Some properties of the action potentials conducted in the spines of the sea urchin Diadema antillarum

Camp. Brachtm. Phpint. Vol. XIA, No. 1, pp. 15-13. 1985 0300~ 9629/%5 $3.00 + 0.00 Printed in Great Britain ‘c_‘; 198s Pergamon Press Ltd

SOME PROPERTIES OF THE ACTION POTENTIALS CONDUCTED IN THE SPINES OF THE SEA URCHIN

DIADEMA A~T~LLA~~~

ANA BERRIOS, DIEDRE BRINK, Jo& DEL CASTILLO

Laboratory of Neurobiology, Medical Sciences Campus, University of Puerto Rico, Boulevard de1 Valle 201, Old San Juan, PR 00901, USA

and DAVID SPENCER SMITH

Department of Zoology, Oxford University. South Parks Road. Oxford OXI 3PS. England

Abstract- 1. Brief (2.-5 msec) electrical pulses applied to the primary spines of the sea urchin Di&etiru anfillurunz elicit graded action potentials (0~‘s).

2. These up’s can be attributed to the electrical activity of a set of 1421 bundles of neurites, each comprising 1000 processes near the spine base and tapering towards the spine tip.

3. The shape of the ap’s varies from a simple diphasic deflection to a complex waveform with 6 or more components. Peak-to-peak amplitude is < ImV.

4. The up’s are conducted at a uniform speed of cn. 27 cm/set. 5. The ap’s are not affected by tetrodotoxin (1 ng/ml) and continue to be produced in Na-free artificial

sea water (ASW). 6. The amplitude of the up’s is greatly reduced or totally abolished in Ca-free ASW. However, some

electrical activity may continue in the absence of external Ca, due to release of Ca?’ ions from the calcium carbonate crystals of the spine shaft.

7. Replacing the Ca content of ASW by barium ions causes an irreversible blockade of the up’s, 8. Spines equilibrated with ASW containing SF ’ ions instead of Ca” produce ap’s of increased

amplitude (up to x 2). 9. The up’s are blocked by La’ ’ , Co”, Cd’ ’ (2-5 mM) and by the organic Ca channel blocker Bepridil

(2 mM). 10. We conclude that the spinal ap’s are due to the summation of Ca spikes produced by the activation of

Ca channels which are blocked by barium and have a high affinity for, or permeability to Sr vs Ca.

INTRODUCTION

Electrical stimulation of the tip of a single spine of the tropical sea urchin Diadema antillurum elicits synch- ronous movements of the adjacent spines (Smith et af., 1984). Since the spread of current to the bases of the spines. where the muscles that move them are located (Smith et al., 1981) could be ruled out, we concluded that a wave of excitation must be conducted along the spines. This was confirmed by the observation that compound action potentials (ap’s) are generated by electrical stimulation of the spines and are conducted along them in both proximal and distal directions.

The above observation was of considerable interest since, almost a century ago, Hamann (1887) described ‘nerve fibers’ ascending the spine from the nerve ring surrounding the spine base, in the echinoid Centroste- @anus longispinus. Since then, this extensive region of the echinoid nervous system has received little atten- tion. Hamann’s finding is mentioned briefly by Hyman (1955) and Smith (1965). but not in later accounts (Smith, 1966; Cobb, 1982) or in the excellent reviews of the echinoderm nervous system by Pen- treath and Cobb (I 972). 1982), which do not consider any conducting system distal to the spine base. Furth- ermore. the presence of nerve fibers within the spine has not been taken into account in experimental work

dealing with the coordination of spine movement (Millet, 1966).

We have confirmed Hamann’s anatomical conclusion (Smith ef al., 1984. light and electron micro- scopic studies have shown the presence of up to 21 regularly disposed nerves within a primary spine of ~jadema. In the basal region of the spine, each nerve comprises > 1000 neural processes, ranging in diameter from < 0. I to cu. 2 pm; the nerves taper distally and end at or near the spine tip as very slender bundles, each including fewer than 100 processes. The small diameter of the processes, together with the fact that they are embedded within the crystalline skeleton of the spine, has necessitated a rather indirect experimental approach-extracellular recording of ap’s well below 1 mV in amplitude. However. the results obtained suggest that the electrical activity of the spinal nerves is due to the operation of a specific type of Ca channel. blocked by Ba’+ ions and having a higher affinity for, or permeability to strontium vs calcium.

MATERIALS AND METHODS

Large adult specimens of i%llrm~ a~7tillarum were col- lected at various sites on the North and Northeast coast of Puerto Rico, and on off-shore reefs. They were transferred

15

16 A. BERRIOS et al

Table I

ASW Na-Free ASW Ca-Free ASW Sr-ASW Ba-ASW Stock solutions Parts stock Fmal cone. Parts Concn Pam Concn Parts Concn Parts C‘oncn

solution M S&l M SOIll M SOIll M S&l M

Compound Concn M

NaCl 0.56 x04 0.450 x04 0.450 x04 0.450 x04 0.450 KCI 0.56 IX 0.010 IX 0.010 IX 0.010 IX 0.010 IX 0 010 CaCI, 0.38 2X 0.01 I 2X 0.01 I M&I: 0.36 I46 0.053 146 0.053 I46 0.053 146 0.053 I 46 0 053 SrC12 0.38 2x 0.01 I B&I, 0.38 2X 0 01 I NaHCO, 0.56 4.6 0.002 46 0.002 4.6 0.002 4.6 0.002 SllWXe 0.93 808 0.752 21.4 0.025

to holding tanks of the laboratory where they lived in good condition for several weeks. Diadema is a hardy organism; its most important requirement is a good oxygen supply, and it generally occurs in the shallows where wave action keeps the water well aerated.

Long (> 15 cm) primary spines were removed while the urchins were still in the holding tank, and kept in dishes of aerated sea water, where they remained excitable for up to 48 hr. However, all experiments were carried out on freshly cut spines. In each experiment 20 different spines were tested.

Solutions

Unless otherwise noted, all experiments were carried out in natural sea water (NSW) taken from the supply to the holding tanks. In experiments where the ionic composition of the sea water was changed, artificial sea water (ASW) was employed. This was prepared by mixing stock solutions isotonic with NSW (see Hodgkin and Katz, 1949). Stock solutions and proportions mixed are given in Table I

Electrical stimulation and recording

The spines were placed over an array of non-polarizable electrodes, as shown in Fig. 1. As most spines show a slight curvature, it was necessary to use a flexible means of contact: each electrode consisted of a piece of glass tubing fitted at one end with a bundle of bristles (from a camel hair brush) about 1 mm in diameter. A chlorided silver wire or pellet was placed in the tube which was then filled with a conductive gel (gelatine 5-lo%, in sea water). When not in use, electrodes were short-circuited and kept immersed in NSW.

Spines were stimulated by brief (2. 5 msec) pulses of supramaximal strength delivered via an insulating trans- former. The recording electrodes were connected directly to a Tektronix 5A22N differential amplifier (1 Ma input resistance) set at a bandwidth of O&300 Hz. The output of the amplifier was connected to a Nicolet Digital Oscilloscope (Model 4094) where the ap’s were averaged to improve the

1

DIF AMP DIG OSC

Fig. I. Arrangement of the stimulating (1, 2j, recording (4, 5) and grounding 131 electrodes. The interelectrode distances in all experiments, with the exception of those designed to measure conduction velocity (see Fig. 2 and text) were I-2, 10 mm; 2-3, 9 mm; 334, 10 mm and 4-S. 9 mm.

signal-to-noise ratio. The amplitude of the recorded potentials did not increase appreciably by interposing a high input impedance preamplifier, or by increasing the bandwidth. Electrical recording was initially made with the spines rest- ing on the electrode array in air; however, the recorded ap’s were found to be larger and more stable in preparations immersed in a bath of paraffin oil.

RESULTS

The spinal action potentials

(i) Shape. As recorded by the above technique, the spinal ap’s appear as graded waveforms of variable amplitude, shape and duration. Their overall shape varies from an almost symmetrical diphasic signal to one showing six or more components. The most commonly observed waveform is that shown in Fig. 2. which may be regarded as typical. It exhibits two positive and three negative deflections.

It is likely that such a complex shape reflects the activity of several groups of nerve fibers, with different conduction velocities. This suggestion is supported by the observation that the different deflections have differing sensitivities to changes in the composition of the ASW, or to the addition of blocking agents to the NSW.

(ii) Size. The largest peak-to-peak amplitude of the ap’s observed is almost 1 mV, although 20&300 uV is far more common. The smallness of the observed amplitude is not surprising, since the circlet of spine nerves, each up to 40 pm dia., is inserted within the crystalline spine shaft, l-1.5 mm dia. and saturated with sea water, which acts as a low-resistance shunt.

(iii) Conduction velocity. The rate of conduction of the ap’s along the spine was measured by changing the distance between the stimulating and recording electrodes, while keeping constant the distances between electrodes 1 and 2, on the one hand, and 3.4 and 5, on the other (cf. Fig. 1). The proximal end of the cut spine was always placed over the stimulating electrodes. In freshly cut spines, the ap’s are conducted at constant speed over distances of at least IO cm. Thus Fig. 2 shows four action potentials recorded at 4 different loci from the stimulating electrodes. The shape of the ap’s remained constant, though the amplitude of the individual deflections decreased, an effect probably due to the decrease in number of neurites within each nerve tract, occurring as the spine diameter decreases from the base to the tip.

In Fig. 3, the time elapsed between the stimulus

Action potentials in Diadema spines 17

D

A/+-- Fig. 2. Four action potentials recorded at increasing distance between electrodes 2 and 4 (see Fig. 1). The distances between electrodes I and 2 on the one hand, and 3.4 and 5 on the other, remained unchanged. In records A, B. C and D the distance between electrodes 2 and 4 was sequentially 19. 30,47 and 55 mm. The left edge of each record corresponds to the position of the stimulus artifact: note increasing distance between this point and the action potentials. (Cali-

brations: vertical 100 uV: horizontal 100 msec.)

artifact and the peak of the first negative deflection is plotted against the distance between the stimulating and recording electrodes, for three spine preparations. The experimental points are fitted reasona- bly well by straight lines with a slope of cu. 27 cmjsec.

The ionic basis of the spinal action potentials

Lack of action of tetrodotoxin. To probe the possible involvement of Na channels in the generation of spinal up’s, we tested the influence of tetrodotoxin (TTX) on the electrical activity of the spine. After recording a control up, the spines were immersed in NSW containing 1 pg/ml TTX. This failed to influence appreciably the recorded up’s, even after 30 min.

To ensure that the TTX stock used ( 10e4 g/ml) was active, we prepared a solution of 1 ug/ml TTX in frog

Ringer: this blocked the up’s of two frog sciatic nerve preparations in about 5 min. Recovery took about 1 hour in fresh Ringer’s solution. Thus it may be concluded that the spinal up’s are not due to the activation of fast Na channels; in which event the ap’s

Fig. 3. Graph obtained by plotting the distances between electrodes 2 and 4 (see Figs. I and 2) against the interval between the stimulus artifact and the peak of the main negative deflection, in 3 different spines. The straight lines correspond to a constant conduction velocity of ca. 27 cm/

sec.

should continue to be produced in the absence of Na’ ions in the medium.

Action potentials in Na-jiree AS W. To test the above possibility, spines were immersed in ASW in which NaCl was replaced by an iso-osmotic amount of sucrose (see Table 1). As shown in Fig. 4 (records A-3, B-3). spines are still excitable in the absence of sodium in the ambient solution. In fact, the ap’s recorded in Na-free ASW with normal Ca content are larger than control ap’s in normal ASW. This is a consequence of the increased resistivity of the external solution, resulting from replacement of NaCl by sucrose.

These results suggest that a cation other than Na’ is responsible for the charge transfer involved in generation of the up’s and Ca’+ was investigated as a candidate for this role.

lnjluence of Ca-free AS W. To test the possible role of Ca’+ ions in spine electrical activity, we studied their behaviour in Ca-free ASW. After recording control up’s from spines equilibrated in ASW, spines were transferred to a solution of similar composition except for the absence of Ca’+ ions (see Table 1). Both the amplitude and time course of the ap’s were seen to change in this medium; they became smaller and slower as shown in Figs. 6 and 7 (records A-2. B-2). Yet, contrary to expectation, a total blockade of the up’s in Ca-free ASW was observed only in a few spines. In addition, in some experiments the up’s were

abolished when the spines were first transferred to Ca- free medium but reappeared, with greatly diminished amplitude, if the spines were left longer in the Ca-free solution.

The spine shaft as a source of calcium

The relative refractoriness of the spinal up’s to Ca- free ASW, and the observation that they could be abolished entirely and yet reappear while still in the same medium suggest that even if Ca” ions have been omitted from the experimental solution, a source of calcium present in the preparation is sufficient to support some degree of electrical activity. This is not surprising, since the slender spinal nerves are sur- rounded by the crystalline skeleton, composed of ca. 90% calcium carbonate and 10% magnesium carbo-

A. BERRIOS et al.

(a) (bf

Fig. 4. Records illustrating(i) the occurrence of electrical activity in a Na-free medium and (ii) the influence of [Ca’+], on the amplitude of spine action potentials. A-l and B-l show control up’s of 2 spines equilibrated with ASW for 10 min. A-2 and B-2 are up’s recorded after equilibration with Na-free ASW with half the normal Ca, Note that despite the slower time course, the peak-to-peak amplitude of the ap’s is increased, due to the higher resistivity of the medium. A-3 and B-3 represent rip’s after equilibration with Na-free ASW containing normal Ca. Note both increased amplitude and conduction velocity. When the spines were equilibrated with Na-free ASW containing x 2 normal. [Cal. the up’s were usually abolished (see text). An exception was noted in spine B; its time course was faster and the amplitude of the 2nd

negative component greatly enhanced. (Calibrations: Vertical, 150 uV; Horizontal, 50 msec.)

nate (Raup, 1966). Furthermore, the spine is invested with a limiting epithelium that may act as a diffusion barrier. Thus any factor that tends to destabilize the calcium carbonate will release Cal’ ions within the spine, which may reach concentrations sufficient to permit electrical activity.

One such factor is the pH of the Ca-free ASW; this was 7.7-1.1 pH units lower than that of NSW. When this was further decreased to pH 6.1 by addition of HCl to the ASW, the spines remained excitable for long periods, and the amplitude of the q’s was only slightly reduced from the control level.

The usual method of eliminating Ca’+ ions from a solution, by use of EDTA, could not be employed in these experiments, since this compound will not only chelate free Ca*+ ions but will also bind crystalline calcium; indeed, we have used EDTA (with EGTA) to decalcify spines prior to structural study of their ‘soft tissue’ components. Instead, to minimize the concentration of Ca*+ in the environment of the spinal nerves,

we used a Ca-ME-free ASW to which 5 mM Na,CU, was added. This compound prevents the release of Ca and Mg from the spine crystals by precipitating both divalent cations. Spines immersed in this solution showed an immediate blockade of ap’s which was largely, but not completely reversible, as shown in Fig. 5.

It may be concluded from these observations that Ca2+ ions are essential for the electrical activity of Diadema spines. The compound ap’s recorded from these structures can be regarded as due to the summation of Ca spikes. If this conclusion is correct, the following predictions, based on the usual criteria for identification of Ca channels (see Hagiwara and Byerly, 1981a. b) can be made:

(1) the amplitude of the ap’s should be dependent upon PI,;

(2) the Ca2+ ions in ASW should be replaceable by both Ba2+ and Sr’+ ions

(3) the spinal ap’s should ‘be blocked by several

Action potentials in Diadem spines 19

w

W

Fig. 5. Illustrating the almost complete blockade of up’s in Ca/Mg-free ASW containing 5 mM N&O,. a control in normal ASW: h record in presence of Na,CO,: c, d records showing partial recovery of up’s in normal ASW. (Calib-

rations: Vertical. 50 pV: Horizontal. 50 msec.)

cations, such as Cd”, Co’+, La’+ and Mn” as well as by several organic compounds known as calcium channel blockers.

However, not ail of these expectations were fulfilled.

Effect of changitzg (CuJ,

To obtain information on the calcium dependence of the spinal up’s, we investigated the effects of solutions with varying Ca concentrations. For this purpose, after recording control ap’s in ASW, the spines were immersed successively in Na-free ASW containing x i, x 2 and x 4 the normal Ca content. Results of one such experiment is illustrated in Fig. 4. Records A-i and B-i are control ap’s, while A-2 and B-2 are recorded from the same spines equilibrated with Na-free ASW containing half the normal calcium; the potentials are now slower but exhibit increased amplitude due to the increased resistivity of the external solution. Records A-3 and B-3 are the up’s recorded from the same spines in Na-free ASW with normal Ca content (11 mM): here the ap’s become larger and faster. Increasing the Ca concentration to x 2 or x 4 normal results in blockade of the ap’s; only a few spines, including spine B in Fig. 4 were still excitable in the presence of twice the normal

[Cal,.

The results with strontium were also unusual, as illustrated in the experiment shown in Fig. 7, which followed the same protocol as in Fig. 6, but with Sr” replacing Ba’+ in the artificial sea water. The spines equilibrated with Sr-ASW were not only electrically excitable, but in addition, the peak-to-peak amplitude of the recorded up’s was about twice as large as that of controls in normal ASW. This observation suggests that the voltage-dependent Ca channels in Diadema spine nerves are more permeable to, or have a greater aflinity for strontium ions than calcium ions.

Eflects qf‘lanthanum, cadmium, cobalt and manganese ions

The abolition of up’s in high external Ca is likely to be due to the inactivation of the Ca channels produced by accumulation of Ca’+ ions on the inner side of the membrane, as first proposed by Brehm and Eckert (1978) and Tillotson (1979). A rapid increase in [Cal, is very likely to occur in these very thin fibres (UZ. 0.1-2 pm dia.) due to their extremely high surface: volume ratio.

These cations block the major proportion of the Ca channels so far studied. Lanthanum ions at a concentration of 5 mM blocked electrical activity of the spines, completely and irreversibly. Likewise. Cd’+ blocked the spine ap’s at a concentration of 2 mM; its effect could not be reversed by washing the preparation in normal ASW. However. in this instance, the ap’s reappeared in most preparations after they were treated with NSW containing 2 mM cysteine. Cobalt ions completely blocked the electrical activity at a concentration of 5 mM, while at 1 mM, this cation slightly potentiated the spine ap’s.

Finally. Mn” ions, which block many Ca channels, failed to block the Diadema preparation even at a concentration of 14 mM. As illustrated in Fig. 8, the only effect of Mn at this concentration was to decrease slightly the velocity of conduction of the up’s, and change slightly the shape of the potentials.

Spines immersed in normal ASW or NSW can be As Mn” ions are known to permeate the Ca stimulated at 1 Hz for periods of I min or more with channels in myoepitheliai cells of .Syt& ~~a~gi~b~~a, a no appreciable change in the amplitude of the up’s. In marine annelid (Anderson, 1979). where they are able fact. a slight potentiation of the rip’s is usually to sustain overshooting ap’s, we replaced the calcium

observed during the first few stimuli. However, a rapid depression occurs during repetitive stimulation of spines equilibrated with solutions of above normal Ca concentration. In most instances, they become electrically inexcitable when the Ca level reaches x 2 normal, and always at x 4 normal.

Eflects qf replaring calcium by barium and strontium

Most, if not ail, Ca channels studied have been found to be permeable not only to Ca’+ ions, but to Ba”” and Sr’+ also. Therefore, these two divalent cations substitute for Ca in maintaining regenerative responses; indeed, this is one of the criteria used to identify calcium spikes (Hagiwara and Byerly, 198ia. b). To determine whether the channels underlying the ap’s conducted in Diadema spines are also permeable to these cations, we carried out standard ion substitu- tion experiments.

Figure 6 illustrates the unexpected results obtained with barium as the substituted ASW constituent (see Table I). Records A-2 and B-2 are the up’s obtained from the same spines, initially, in Ca-free ASW. Records A-3 and B-3 show the absence of electrical activity when spines were equilibrated with ASW in which all the Ca is replaced by an equivalent amount of Ba’- ions. These experiments showed not only that E&a’” is unable to replace Ca”, but. furthermore, irreversibly blocks electrical activity of Diudema spines. Indeed, the ap’s did not reappear even after 30 min immersion of the spines in normal ASW.

20

(a)

A. BERRIOS et al.

(b)

Fig. 6. Illustrating the blocking action of Ba” ions on spinal up’s, A-l, B-l represent the ap’s in NSW. Each ap is the average of 10 traces. A-.. , 7 B-2 show the up’s elicited in the same spines after equilibration in Ca-free ASW. The ap’s do not always disappear in this solution: however, repetitive stimulation tends to depress the response. Thus, A-2 and B-J are the average of only 3 traces. A-3, B-3: following replacement of the normal calcium content of ASW with an equivalent concentration of Ba’+. ap’s are completely blocked. This effect is irreversible, as shown in the next pair of records: A-4 and B-4 in which up’s fail to reappear in

NSW. (Calibrations: Vertical, 100 uV; Horizontal. 100 msec.)

content of the medium with manganese, in some experiments. However, no electrical activity in Dia- dema was detected in Mn-ASW.

Organic Ca channel blockers

Two compounds of this group of drugs were tested on Diadema spines. Verapamil at a concentration of 2 mM slightly slowed the time course of the ap’s, while at the same concentration, Bepridil markedly decreased the peak-to-peak amplitude of the up’s and in some instances abolished electrical activity of the spines, completely and irreversibly (Fig. 9).

DISCUSSION

The ‘rediscovery’ of nerves in the spines of Dia- dema, corresponding to those originally described in Centrostephanus by Hamann in 1887 is of general neurobiological interest, since they represent a major but overlooked division of the echinoderm nervous system. All our structural and electrophysiological work at present relates to Diadema antillarum equipped, as is the sea urchin studied by Hamann, with very long slender spines. We have yet to ascer- tain whether similar spinal nerves are present in other genera; no electrical activity was seen in preliminary observations on the locally available Echinometra and Eucidaris, but this is inconclusive. since due to the much lower length : diameter ratio of these spines, the

shunt represented by the calcified skelton may prevent the recording of the action potentials. Structural studies on spines of these genera are in progress.

The action potentials in the spines of Diadema appear to be produced by the summation of Ca spikes. Their properties serve to emphasize the diver- sity of Ca channels found in nature. The channels of the spinal nerves are blocked not only by La’-, Co’- and Cd”, but also by barium, a cation that in almost all instances is able to replace calcium. Further, Mn’* ions, which block many Ca channels, failed to block the spinal up’s in Diudema, even at high concentration. However, Mn is not able to replace Ca, as it is in cells of the proventriculus of the marine annelid Syllis spongiphila (Anderson, 1979).

Strontium is the only divalent cation able to replace Ca in Diadema. and while our observations do not allow quantitative conclusions, they suggest that either the permeability to, or affinity of Sr“ for the Ca channels is higher than that of calcium ion itself.

The overall shape of the spinal ap’s and their properties suggest that they are due to the activity of several groups of neurites, differing not only in their diameter and velocity of impulse conduction, but also in their sensitivity to calcium channel blockers. Trans- verse thin sections of the nerves substantiate the first conclusion: while most of the neural processes are in the <0.1-0.3 urn dia range. less frequent profiles up to 2 pm dia. are present.


(a) (b)

Fig. 7. Illustrating the effect of replacing Ca2’ ions by ST’+ on the spinal action potentials. Records in columns A and B were obtained from two different spines. A-l and B-I are control up’s in ASW. A-2 and B-2 after 10 min in Ca-free ASW. A-3 and B-3 after equilibration in ASW in which ail calcium was replaced by an equal amount of strontium. The amplitude of the up’s increased so much that the gain was decreased by a factor of one half. A-4 and B-4 after re-equilibrating the spines with Ca-free ASW for 10 min.

(Calibrations: Vertical, 100 pV records I , 2, 4 and 200 pV record 3: Horizontal, IO0 msec.)

It is interesting to note that some of the results described here in the context of a ‘novel’ echinoderm neural system are in general agreement with previous reports in the literature on aspects of neurophysio- logy within the Phylum. Thus, the conduction velocity recorded here does not greatly exceed the upper range of velocities (2-20 cm/s) of the graded compound action potentials recorded from the radial nerve cords of asteroids, echinoids and ophiuroids, although single unit spikes propagating at the rate of up to 78 cmjsec have been detected in the radial nerves of Ophiosila calfornica (refs in Pentreath and Cobb, 1982). Also, the role of calcium ions in the generation of action potentials in echinoderms was first proposed by Millott and Okumara (1968) and Binyon and Hasler (1970).

Our structural studies (which will be described elsewhere) suggest that many of the neurites comprising the spine tracts in Diudema are distai extensions of cell bodies situated near the apex of the cone of ‘ligament’ and muscle (Smith et al., 1981) surrounding each spine articulation. Others may have a more proximal origin, possibly at the level of the nerve ring surrounding each spine base (Hyman, 1955), in accord with the original description by Hamann (1887). The number of neurites within each tract, more than

1000 in the basal region of the spine, is reduced as the spine tapers until each nerve includes fewer than 100 processes at the tip; an unusual anatomical arrangement, perhaps suggesting that detection and trans- duction of appropriate stimuli by the neurites takes place at all levels along the spine.

We have examined the fine structure of the tip of Diadema primary spines, but have noted no structural specializations. there or elsewhere along the spine, that offer a clue to the function of the spine nerves. However, Pentreath and Cobb (1982) have pointed out that sensory receptor elaboration is rarely encountered in echinoderm nervous systems. In the context of organization of the spine tip, it may be noted that we have found no evidence of structural elaboration for venom secretion in Diadema such as that described in Ast~ze~osoma by Sarasin and Sarasin (1888) (cited by C&mot, 1948 and Hyman, 1955). This is in accord with the statement by Russell (1984) that primary spines of diadematid sea urchins are injurious because of their length and fragility rather than through wound envenomation.

In principle, if the spine neurites are, as we suggest, primarily distal sensory processes. they might trans- mit mechanical, chemical or thermal information. The only evidence at present available, and that

22 A. BERRlOS rt ul.

Fig. 8. Illustrating the influence of Mn’ ’ ions on the spinal up’s A-l, B-I represent controls in NSW. A-2, B-2 and A-3, B-3 represent up’s recorded after exposure of 10 and 30 min respectively to ASW containing 7 mM Mn2’ ions. Note (i) increased interval between the stimulus and the up’s, (ii) bifurcation of the main negative peak and (iii) abolition of the late negative humps. A-4, B-4 were takefi 10 min after increasing the [Mn] to 14 mM and A-5, B-S after 4 hr exposure to this solution. (Calibrations: Vertical, 50 uV; Horizontal.

50 msec.)

(a) (b) 1

2 2

I- -,\-‘_

Fig. 9. Illustrating the effect of the Ca channel blocker Bepridil on Diadema spine action potentials. A-l and B-l are control up’s recorded from 2 spines equilibrated with NSW. A-2 and B-2 show the effect of exposure

to 2 mM Bepridil for 10 min. (Calibrations: Vertical, 50 uV; Horizontal, 50 msec.)


circumstantial, may point to the first of these possibi- lities. Some behavioral observations on echinoids have suggested that spines may receive tactile mechanical information. Thus, von Uexkiill(l900) regarded contact between spines as an important factor in the coordination of spine movement. Furthermore, Bul- lock (1965). in a study of spine-to-spine communica- tion involved in the spinexonvergence reflex noted, in representatives of several genera. that ‘very slight tactile stimulation at the side of the spine, sometimes even near its tip, is a strong stimulus for movement of neighboring spines’.

Experiments are in progress in quest of more rigorous information on the sensory modality con- ferred on Diadema by extension of its nervous system by up to 20 cm from the surface of its test.

AcX-no~led~emeni.~~This work has been supported by NIH Grants NS-07464 and NS-14938. Our thanks are due to Mr. F. McKenzie for collecting the specimens, to Mr G. Garcia de la Noceda for help with the illustrations and to Ms. Minerva Rodriguez for help in preparation of the manus- cript. The Verapamil and Bepridil (W-2799 CERM) used were gifts from Knoll Pharmaceutical Co., and Wallace Laboratories, respectively.

REFERENCES

Anderson M. (1979) Mn” ions pass through Ca” channels in myoepithelial cells. J. e.wp. Biol. 82, 2277238.

Binyon J. and Hasler B. (1970) Electrophysiology of the starfish radial nerve cord. Camp. Biochem. Physiol. 32, 747-753.

Brehm P. and Eckert R. (1978) Calcium entry leads to inactivation of calcium channels in Paramecium, Science 202, 1203-I 206.

Bullock T. H. (1965) Comparative aspects of superficial conduction systems in echinoids and asteroids. Am. Zool. 5, M-562.

Cobb J. L. S. (1982) The anatomical basis of integratory mechanisms in echinoderms. In Echinoderms: Proc. Int. Corzf: Tampa Buy (Edited by Lawrence J. M.), pp. 409412. A. A. Balkema. Rotterdam.

Cuenot L. ( 1948) Echinodermes, Stomochordks, Prochord&. in Trait& de Zoologie. Anatomic. Sysknatique. Biologic (Edited by Gras@ P. P. ). Vol. II. p. 172. Masson, Paris.

Hagiwara S. and Byerley L. (1981a) Calcium channels.Ann. Rev. Neurosci. 4, 69-125.

Hagiwara S. and Byerley L. (1981b) Membrane biophysics of calcium currents. Fed. Proc. 40, 2220-2225.

Hamann 0. (1887) Beitrlge zur Histologie der Echinoder- men. Heft 3. Anatomie und Histologie der Echiniden und Spatangiden. Jena Ztschr. Naturwiss. 21, 114176.

Hodgkin A. L. and Katz B. (1949) The effect of sodium ions on the electrical activity of the giant axon in the squid. J. Phgsiol. (Lond.) 108, 37-77.

Hyman L. H. (1955) The Invertebrates: Echinodermata,Vol. 4. McGraw-Hill, New York.

Millott N. (1966) Coordination of spine movement in echinoids. In Physiology of Echinodermatn (Edited by Booloo- tian R. A. ), pp. 465485, Interscience, New York.

Millott N. and Okumara H. (1968) The electrical activity of the radial nerve in Diadema antillarum and certain other echinoids. J. exp. Biol. 48, 2799287.

Pentreath V. W. and Cobb J. L. S. (1972) Neurobiology of Echinodermata. Biol. Rev. 47, 363-392.

Pentreath V. M. and Cobb J. L. S. Echinodermata. In Electrical Conduction and Behaviour in ‘Simple’ Inverte- brates (Edited by Shelton G. A. B.), pp. 44G472. Claren- don Press, Oxford.

Raup D. M. (1966) The endoskeleton. In Physiologv of Echinodermata (Edited by Boolootian R. A.). pp. 379-395. Interscience. New York.

Russell F. E. (1984). Marine toxins and venomous and poisonous marine plants and animals (Invertebrates). Adv. Mar. Biol. 21, 6&217.

Smith D. S., Wainwright S. A., Baker J. and Cayer M. L. (198 1) Structural features associated with movement and ‘catch’ of sea urchin spines. Tiss. Cell 13, 299-320.

Smith D. S.. Brink D. and de1 Castillo J. (1984) Nerves in the spines of sea urchin: a neglected division of the echinoderm nervous system. Proc. natnl Acad. Sci. U.S.A. (in press).

Smith J. E. (1965) Echinodermata. In Structure andFuncfion of the Nervous Systems ofInvertebrate. (Edited by Bulmck T. H. and Horridge G. A. ). Ch. 26 Freeman. San Francisco.

Smith J. E. (1966) The form and function of the nervous system. In Physiology of Eehinodermatu (Edited by Boo- lootian R. A.), pp. 503-5 11. Interscience, New York.

Tillotson D. (1979) Inactivation of Ca conductance depends on entry of Ca ions in molluscan neurones. Proc. Natnl Acad. Sci. U.S.A 76, 1497-1500.

von Uexkiill J. (1900) Die Physiologie des Seeigelstachels. Z. Biol. 39, 73-112.

Documents

Some properties of the action potentials conducted in the spines of the sea urchin Diadema antillarum