode. This implies that under normal conditions of undisturbed ventilation, feedback information can be provided by small oscillations in membrane potential without the intervention of spikes.

Fibre X has the least labile response during habituation, and though spiking may disappear, the graded response retains a steep dynamic component so that the ampli tude maximum undergoes the least phase shift of the three fibres. X is strongly implicated in the role of beat marker and is able to continue signalling beat frequency throughout long periods of regular ventilation.

Water flow through the gills however is not invariant, and several authors have demonstrated changes in stroke volume, scaphognathite reversals, sudden frequency changes and sporadic apnoeic periods (Hughes, Knights & Scammel, 1969; Wilkens & Young, 1975; Best, 1982; Simmers & Bush, 1983b). Sometimes these are reflex responses to noxious elements in the respiratory current, at other times these varia tions in rhythm appear to be spontaneous. Any such perturbations in the regular cycle of stretch and relaxation of the oval organ would disrupt habituation and restore spiking. Temporary obstructions, which might be caused by detritus lodging in the pumping chamber, forcing the scaphognathite into an abnormal configuration, would cause the transient appearance of spikes for a few cycles, while the commonly obser ved respiratory pauses are of sufficient duration to permit the restoration of the fully non-habituated response. Hyperventilation involving a faster, more powerful stroke, could be expected to elicit trains of bursts, with prolific spiking at any change in rhythm.

Thus the graded potentials would predominate during bouts of regular beating, while spiking would provide an enhanced release of transmitter at any perturbation in rhythm. This interplay between the two methods of signalling would seem to be an effective means of discriminating between a regular and changing ventilatory activity. Such immediacy could hardly be attained by either the analogue or the impulsive mode alone.

Sensory habituation is likely to occur in any receptor subject to rhythmical stimula tion at moderate or fast repetition rates. Insect song patterns, for example, frequently provide such a stimulus situation, and Esch, Huber & Wohlers (1980), recording from single primary auditory units of Gryllus campestris, noted declining responses during groups of sound syllables. There was a significant decrease in the number of spikes per response with a corresponding increase in latency. The 325 ms silent period between groups was sufficient for complete recovery of the receptors before the arrival of the first syllable of the next group. Sensory habitation has also been reported in Pacinian corpuscles (Lowenstein, 1961), frog tactile receptors (Catton, 1961), crab T-C MRO (Bush & Roberts, 1971; Cannone & Bush, 1980) and cat muscle spindles (Hunt & Ottoson, 1976). All these receptors would receive repetitive stimulation during rhythmical locomotion.

Gill withdrawal in Aplysia is perhaps the best known example of behavioural habituation to have been analysed at the cellular level (reviewed by Kandel & Schwartz, 1982). Byrne, Castellucci, Carew & Kandel (1978) have shown conclusive ly that the mechanoreceptors mediating the withdrawal reflex show no decrement in afferent discharge at stimulation rates producing habituation. Thus, in contrast to the peripheral mechanism described for the oval organ and the other examples listed above, in Aplysia the progressive decline in motor output to the gills during repeated skin stimulation is a function of central synapses between sensory neurones and interneurones. A similar situation obtains in the crayfish escape reflex (Zucker, 1972).

Our study has established an example of habituation occurring at the receptor level which is amenable to intracellular investigation. The significance of sensory habitua tion and the consequent reduction in feedback to the respiratory pattern generator is currently under investigation.

This work was supported by research grants to BMHB from the Royal Society and the Science Research Council (U.K.) and to VMP from the Natural Sciences and Engineering Research Council of Canada. We thank Alison Walford for technical assistance.