ABSTRACT
The metathoracic anterior coxal adductor (a.c.a.) muscle of the locust and the grasshopper is innervated by a peripheral inhibitory axon similar to the inhibitory axon which innervates the metathoracic extensor tibiae muscles of these insects. No evidence was found to justify calling this axon an inhibitory-conditioning axon.
Hyperpolarizing inhibitory postsynaptic potentials (IPSPs) are normally recorded from a.c.a. muscle fibres during stimulation of this axon, and if the bathing medium contains a high concentration of potassium ions the tonic fibres of the a.c.a. muscle relax slightly during inhibitory stimulation.
The IPSPs are chloride potentials and can be converted to depolarizing responses by changing either the external or internal chloride concentration of the a.c.a. muscle fibres. Depolarizing IPSPs are frequently accompanied by small contractions of a.c.a. muscle fibres innervated by the inhibitory axon.
The a.c.a. muscle fibres are permeable to potassium and chloride ions but influx of potassium chloride is much faster than efflux. Therefore when a.c.a. muscle fibres are loaded with chloride by exposing them to high-K saline (20−100 m-equiv. potassium/I.) and are then returned to normal (10 m-equiv. potassium/1.) saline the internal chloride concentration remains elevated for some time and during this period the equilibrium potential for the inhibitory response is less negative than the resting potential and the IPSPs are depolarizing.
Depolarizing IPSPs are usually recorded from a.c.a. muscle fibres of locusts and grasshoppers when these fibres are transferred from their normal bathing medium, haemolymph, to 10 K saline. Probably the main reason for this reversal of the IPSPs is the entry of KC1 into the muscle fibres during dissection of the nerve-muscle preparations. Large quantities of KC1 would be released into the environment surrounding these preparations from muscle fibres cut and removed during dissection.
Depolarizing IPSPs were more frequently recorded from muscle fibres of grassfed locusts than from fibres of starved locusts. The potassium concentration of haemolymph of grass fed locusts is higher than that of locust saline (10 m-equiv./l.).
The potassium concentration of locust haemolymph presumably fluctuates in vivo but these fluctuations are too slow to affect the sign of the IPSP. The IPSPs are therefore always hyperpolarizing in vivo.
The effect of changes in the potassium concentration of the bathing medium on the magnitude and polarity of the IPSP could account for the diverse responses recorded previously from a.c.a. muscle fibres of locusts and grasshoppers.
INTRODUCTION
Usherwood & Grundfest (1964, 1965) and Ikeda & Boettiger (1965a, b) have proposed that some insect muscles are controlled in part by peripheral inhibitory axons. For example, some fibres of the metathoracic extensor tibiae muscle of locusts and grasshoppers are innervated by an inhibitory axon, and the excitatory electrical responses and mechanical responses of these fibres are attenuated during stimulation of this axon (Usherwood & Grundfest, 1964, 1965). The extensor tibiae muscle is innervated by two excitatory axons, designated ‘fast’ and ‘slow’ (Hoyle, 1955a; Cerf, Grundfest, Hoyle & McCann, 1959). Only fibres innervated by the ‘slow’ excitatory axon receive endings from the inhibitory axon. Peripheral inhibition of the ‘fast’ response is therefore minimal and the main effect of inhibitory activity is on the ‘slow’ responses of the extensor tibiae muscle (Usherwood & Grundfest, 1965). The electrical response of the extensor tibiae muscle fibre to stimulation of the inhibitory axon consists of an inhibitory postsynaptic potential (IPSP), which is usually hyperpolarizing. Hyperpolarizing IPSPs have been recorded from other muscle fibres of locusts and grasshoppers (Usherwood & Grundfest, 1965), from cockroach muscle fibres (Usherwood & Grundfest, 1965; Guthrie, personal communication), from bee muscle fibres (Ikeda & Boettiger, 1965 a) and from beetle muscle fibres (Ikeda & Boettiger, 1965b). In orthopteran muscles the IPSP results from a transient increase in the chloride permeability of the muscle fibre, or more precisely the inhibitory postsynaptic component of the muscle fibre membrane (Usherwood & Grundfest, 1965). Therefore during the IPSP the membrane potential of the muscle fibre moves toward the chloride equilibrium potential (ECI) which is also the inhibitory equilibrium potential (EIPSP). γ-Aminobutyric acid (GABA) mimics the transmitter at orthopteran inhibitory neuromuscular synapses while picrotoxin blocks activation of the inhibitory postsynaptic membrane both by GABA and the inhibitory transmitter (Usherwood & Grundfest, 1964, 1965).
On the basis of these observations the occurrence of peripheral inhibition in insects seemed firmly established although the functional significance of the inhibitory axons was not established. Nevertheless two recent publications by Hoyle (1966 a, b) in which the properties of the metathoracic anterior coxal adductor (a.c.a.) muscle in locusts and grasshoppers are described, are somewhat critical of this conclusion. The a.c.a. muscle is innervated by two axons, one of which is clearly excitatory. The properties of the other axon are apparently less clearly defined since it sometimes enhances (excites) and at other times attenuates (inhibits) the contractions of the a.c.a. muscle.
In view of its apparent diverse properties Hoyle (1966a, b) has called this axon inhibitory-conditioning (I-C) rather than inhibitory. The electrical responses recorded intracellularly from fibres of the a.c.a. muscle during I-C stimulation ranged from large depolarizing postsynaptic potentials to large hyperpolarizing postsynaptic potentials. In some muscles the I-C responses were exclusively hyperpolarizing and inhibitory while in other preparations, where depolarizing I-C responses predominated, a slight enhancement of the contractions of the a.c.a. muscle in response to stimulation of the excitatory axon was recorded during I-C stimulation. In some preparations the electrical responses to I-C stimulation changed during the course of the experiment from hyperpolarizing to depolarizing potentials. Hoyle (1955b) has previously used the term ‘conditioning’ to describe one of the axons which innervates the locust metathoracic extensor tibiae muscle. Although this axon is clearly identical with the inhibitory axon described by Usherwood & Grundfest (1964, 1965), Hoyle (1955 b) failed to demonstrate its inhibitory properties. Nevertheless he did record hyperpolarizing postsynaptic potentials from fibres of the extensor tibiae muscle during stimulation of his ‘conditioning’ axon although these, did not apparently attenuate the depolarizing excitatory postsynaptic potentials (EPSPs). In fact a slight enhancement of the ‘fast’ and ‘slow’ contractions of the extensor tibiae muscle was observed during stimulation of the ‘conditioning’ axon. Hoyle (1966 b) suggested that this axon might be an evolutionary relic which has lost its primeval function of peripheral inhibition. However, it is difficult to understand how stimulation of this axon could have much effect on the ‘fast’ response of the extensor tibiae muscle since it does not innervate most of the fibres which receive endings from the ‘fast’ excitatory axon (Usherwood & Grundfest, 1965). In fact the evidence favouring an inhibitory function for this axon is so overwhelming that there seems no reason to continue using the term ‘conditioning’. However, the suggestion that the a.c.a. muscle in locusts and grasshoppers is innervated by an ‘inhibitory-conditioning’ axon rather than an inhibitory axon cannot be rejected quite so easily, and although much of the evidence for this is inconclusive the suggestion does warrant further investigation. In view of this the properties of the metathoracic a.c.a. and extensor tibiae nerve-muscle preparations have been re-examined in locusts and grasshoppers. The results of these studies indicate conclusively that the a.c.a. muscle and extensor tibiae muscle are both innervated by inhibitory axons of the type described by Usherwood & Grundfest (1964, 1965). It is possible that all the diverse responses obtained by Hoyle (1966a, b) from his a.c.a. nerve-muscle preparations can be accounted for partly by variations in the haemolymph-potassium levels in locusts and grasshoppers (Hoyle, 1953) and partly by an increase in the chloride content of the a.c.a. muscle fibres as a result of KC1 leakage from muscles removed during dissection of the a.c.a. muscle.
MATERIAL AND METHODS
In situ and isolated preparations of the metathoracic a.c.a. nerve-muscle (Hoyle, 1966a, b) and in situ preparations of the metathoracic extensor tibiae nerve-muscle (Usherwood & Grundfest, 1965) were examined in the locusts Schistocerca gregaria and Locusta migratoria and in the grasshopper Ramalea microptera. Intracellular recordings from a.c.a. and extensor tibiae muscle fibres were made with glass micro electrodes filled with either 3 M potassium chloride or potassium propionate. The recording and display equipment was standard (e.g. Katz, Webb & Sorem, 1964). Mechanical responses of the muscles were obtained using a Grass Strain Gauge (Usherwood & Grundfest, 1965).
Most preparations were bathed for periods in locust saline (10 K saline) containing lomM-KCl, 140 mM-NaCl, 2 mM-CaCl2, 6 mM-NaH2PO4, 4 mM-NaHCO8. This medium is similar to Hoyle’s (1953) locust saline except that MgCl2 was omitted. In some experiments the potassium content of the saline was increased. Salines containing an increased potassium concentration were prepared in one of two different ways: (1) potassium was substituted for sodium whilst maintaining the chloride concentration constant (154 m-equiv./l.); this type of saline will be referred to as high-K saline; (2) potassium was substituted for sodium whilst maintaining the product of the potassium and chloride concentration constant, i.e. an increase in the potassium concentration was compensated by a decrease in the chloride concentration. The impermeant anion propionate was substituted for chloride. This type of saline will be referred to as high-K (constant product) saline. All the salines were effectively isotonic and buffered at pH 6·8. The preparations were perfused continuously throughout the experiments with a constant stream of saline.
RESULTS
The main object of these studies was to determine whether the diverse inhibitory-conditioning (hereafter called inhibitory) responses obtained by Hoyle (1966a, b) from his preparations of the a.c.a. nerve-muscle of locusts and grasshoppers represent the normal responses of these muscles. As a starting-point the possibility was examined that the actual preparation of in situ and isolated a.c.a. muscles alters the responsiveness of these muscles to the inhibitory axon compared with the in vivo muscle. For example, significant changes in the ionic environment of the a.c.a. muscle could occur when the normal bathing medium of the muscle, haemolymph, is replaced by locust saline, and these changes might have implications for the response of the a.c.a. muscle fibre to inhibitory stimulation. Exposure and isolation of the a.c.a. muscle and associated nerves might also modify the inhibitory responses of the a.c.a. muscle fibres. To test these possibilities about twenty in situ a.c.a. muscles with their associated nerves were prepared and the inhibitory responses of the a.c.a. muscle fibres were examined at various stages during the dissections. The preparations were bathed in their own haemolymph at all times. Similar studies were made on six locust extensor tibiae nerve-muscle preparations.
The results obtained from these preparations were conclusive. The contractions of the a.c.a. muscles and the ‘slow’ contractions of the extensor tibiae muscles were always attenuated during inhibitory stimulation and the electrical responses to stimulation of the inhibitory axon were exclusively hyperpolarizing IPSPs (Fig. 1). Depolarizing IPSPs were never recorded from muscles bathed in haemolymph. Considerable resting tensions (c. 3 g. in the locust and c. 5 g. in the grasshopper) were often recorded from both a.c.a and extensor tibiae muscles when these muscles were bathed in haemolymph and stretched to maximal body length. The largest ‘resting’ tensions were recorded from muscles of insects fed on a diet (e.g. grass) containing a high concentration of potassium. Hoyle (1954) found a relationship between the potassium content of the haemolymph and the diet of the locust and showed that the haemolymph-potassium concentration is elevated following a meal containing a high concentration of potassium. Usherwood & Little (to be published) have recently confirmed that the potassium content of locust and grasshopper haemolymph is dependent to some extent on diet although large variations in haemolymph-potassium levels have not been recorded. Muscle resting potentials are often significantly lower in grass-fed locusts than in starved locusts. A correlation between haemolymph-potassium levels and muscle fibre membrane potentials is of course to be expected since locust muscle fibres are known to be reasonably good potassium electrodes for a wide range of potassium concentrations (Usherwood, 1967a, b). The extensor tibiae muscles in locusts and grasshoppers contain some tonic fibres (Usherwood, 1967a) and it seems reasonable to assume that the a.c.a. muscle also contains some tonic fibres since it gives a tonic contraction during potassium depolarization although in good preparations the tonic contraction is always preceded by a brief phasic response. In fact Hoyle (1966 b) has reported the occurrence of at least two types of fibre in the a.c.a. muscle. Presumably the tonic fibres respond to potassium depolarization with a maintained contracture and this could account for the abnormally large ‘resting’ tensions recorded from muscles of locusts fed on a high-potassium diet. Stimulation of the inhibitory axon to both a.c.a. (Fig. 2) and extensor tibiae muscles of grass-fed locusts transiently relaxes these muscles. Studies of free-walking locust preparations (Runion & Usherwood, 1966; Galloway, Runion & Usherwood, 1966) indicate that the inhibitory axons to both the a.c.a. and extensor tibiae muscles are active when these muscles are not being excited by the excitatory motor innervation (Usherwood & Runion, to be published). Inhibitory impulses could bring about small changes in the ‘resting’ tension of these muscles by virtue of their action on the tonic fibres especially if the haemolymph-potassium concentration is high, since the IPSPs are always hyperpolarizing and inhibitory regardless of the level of potassium in the haemolymph. Enhancement of the excitatory responses during inhibitory stimulation, like that seen by Hoyle (1966a, b) in his a.c.a. preparations, never occurred in the present experiments when the muscles were in haemolymph. It seems reasonable therefore to assume that in vivo the inhibitory axons to a.c.a. muscles and extensor tibiae muscles in locusts and grasshoppers have similar properties. The term ‘conditioning’ as defined by Hoyle (1966a, b), cannot be used to describe these axons in vivo since they never ‘condition’ the muscle to produce a larger response to an excitatory stimulus.
The effects of locust saline on the properties of the a.c.a. and extensor tibiae muscles were examined next. The electrical and mechanical responses of over 100 a.c.a. nerve-muscle preparations and about 10 extensor tibiae nerve-muscle preparations from locusts and grasshoppers were studied first in haemolymph and then more critically in locust saline (10 K saline). In haemolymph the electrical responses during inhibitory stimulation were, as before, exclusively hyperpolarizing IPSPs and were often accompanied by transient relaxations of the muscles (Fig. 3). However, when the haemolymph was replaced with 10 K saline changes in the responses to the inhibitory axon often occurred. In some preparations the IPSPs became slightly smaller and the relaxations of the muscles in response to inhibitory stimulation slowly diminished (Fig. 3). At the same time a slight fall in muscle ‘resting’ tension was recorded. These results were most frequently obtained from muscles of starved insects with low haemolymph-potassium concentrations (c. 7 m-equiv./l.). In most preparations from animals fed on a high-potassium diet, reversal of the muscle IPSPs from hyperpolarizing to depolarizing responses occurred in many fibres when the preparation was exposed to 10 K saline (Fig. 4). Reversal of the IPSPs also occurred in a few preparations from starved locusts. With the appearance of depolarizing IPSPs small contractions were frequently recorded from the muscles during inhibitory stimulation although contractions did not invariably occur. Interaction between depolarizing IPSPs and EPSPs was complex (Figs. 4, 5). On occasions the EPSP and IPSP summated to give what might appear to be an enhanced excitatory response (Fig. 5). At other times, provided the EPSP was large, the EPSP was attenuated by the depolarizing IPSP (Fig. 4). The effects of interaction between the EPSP and IPSP on the mechanical responses of the a.c.a. and extensor tibiae muscles were unpredictable since these muscles contain many fibres innervated by both excitatory and inhibitory axons and the EPSPs and IPSPs recorded from the different fibres are never identical. The ionic basis for interactions between the EPSPs and IPSPs will be discussed later. Depolarizing IPSPs first appeared about 10 min. after the haemolymph was replaced by 10 K saline and were recorded from some preparations up to 6 hr. later. However, after the muscles had been soaked for about 8 hr. in 10 K saline the IPSPs were once again mainly hyperpolarizing. Stimulation of the inhibitory axon at this time usually failed to elicit any mechanical response from the muscles. Resting potentials of a.c.a. and extensor tibiae muscle fibres of grass-fed locusts increased progressively in 10 K saline but usually did not attain the predicted value of 60 mV. (Usherwood, 1967 a, b) for some hours.
It appears from these results that 10 K saline can have a marked effect on the responses of muscle fibres of locusts and grasshoppers to inhibitory stimulation. This could undoubtedly explain why the inhibitory responses recorded by Hoyle (1966 a, b) from his a.c.a. muscles were so diverse. Hoyle (1966a, b) reported that his a.c.a. preparations were examined in ‘standard’ locust saline which contains 10 m-equiv./l. potassium (Hoyle, 1953) and which he suggests should give similar results to average results obtained from muscles bathed in their own haemolymph. However, in view of the variable potassium content of locust haemolymph a significant fall in the potassium concentration of the external environment of the muscle could occur when the muscle is transferred from haemolymph to 10 K saline which could influence the magnitude and sign of the IPSP. This possibility was examined by observing the effect of controlled changes in the potassium concentration of the medium bathing a.c.a. or extensor tibiae nerve-muscle preparations on the inhibitory responses of these preparations.
About 50 a.c.a. and extensor tibiae nerve-muscle preparations from starved locusts were equilibrated for 1−2 hr. in 10 K saline. The haemolymph of starved locusts contains about 7 m-equiv./l. potassium and changes in the characteristics of the muscle IPSP occur less frequently when nerve-muscle preparations from these insects are exposed to 10 K saline. At the end of the equilibration period resting potentials of about 60 mV. were recorded from most superficial muscle fibres and the electrical responses to inhibitory stimulation were always hyperpolarizing IPSPs. Stimulation of the inhibitory axon of these ‘equilibrated’ preparations had little or no perceptible effect on muscle tension (Fig. 6). The 10 K saline was then replaced with a high-K saline containing 20−100 m-equiv./l. potassium and the preparations were perfused with this high-K saline for periods ranging from 10 min. to 2 hr. The superficial muscle fibres quickly depolarized in the high-K saline but with potassium concentrations less than 50 m-equiv./l. the excitatory and inhibitory axons were still capable of conducting impulses. The EPSPs were immediately reduced in magnitude but the IPSPs were initially enhanced (Fig. 7) although they also became smaller eventually (Fig. 8). Nevertheless the IPSPs were always hyperpolarizing in high-K saline. When the potassium concentration of the high-K saline exceeded 50 m-equiv./l. axonal conduction was blocked. After treatment with the high-K saline the preparations were returned to the control 10 K saline. Repolarization of the muscle fibres in 10 K saline was slow, the rate of repolarization being related to the concentration of potassium in the high-K saline and to the length of time in the high-K saline, i.e. repolarization of the muscle fibres was slowest following a 2 hr. exposure to roo K saline. The effects on the IPSP of exposing a preparation for 10 min. to 20 K saline and then returning it to 10 K saline were minimal. In fact the only noticeable difference seen before and after treatment with the 20 K saline was a slight transient reduction in the magnitude of the IPSP. After longer periods in high-K saline, particularly in saline containing more than 50 m-equiv./l. potassium, the polarity of the IPSP was altered when preparations were returned to 10 K saline (Figs. 7, 8). Stimulation of the inhibitory axon now evoked a small contraction from the muscles (Figs. 6, 9, 10), whereas before high-K treatment either a small relaxation or no mechanical response at all was observed during inhibitory stimulation. In some experiments the preparation was treated with high-K saline and then the potassium concentration of the saline was lowered by 5 m-equiv./l. every 30 min. to a final value of 10 m-equiv./l. Under these conditions the response to the inhibitory axon was at all times a hyperpolarizing IPSP. Depolarizing IPSPs therefore only occur as a result of a large fast reduction in the potassium concentration of the bathing medium. Significantly, when depolarizing IPSPs appear they can be quickly reconverted to hyperpolarizing responses merely by increasing the potassium concentration of the bathing medium.
Recent studies of metathoracic retractor unguis muscles in locusts and grasshoppers have shown that an increase in the concentration of the potassium ions in the bathing medium results in swelling of these muscles as a result of influx of potassium chloride into the muscle fibres accompanied by an osmotic influx of water (Usherwood, 1967a, b). Changes in muscle volume occur only when the potassium concentration of the bathing medium is increased whilst maintaining the chloride concentration constant. If the potassium concentration is increased and compensatory changes in the chloride concentration are made in order to maintain the product of the potassium and chloride concentrations constant there is no influx of potassium chloride and water, and the muscle fibres do not swell (Usherwood, 1967a, b). An increase in the internal chloride concentration of the muscle fibres would be expected therefore during treatment with high-K saline. This change in could account for the appearance of depolarizing IPSPs when the muscles are returned to 10 K saline since the IPSP is a chloride potential. If this is the case, then the inhibitory responses should not be altered after treatment with high-K (constant product) salines since this should not alter . This was tested by treating a.c.a. and extensor tibiae nerve-muscle preparations, which had been equilibrated with 10 K saline for 2 hr., with high-K (constant product) salines containing potassium concentrations of 20-100 m-equiv./l. The muscle fibres were quickly depolarized by the high-K (constant product) salines and there was some reduction in the magnitudes of the IPSPs and EPSPs. The IPSPs were not enhanced initially in the high-K (constant product) salines as they were in the high-K salines. On returning the preparations to 10 K saline the muscle fibres quickly repolarized and the IPSPs and EPSPs returned to normal (Fig. 10). The IPSPs were always hyperpolarizing even following prolonged periods (up to 4 hr.) in 100 K (constant product) saline.
DISCUSSION
The IPSPs of the metathoracic a.c.a. and extensor tibiae muscles in locusts and grasshoppers can be converted from hyperpolarizing to depolarizing responses by soaking these muscles for a short period in saline containing a high concentration of potassium ions and then returning them to saline containing a low concentration of potassium ions, provided that is maintained constant. However, if the products of the potassium and chloride concentrations of these salines are equal (constant product salines) the IPSP is relatively unaffected. Entry of potassium chloride into the muscle fibres during treatment with saline containing a high concentration of potassium appears essential for reversal of the IPSP.
Most of the current during the IPSP is carried by chloride ions (Usherwood & Grundfest, 1965). When the chloride in the medium bathing one of these insect muscles is replaced by the impermeant anion propionate, the IPSPs reverse in sign to become depolarizing potentials (Usherwood & Grundfest, 1965) and the muscle contracts slightly during stimulation of the inhibitory innervation (Fig. 11). The equilibrium potential for the IPSP (EIPSP) is normally more negative than the resting membrane potential (EM) (Usherwood & Grundfest, 1965) and since the IPSP is a chloride potential (Fig. 12) the chloride equilibrium potential (Ec1) must also be negative to EM (Fig. 13). The resting muscle fibre in locusts and grasshoppers is permeable to both potassium and chloride ions (Usherwood, 1967a, b) so that some form of outward chloride pump could account for the discrepancy between EM and ECI. Since these insect muscle fibres are relatively good potassium electrodes it is reasonable to assume that the potassium equilibrium potential (FK) is close to EM. When locust and grasshopper muscle fibres are exposed to high-K (constant product) salines the EMF of the potassium battery and the EMF of the chloride battery should change simultaneously. EK and Ec1 will become less negative and the muscle fibres will be depolarized (Fig. 13). On the assumption that EK and EC1 change to the same extent the magnitude of the IPSP should not change although the increased conductance of the depolarized muscle fibre could account for the reduced magnitude of the IPSP observed in the high-K (constant product) saline. The EPSP is always reduced in magnitude when the muscle is depolarized since EM is then closer to the equilibrium potential for the EPSP (EEPSP) (Cerf et al. 1959) and the driving force on the ions which carry the current during the EPSP is reduced. Potassium chloride does not enter muscle fibres exposed to high-K (constant product) saline since the distribution of potassium chloride in the intracellular and extracellular environments is determined by the equilibrium (where and are the internal and external potassium concentrations (activities) and and are the internal and external chloride concentrations (activities)) and replacement of 10 K saline with a high-K (constant product) saline does not upset this equilibrium. When muscles equilibrated in high-K (constant product) saline are returned to 10 K saline the superficial fibres repolarize very rapidly as EK and EC1 return to more negative levels simultaneously. Therefore EIPSP is always more negative than EM and the IPSP is at all times a hyperpolarizing response (Fig. 13).
When the a.c.a. or extensor tibiae muscle is exposed to a high-K saline with a normal chloride concentration (154 m-equiv./l.) the superficial muscle fibres are quickly depolarized, presumably as a result of the change in the EMF of the potassium battery across the muscle fibre membrane (Fig. 12). The chloride concentration of the bathing medium has not been altered and Ec1 does not therefore change immediately although it slowly becomes less negative as potassium chloride enters the muscle fibre and increases. Since EC1 changes more slowly than EM or EK, the driving force on the chloride ions is temporarily increased and the IPSP is enhanced. The IPSP then slowly declines in magnitude as the equilibrium is re-established. The extra potassium chloride and water which enters the muscle fibres during high-K treatment will be lost when the muscle is returned to the 10 K saline. However, efflux of potassium chloride from muscle fibres of locusts and grass-hoppers is much slower than influx of potassium chloride (Usherwood, 1967a, b). When a muscle fibre loaded with potassium chloride is returned to 10 K saline the extra potassium chloride leaves the fibre only very slowly and EM changes only very slowly (Usherwood, 1967 a, b). Since has not altered significantly during treat-ment with high-K saline, EK will immediately assume the value (60 mV.) appropriate for the new low external potassium concentration according to the Nernst equation EK = (RT/F) log where R, T and F have their usual meanings. increases during treatment with high-K saline and since potassium chloride efflux is slow EC1 and therefore EIPSP will not change very rapidly when a muscle fibre loaded with potassium chloride is returned to 10 K saline. EM should assume a value somewhere intermediate between EK and EC1 and therefore for some time after the fibre is replaced in 10 K saline, EC1 will be less negative than EM and depolarizing IPSPs will appear during inhibitory stimulation. As potassium chloride diffuses out of the fibre EC1 will become more negative but so will EM and therefore the IPSP will remain depolarizing. However, when EM, EK and EC1 become equal the IPSP will disappear and then as the rest of the extra potassium chloride diffuses or is pumped out of the fibre, Ed will become negative to EM and EK, and the IPSP will reappear as a hyperpolarizing response.
If at the end of high-K treatment the potassium concentration of the saline is reduced only slowly then the rate of potassium chloride efflux will be high enough to maintain EC1 below EK and the IPSP will remain hyperpolarizing throughout the period of potassium chloride efflux.
According to these arguments the diverse results obtained by Hoyle (1966 a, b) from his a.c.a. and extensor tibiae nerve-muscle preparations could be accounted for by the different haemolymph-potassium concentrations of his insects. Hoyle (1954) has suggested that the potassium content of locust haemolymph varies between 5 and 40 m-equiv./l. according to the potassium content of the diet. Usherwood & Little (to be published) have failed to obtain such large variations in haemolymph-potassium levels by feeding locusts and grasshoppers on diets containing different potassium concentrations, although haemolymph from insects fed on grass (a high-potassium diet) usually contains about 4 m-equiv./l. more potassium than haemolymph from starved insects. However, fluctuations in the haemolymph-potassium levels are not apparently accompanied by changes in the chloride content of the haemolymph (Usherwood & Rees, unpublished). Therefore any variations in haemolymph-potassium concentration will be accompanied by changes in the chloride concentration of the muscle fibres. Since the haemolymph-potassium concentration presumably varies only very slowly EC1 will always be more negative then EM and the in vivo response to inhibitory stimulation will be at all times a hyperpolarizing IPSP.
Changes in the responses of these insect muscle fibres to inhibitory stimulation probably result mainly from leakage of KC1 from muscles cut during dissection of the nerve-muscle preparations. A number of closely adjacent muscles must be removed to obtain a.c.a. nerve-muscle preparations in locusts and grasshoppers, and since these muscles contain a high concentration of potassium it is possible that some of this potassium could leak out from the cut ends of the fibres during dissection to produce local increases in the potassium concentration of the medium bathing the a.c.a. muscle. Some fibres of the a.c.a. muscle could therefore be subjected for short periods to a high concentration of potassium with resultant increase of the chloride content of these fibres. Perhaps this would explain why in the present experiments depolarizing IPSPs were recorded more frequently from isolated a.c.a. nerve-muscle preparations which involve a considerable amount of dissection. It could also account for the varied inhibitory responses obtained from fibres of the same muscle preparation and also for the fact that depolarizing IPSPs were sometimes recorded also from isolated a.c.a. muscles from starved locusts despite the fact that the potassium concentration in the haemolymph of these insects should be similar to or less than that of 10 K saline. Reversal of the IPSPs to depolarizing responses as a result of potassium leakage from cut fibres could presumably occur even when the muscles were in haemolymph although this did not happen in the present experiments. Artifacts of this type are less likely to arise with the extensor tibiae muscle since this muscle is more easily exposed and in any event is protected to some extent, by the tracheolar system, from any potassium that might leak out during removal of the flexor tibiae and retractor unguis muscles during preparation.
Disregarding any possible alteration of during dissection the responses recorded from a.c.a. and extensor tibiae muscle fibres when the haemolymph is replaced by ‘standard’ locust saline (10 K) will depend on the haemolymph-potassium concentration or, more exactly, on the difference between the products of the potassium and chloride concentrations of the haemolymph and 10 K saline. The possibility that ionized amino acids play some part in determining the membrane potentials of insect muscle fibres cannot be ignored, but the role, if any, that these substances play is not yet understood. If replacement of haemolymph by 10 K saline reduces the external potassium concentration of the a.c.a. muscle and thereby alters the product of the external potassium and chloride concentrations then EK will change faster than EC1 (EIPSP) and EC1 could become less negative than EM. The IPSP would change from a hyperpolarizing to a depolarizing response and contractions rather than relaxations could be recorded from the muscle during inhibitory stimulation. The nature of the interaction between an EPSP and a depolarizing IPSP depends to some extent on the magnitude of the excitatory response. If the EPSP is large then it is attenuated by the depolarizing IPSP. Presumably the peak of the EPSP in this instance is above (less negative than) EIPSP. When the EPSP is small and the peak of the response is below (more negative than) EIPSP, then summation of the inhibitory and excitatory response will occur. The magnitude of the EPSP varies from fibre to fibre in both the a.c.a. and extensor tibiae muscles, and since these muscles are large, considerable differences in EIPSP and EM of different a.c.a. muscle fibres might be expected for some time after exposure to 10 K saline. These differences could account to some extent for the many different inhibitory and excitatory potentials recorded by Hoyle (1966 a, b) from different fibres of his a.c.a. nerve-muscle preparations.
Usherwood & Grundfest (1964, 1965) found that the IPSPs of the extensor tibiae muscle fibres in locusts and grasshoppers were always hyperpolarizing even in 10 K saline, although some differences in the magnitude of the resting potential, EPSP and IPSP were seen in different fibres. This occurred despite the fact that the locust preparations used by Usherwood & Grundfest (1964, 1965) were obtained from locusts that had been starved and were always equilibrated for about 1 hr. in 10 K saline before any recordings were made from the muscle fibres. Any effect on of differences in the potassium concentrations of the haemolymph and 10 K saline should therefore have been minimized. The low resting potentials recorded from many a.c.a. and extensor tibiae muscle fibres (Hoyle, 1966a, b; Usherwood & Grundfest, 1964,1965) were probably in many cases the result of inferior penetration of these fibres with the recording electrodes rather than to an abnormally high . It is possible of course that tonic fibres have lower resting potentials than phasic fibres (Usherwood & Grundfest, 1965). The fact that Usherwood & Grundfest (1965) found that the reversal potential for the IPSP was the same in fibres with low and high resting potentials suggest that the chloride concentrations of these fibres were identical. In the present experiments considerable care was taken to select electrodes with relatively high resistances (c. 20 M Ω) and low tip potentials (c. 5 mV.). Using these electrodes the resting potentials of superficial fibres in any given muscle were relatively constant ( ± 2 mV.), provided of course that these muscle fibres had been completely equilibrated with the saline environment.
In view of the large size and complexity of the a.c.a. and extensor tibiae muscles, considerable errors in measurement of membrane potentials must arise, especially when these involve changes in the ionic environment of the muscles. Changes in membrane potential of the more deeply situated a.c.a. muscle fibres were much slower than for the superficial fibres of the muscle. This was especially true for decreasing . Presumably the full effects of a reduction of resulting in potassium chloride efflux from the fibres are delayed to some extent by the slow rate of diffusion of potassium from the fibres and from the extracellular spaces between the fibres. Quantitative information on the permeabilities of a.c.a. and extensor tibiae muscle fibres cannot therefore be obtained easily from whole-muscle preparations. Obviously, isolated single-fibre preparations are the complete answer to this problem but preparations of this type are very difficult to obtain, at least in the grasshopper and locust. Usherwood (1967 a, b) and Cochrane & Elder (1967) have investigated the potassium and chloride permeabilities of isolated retractor unguis muscles in locusts and grasshoppers. These preparations contain only seventeen fibres and are considerably better than the a.c.a. and extensor tibiae muscles for investigation of ion fluxes. Unfortunately the retractor unguis muscle is not innervated by an inhibitory axon and the muscle does not contain any tonic fibres.
The main object of this publication was to substantiate claims by Usherwood & Grundfest (1964, 1965) that some orthopteran muscles are innervated by peripheral inhibitory axons similar to those found in some crustaceans. The experiments described herein were not done with a view to arriving at a definition of the role of peripheral inhibitory innervation in insects. This aspect will be considered at some length in a later publication.
ACKNOWLEDGEMENT
The author is indebted to Professor H. Grundfest for his criticism of a draft of this manuscript.