Action potential characteristics of demyelinated rat sciatic nerve following application of 4-aminopyridine

Bram Research, 363 (1986) 1-9 Elsevier

1

BRE 11358

• Research

Reports

Action Potential Characteristics of Demyelinated Rat Sciatic Nerve following Application of 4-Aminopyridine ELISABETH F TARG and JEFFERY D KOCSIS Department of Neurology, Stanford Umverslty School of Medtcme, and Veterans Admmtstrat:on Medical Center, Palo Alto, CA 94304 (U S A ) (Accepted May 7th, 1985) Key words demyelinatton - - 4-am]nopyrtdme - - penpheral nerve - - lysophosphat]dylcholine

The sciatic nerves of rats were demyelinated by micromlect~on of lysophosphatldylchohne A variety of abnormahtles such as conduction slowing and block were present Application of the potassmm channel blocker 4-amlnopyndine (4-AP) to the lesion site, led to an increase in area of the compound action potential recorded across the site of demyehnation Single axon recordings revealed three types of changes that may account for the 4-AP-mduced increase m the compound response. One group showed broadening of the acUon potential Other axons showed hyperexcltabd]ty following 4-AP, as manifest by spontaneous firing and multiple spike discharge following a single stimulus. In some of the axons studied, 4-AP led to overcoming of conduction block Although many axons showed increased excltabihty properties m the presence of 4-AP, the frequency-following ab]hty of the axons was reduced, and the absolute refractory period of the axons was increased. These results ln&cate that pharmacolog]cal blockade of potassmm channels with 4-AP not only leads to action potentml broademng m demyehnated axons, but to a variety of excitability changes. These heterogeneous effects of 4-AP should be considered m the rauonale for its climcal use INTRODUCTION The potassium channel blocking agent, 4-aminopyridine (4-AP), is being used in clinical trials on multiple sclerosis (MS) patients12. 23. T h e rationale for its use is that b l o c k a d e of potassium conductance will lead to b r o a d e n i n g of the action potential, thereby increasing the probability of impulse p r o p a g a t i o n through a region of low safety factor such as a site of demyelination1,4,22,25. I n d e e d , application of potassium channel blocking agents 1,2s,3° or sodium channel inactwation inhibitors I can lead to restoration of conduction in experimentally d e m y e l i n a t e d nerve fibers. While 4 - A P has minimal effects on n o r m a l m a t u r e myelinated axons, it causes b r o a d e n i n g of the action potential in demyelinating, regenerating and developing m a m m a l i a n axons 2,16.20,30. H o w e v e r , 4 - A P has other effects on axon excitability aside from action potential b r o a d e n i n g that m a y limit its clinical usefulness. F o r example, in young animals while m o t o r fi-

bers show spike b r o a d e n i n g following 4 - A P application, sensory fibers can give rise to multiple spike discharge or bursting following a single stimulus3,16. A d ditionally, 4 - A P can lead to increased spontaneous impulse activity and spike b r o a d e n i n g in mature animals with e x p e r i m e n t a l demyelinating lesions 30. It has been suggested that the severe paresthesias, or tinghng, r e p o r t e d for patients given 4-Ap12,18 may be related to the propensity of sensory fibers to give rise to impulse burst activity following 4 - A P application 3,30. In this report, we describe alterations in conduction properties of rat sciatic nerve fibers demyelinated by focal injection of lysophosphatidylcholine, and the influence 4 - A P has on modifying the excitability of these fibers. O u r results suggest that while more reliable conduction m a y occur m d e m y e h n a t e d axons m the presence of 4-AP, following single stimuli, conduction in response to multiple stlmuh may be less reliable.

Correspondence" J.D. Kocsis, Stanford Umverslty Neurological Umt, VA Medical Center (127), 3801 Miranda Avenue, Palo Alto, CA 94364, U S A 0006-8993/86/$03.50 ~) 1986 Elsevier Science Pubhshers B V (Blome&cal Dwlslon)

MATERIALS A N D M E T H O D S

Demyehnating lesions were made by microinjection of 1-2 /~1 of lysophosphatidylcholine (1%), mixed in normal Ringer solution (NS), mto the exposed sciatic nerves of 8-10-week-old female Wistar rats (29). The injection device consisted of a glass mlcroelectrode with a tip diameter of about 0.1 mm fixed to the needle of a 2-/~1 syringe. Five to 10 days after the injection, the animals were exsanguinated and a 2-cm segment of sciatic nerve containing the lesion was removed, desheathed, and placed in a modtried Krebs solution (NS) (in mM: 124 NaC1, 3.0 KCI, 2.0 MgSO 4, 2.0 CaCI z, 26.0 N a H C O 3, 1.8 NaH2PO 4" HEO and 10.0 dextrose) saturated with 95% oxygen and 5% CO2). This procedure is similar to that described in a previous paper25 Each end of the nerve was placed across b~polar silver electrodes that could be used for either whole nerve recording or whole nerve sumulation. The ends of the nerve were covered with petroleum jelly for electrical isolation, and the central region of the nerve was continuously superfused with NS or a NS containing 1.0 mM 4-AP. The distal end of the nerve was crushed between the recording electrodes so as to obtain monopolar recordings. Intra-axonal and extra-axonal recordings were obtained with glass mlcroelectrodes filled with a 3 M KC1 solution and amplified with a high impedence amplifier. Intra-axonal impalements could be verified by the presence of a resting potential and by modulation of spike amplitude by injection of hyperpolarizing current through the mlcroelectrode tip. Single axon recordings were obtained from nerve regions proximal or distal to the lesion site; both ends of the nerve were stimulated. This recording arrangement allowed comparison of conducUon properties followmg stimulation on either the proximal or &stal side of the lesion. Thus, conduction properties through the lesion could be compared with conduction properties m a normal segment of the same fiber. More details of the recordmg techniques can be found m ref. 16 In eight experiments designed to look for possible mter-axonal activation ('cross-talk'), the proximal 1.5 cm of the nerve was split creating a Y shape Each branch was laid across a bipolar electrode such that one branch could be stimulated while recordings

were simultaneously made both from the other branch and from the proximal (unsplit) end of the nerve. Neuromas were prepared as described in Kocsis et al. 15 for comparison using the same arrangement RESULTS

Conduction in the nerve segments proximal (Fig. 1A) and distal (Fig. 1C) to the lysophosphatidylcholine (LPC) lesion is relatively normal. However, responses recorded through the LPC lesion site are charactenzed by conduction slowing and impulse block. This can be seen m Fig. 1B 2 where an attenuated whole nerve response showing temporal dis-

prox.

demyel.

dist.

2 0

mV

20 msec

_

(

l

)

20 [mV

Cl (

[1

~

2

.

l

I

mV

'?

2

Fig 1 Whole nerve recordings of propagation along nerve segments proximal (A) to the LPC lesion, distal to the LPC lesion (C), and through (B) the LPC lesion The schematics m&cate the relatwe posmons of stimulating and recording electrodes D 1 single axon recordings were obtained at the end of an LPC les~on A short latency single axon response was recorded following proximal stimulation, and a long latency response was recorded following &stal stimulation when conduction was through the demyehnated zone D 2 &stal sumulatlon m a different axon gave no response at all, indicating conduction block at the site of the les~on The proximal stimulus resulted m a normal short latency response

A

B @

20

2.0

m@ec

ms@c

C ao

"~ 2q E

o o. o

. o:

._= 0

<

8 @ 4

eo ~e

•

• •

".. :.

•

•

•

2 I

I

2

l

i

4

I

i

6

Latency(ms)

20" ~o

Ftg 2 A 1 the trans-les]on response was often characterized by an early component and a late component (arrow). When pmred stlmuh were presented wnth an rater-stimulus interval of 3 0 ms (A2), the early component showed nearly complete recovery m amphtude, whde the late component remained refractory B upon high-frequency stnmulation (200 Hz), the late component showed decreased amphtude and increased latency, whde the early component followed the raptd stimulation wnth only minor changes m amplitude and latency B2 ~s at a higher panr to show the latency shift of the late component during high-frequency stimulation C. a scatter plot of the absolute refractory period vs latency taken from a series of single axon recordings m the vlcmtty of an LPC lesion The open circles show responses from normal nerves and the closed cnrclesfrom demyehnated nerves. perslon is present. In addition to the whole nerve recordings, single axon responses were recorded with glass microelectrodes in order to assess conduction properties of the demyehnated axons. Stimuli were delivered at both ends of the nerve segment, and recordings were made at a point proximal to the lesion. D~stal stimulation (conduction through the lesion) often resulted in a long latency response, while the

proximal stimulus resulted in a short latency response (Fig. 1DI). A n estimate of average conduction velocity for these two responses is 50 and 2.0 m/s, respectively. In Fig. ID2, are responses recorded from another axon in the same nerve. While a short latency normal response was obtained following proximal stimulation, there was no response followmg distal stimulation. These single axon recordings corroborate our observations from the compound response which in&cate conduction slowing and block. The whole nerve response of LPC-injected nerves often showed an early and a late component (Fig. 2A1). The early component may represent relatively unaffected axons, and the late component may represent the demyelinated axon population. This is in agreement with previous morphological studies 10, which indicate that not all of the axons are affected by the injections. The early and late components had different excitability properties as determined by paired stimulation experiments. When the interstimulus interval (ISI) was set to 3.0 ms, the fast component showed almost complete recovery in amplitude, but the slow component was nearly obliterated (Fig. 2A2), indicating a prolonged refractory period for the late acttvity. The ability of the late component to follow high-frequency stimulation was also less than that of the fast component. In addition to a greater amplitude reduction of the late component during high-frequency stimulation (Fig. 2B), greater conduct]on slowing of the late component occurred during repetitive stimulation (Fig. 2B2). The absolute refractory period (ARP) of the fiber was studied in greater detail using single axon recordings. A R P s of up to 30 ms were observed among single axon ,responses for propagation through the lesion. Fig. 2C is a scatter plot of the A R P vs latency from a series of axons from which single unit recordings were obtained. The open circles are from axons of a normal control nerve. The control nerve axons have relatively short A R P s and latencies. Some of the demyehnated axons (closed circles) had short latencies and short A R P s m a range comparable to those recorded from normal fibers (open circles). It may be that recordings of short latencies in the LPC-injected nerves represent either axons which were not in}ured by the LPC injection, or cases in which the stimulus and recording electrode are on the same side of the lesion area for a given axon. The population of fibers with

A

C

3.0

~,, v A . - . ~

~"

~'J

dist.

msec

D

B 1.0 mV

msec

2.01 mV

Fig 3 A a whole nerve recording from a mature normal nerve before and after (lower trace) 4-AP apphcat~on Note the shghtly mcreased late negativity after 4-AP. B. 4-AP application to a nerve with an LPC-induced demyehnatmg lesion resulted m a conszderable enhancement of the compound response (lower trace) C' schema for isolation of proximal, &stal and lesion s~tes of the nerve wzth petroleum jelly Pools were formed to isolate the proximal (1) and distal (3) nerve segments, and the leszonsite (2) D selective apphcation of 4-AP to the isolated pools on either side of the lesion site did not result m significant departure from the response recorded m normal solutmn (superimposed top traces), while direct application to the lesion site (pool 2) resulted in an enhancement of the compound response (lower trace) long latencies and long A R P s are a b n o r m a l and may correspond to axons comprising the late c o m p o n e n t of the c o m p o u n d response, i.e. demyelinated axons. The A R P was also studied m axons where a short latency single axon response was recorded from stimulation of one end of the nerve, but a long latency response was recorded from stimulating the other side. The short latency response had relatwely short A R P s (M = 0.86 + 0.6 ms, n = 15) and the longer latency response recorded from the same axon had longer ARPs (M = 4.7 + 3.5 ms; n = 15) This indicates that the excitability changes were localized to conduction within the les~on site and not along the entire course of the axon. In contrast to the effects of the potassmm channel blocker 4-AP on normal mature axons, where the effects are minimal (Fig. 3A), introduction of 4-AP into the bath of demyelinated axons resulted in an augmentation of the c o m p o u n d action potential (Fig. 3B). To test the specificity of 4-AP sensitivity to the les]on area, the proximal, distal and lesion sites of the nerve were isolated with petroleum jelly. W h e n 4-AP was applied focally to the proximal and distal por-

tlons of the nerve segment, only minor changes in the action potential were seen (Fig. 3D). However, when 4-AP was apphed directly to the isolated lesion site, a large increase in the area of c o m p o u n d response occurred (Fig. 3B). This indicates the effects of 4-AP are specific to the lesion site and not to the non-lesion region of the nerve. Although 4-AP led to a considerable e n h a n c e m e n t of the c o m p o u n d action potential following a single stimulus, recovery of the action potentml amplitude during double shock experiments was much delayed in the demyelinated nerves after apphcation of 4-AP. In Fig. 4A, paired stlmuh are delivered to a demyelinated nerve at interstimulus intervals ranging from 1 to 16 ms. In NS, the conditioned response amphtude was nearly completely recovered at an lnterstimulus interval of 14 ms. However, when the nerve segment was superfused with 4-AP, both the absolute and relative refractory periods increased, and gradual amplitude recovery of the conditioned response continued for more than 70 ms (Fig. 4B, C). Axonal cross-talk me&ated by electrical mteractlons has been demonstrated for axons entering ex-

A

4.0 msec

2.0ImV

B

C lo0

•

•

o o

o

o

oOo °

.o

• NS o 4-AP

o

i,r

o o 40

•

~

•

20'

o

Irlter-M|muiuIIInterval(ms) Fig. 4. A: the whole nerve response of a demyelinated nerve to paired stimuli delivered at interstimulus intervals ranging from 1 to 16 ms in the absence of 4-AP. B: the response of the same nerve to paired stimuli m the presence of 4-AP. Recovery of the late component of the acnon potential is delayed. C: graph of percent recovery of response amplitude of the condmoned response in a demyelinated nerve m the presence and absence of 4-AP The relatwe refractory period continued for more than 70 ms

perimental neuromalS.24. The question has been raised as to whether demyelinated axons show crosstalk. In order to test for cross-talk, the LPC-demyelinated nerves were removed, and the proximal segment leading into the LPC lesion was longitudinally split. One branch from the proximal end could be stimulated and the other used for recording (Fig. 5A). In normal nerve, no activity can be recorded from one branch to the other with such an arrangement~5. However, in the case of neuroma, a discrete response can be recorded from one branch following

activation of the other (Fig. 5B), indicating transneuroma activity 15. While the split LPC-treated nerves showed the typical compound response to proximal stimulation, i.e. conduction slowing and block, when recording was distal to the lesion (Fig. 5C1), no response could be recorded in the non-stimulated branch (Fig. 5C2), indicating an absence of cross-talk. Application of 4-AP to the demyelinated nerve segment, which should enhance the electrical activity in the lesion site and thereby increase the trans-lesion response, did not lead to the elicitation of a trans-lesion response (Fig. 5C3). We studied seven LPC-demyelinated nerves in this manner, and none showed branch to branch activation. 4-AP application had two main effects on action potential waveform. In one set of axons, a late depolarizing response developed that often gave rise to bursts of action potentials. This is evident from the intra-axonal recording in Fig. 6A. Note the large amplitude delayed depolarizing response with an abortive spike. The extracellular recording in Fig. 6D shows spikes recorded before (Fig. 6D1) and after (Fig. 6D2) 4-AP. A single spike was induced in NS, but a burst of three spikes occurred after 4-AP was applied. In other axons, 4-AP application led to broadening of the action potential. The spikes m Fig. 6B were recorded before and after (broader spike) 4-AP application. Demyelinated axons bathed in NS did not display spontaneous impulse activity. However, after the application of 4-AP, these same axons were capable of giving rise to spontaneous impulse activity (Fig. 6C) and multiple spike discharge (Fig. 6D). The axon shown m Fig. 6D gave rise to a steady rhythmic firing at 3 cps. 4-AP application also led to the overcoming of conductxon block in some axons. It was often observed that a proximal stimulus would elicit a spike that could be recorded with a microelectrode at the proximal edge of the lesion, but distal stimulation which would require conduction through the lesion failed to elicit a response, thus indicating conduction block (Fig. 6El). The spike in Fig. 6E1 was elicited by proximal (S1) stimulation in normal Ringer solution. Distal stimulation ($2) failed to elicit a response, indicating conduction block through the lesion. After 4-AP was applied, the spike width of the response elicited by proximal stimulation was increased, and the distal stimulus ($2) now was capable of eliciting a

A

branch1~.~

r i L~ 2"01mV n ~

2 branch2~ B

0.4ImV

4.0 rnsec

' O.11mV

3 4

Fig 5 A scheme showing longitudinal spilt of proximal nerve segment, and site of stlmulatzon (branch 1) and recording sites (distal to lesion and branch 2), In the case of neuroma, stzmulation of branch 1 leads to actwzty m branch 2 (B) In the LPC-treated nerve, a response was recorded distal to the site of demyelinatlon (Cl), but not m the other branch Even when the compound response was augmented by 4-AP treatment, no branch to branch actiwty could be recorded C4 is a basehne recorded without a stimulus to show background noise level at this high gain spike, thus indicating the overcoming of conduction block. Note the long latency of the response elicited by $2 stimulation. DISCUSSION T h e r e has been much interest in the effects potassium channel blocking agents such as 4 - A P have on conduction properties of d e m y e h n a t e d axons. Part of th~s stems from voltage-clamp observations on normal m a m m a l i a n nodes of Ranvler. These studies indicate that there is minimal v o l t a g e - d e p e n d e n t potassium conductance (g~:) at the normal m a t u r e mammahan node6, 7. F u r t h e r m o r e , Brismar 5 noted a relatively large potassium current in the alloxan diabetic rats where myelin is disrupted. Chiu and Ritch~eS,9 were able to voltage-clamp d e m y e l i n a t e d internodal m e m b r a n e m isolation of the nodes. They n o t e d a distract potassium current, but no sodium current after demyelinatlon. Their observations confirmed the proposal suggesting that the node of Ranvier may have primarily sodium channels, and that p o t a s s m m channels may be locahzed at internodal (submyelinic) axon regions 5-7. In a g r e e m e n t with this proposal, 4-AP has a minimal effect on action potential waveform of normal m a t u r e m y e h n a t e d fibers, but significantly alters the action potential waveform and firing characteristics of d e m y e l i n a t e d axons 2.20.21,30,32 in which internodal p o t a s s m m channels may be exposed

and therefore available to a potassium channel blocker. Bostock et al. 2 have shown that the blocking temperature of rat d e m y e l i n a t e d dorsal root fibers is raised after 4 - A P application. This taken together with the observation that normal m y e l i n a t e d axons are only modestly affected by 4 - A P led to the suggestion that 4 - A P may have possible s y m p t o m a t i c therapeutic value m demyelinating diseases2,7,23,25, 30 Clinical trials have been carried out to test this p r o p o sal12,23. Prior to testing on MS patients, 4 - A P had been used on patients with neuromuscular disorders, such as myasthenia gravis 17,1s. The rationale for its use in neuromuscular disease is that b r o a d e n i n g of the action potential should result m greater transmitter release, thereby leading to greater synaptic activation of muscleaL While some beneficial effects have been reported, it should be pointed out that a n u m b e r of undesirable effects such as seizures]8, 23 and sensory complaints (severe paresthesla and dysesthesla12,18) have also been described for patients given 4-AP. Results in this study from single axon recordings of demyelinated axons b a t h e d in 4 - A P provide information that may account for both the beneficial and nonbeneficial effects of systemic 4 - A P apphcation in patients. In some axons, conduction block was overcome following 4 - A P application. Two types of response change could occur after 4 - A P application' acUon potential b r o a d e n i n g or multiple spike dis-

A

m~ s.o

°

2.0

1

_

401 mV

mNc

E

NS

~

Ns

B

r ~

sl

2°1 my

s'2

C

O5

see

Fig 6. Apphcatlon of 4-AP resulted m several different changes m single axon response recorded from demyellnated nerve A' a large sustained delayed depolarization with an abortwe spike m an axon in the presence of 4-AP. B: broadenmg of the mtracellular response to stimulus following application of 4-AP. C. spontaneous spike activity recorded m the presence of 4-AP No spontaneous actw~ty was recorded in normal Rmger D' the single spike induced by a single stimulus before 4-AP apphcatlon (D 0 and the triple spikes mduced after (D2) 4-AP apphcatlon to the demyehnated nerve. E l' stimulation of the proximal end of the nerve resulted in a relatively short latency extracellular spike, but distal stimulation was unable to ehctt a sp]ke However, following 4-AP apphcatlon (E2) the d~stal stimulus now became effective m ehcmng a spike, thereby lndtcatmg the overcoming of conduction block

charge following a single stimulus. Restoration of conduction could occur m demyelinated axons that displayed either the broadening or the bursting effect after application of 4-AP 30. We did not determine the functional category of the axons we recorded from in this study. However, previous studies on sensory (dorsal) and motor (ventral) root fibers suggest that 4-AP-elicited action potential broadening is characteristic of m o t o r fibers, and that 4-AP-ehcited repetitive firing and burst activity is characteristic of sensory fibers 3,16. Furthermore, many of the demyelinated axons gave rise to spontaneous impulse activity in the presence of 4-AP. These hyperexcltability phenomena, i.e. spike bursting and spontaneous activity, could account for the sensory complaints such as paresthesia described by subjects gwen 4-AP

Several types of hyperexcitability phenomena have been described for demyelinated axons, and it has been suggested that these properties may account for the positive neurological signs in demyelinating disease. Increased mechanosensitive and spontaneous impulse activity have been noted in experimentally demyelinated sensory axons of rat spinal cord 28. Smith and McDonald 20 have suggested that Lhermitte's sign (paresthesia induced upon neck flexion in some MS patients) may be related to the increased mechanosensitivity of demyelinated axons. Additionally, it has been noted that nearly all MS patients have, at some point, had paresthesias 26. In our in vitro preparation of demyelinated axons, there is a conspicuous absence of spontaneous impulse activity. It is only after blockade of gK in demyelinated nerve with 4-AP that spontaneous activity becomes prominent. Indeed, part of the rationale for the use of 4-AP in MS patients is that at an appropriate concentratlon it may selectively affect demyehnated axon regions and not normal myelinated axons. Th~s is, to some extent, supported by our ammal experiments; there ~s a selective effect of 4-AP on the axon segment containing the demyelinating lesion site. In some peripheral nerve disorders, such as the buckthorn (Tullidora) neuropathyll, neuroma15.24 and amyelinated dystrophic axons 19, evidence has been presented for axonal cross-talk. Actwat~on of one axon, or set of axons, can lead to activation of neighboring quiescent fibers. The rationale for crosstalk In demyelinated axons is relatively strmghtforward; loss of myelin leads to short-clrcumng of axons and thereby increases the probability of direct interactions between the demyelinated axons. These interactions can be either electrical or iomc ]3.15 However, no evidence for cross-talk was observed in LPC-demyelinated axons studied in vitro in the present experiments. Demyelination following LPC injections presents a morphological picture that differs from amyehnated axons and other forms of demyelination where cross-talk was demonstrated. The fibers m the LPC lesion are separated from each other by myelin debris, cellular elements and endoneurial fluid within the endoneurial compartment. The relatively large volume of the endoneunal space could lead to spatial buffering whereby actwlty-dependent increases, e.g. in extracellular potassium concentration, would be small as compared to the concentra-

tion changes that might occur in a confined extracellular space such as in a tightly packed non-myelinated fiber tract 14. This would prevent, for example, crosstalk due to m e m b r a n e depolarization caused by such

dicate that u n d e r certam conditions 4-AP ts capable of overcoming conduction block in animal models of demyelination. The increased spontaneous impulse activity and the burst activity elicited by a single stim-

an accumulation of potassium. Furthermore, a rela-

ulus to the demyelinated axons in the presence of 4-

tively large extracellular (endoneurial) space would

A P described in this report may be the electrophysiological correlate of the sensory complaints reported

tend to reduce the likelihood of extracellular current being shunted across neighboring axon membranes. Smith and Hall 27 have noted that the security of conduction in demyelinated axons should be assessed in terms of refractory period. Although 4-AP application led to an increased area u n d e r the c o m p o u n d action potential and to restoration of conduction in at least some of the demyelinated axons, our results in&cate that the refractory period for whole nerve transmission was increased after 4-AP application. This implies that while 4-AP may lead to increased excitability and even restoratton of conduction through some s~tes of demyelinatlon, that the safety factor following another impulse may actually decrease after 4-AP application. This may be because of collision with spontaneous spikes, prolonged depolarization, or because axons that show restored conduction after 4-AP application have a lower safety factor. As m e n t i o n e d above, clinical studies have been carried out to study the effectiveness of 4-AP as a posstble symptomatic therapy in neuromuscular diseaselT, 18 and multiple sclerosis 2.12. The re-

m patients to whom 4-AP has been administered. Furthermore, while 4-AP application to demyehnated axons can lead to hyperexcitability p h e n o m e na, an increase in refractory period and therefore a reduction in excitability has also been noted. The effects of 4-AP on m a m m a l i a n demyelinated axons are diverse and not simply a prolongation of the action potential as might be predicted for gK blockade of some axons. In addition to action potential broadening, 4-AP can lead to multiple spike dtscharge, spontaneous impulse activity and alteratton m refractory periods. These heterogeneous effects of 4-AP on m a m m a l i a n myelinated axons should be considered in the rationale for the clinical use of 4-AP. ACKNOWLEDGEMENTS This work was supported in part by the N I H , the Medical Research Service of the Veterans Administration, and the Folger Foundatton.

sults of Bostock et al. 2 together with our results 30 in-

REFERENCES 1 Bostock, H and Sears, T A , Overcoming conductJon block in demyehnated fibres by prolonging acUon potentials, Nature (London), 274 (1978) 385-387 2 Bostock, H , Sears, T A and Sherratt, R M , The effects of 4-amlnopyndlne and tetraethylammomum sons on normal and demyehnated mammahan nerve fibres, J Physlol (London), 313 (1981) 301-315 3 Bowe, C M , Kocsls, J D and Waxman, S G , Differences between mammalian ventral and dorsal spinal roots m response to blockade of potassmm channels durmg maturation, Proc R Soc Lond Ser B, 224(1985)355-366 4 Brdl. M H , Waxman, S G , Moore, J W and Joyner, R W , Conduction velocity and spike configuration m myehnated fibres computed dependence on mternode distance, J Neurol Neurosurg Psychtatr, 40 (1977) 769-774 5 Bnsmar, T , Potentml clamp experiments on myehnated nerve fibers from alloxan dmbetlc rats, Acta Phystol Scand, 105 (1979) 384-386 6 Brlsmar, T , Potentml clamp analysis of the membrane currents m rat myellnated nerve fibres, J Phystol (London),

298 (1980) 171-184 7 Chin, S Y , Rltchle, J M., Rogart, R B and Stagg, D.A, A quanmatlve description of membrane currents m rabb~t myehnated nerve, J Physzol (London), 292 (1979) 149-166 8 Chin. S Y and Rltchle, J M , Evidence for the presence of potassium channels m the paranodal region of acutely demyehnated mammahan single nerve fibers, J Phys:ol (London), 313 (1981) 415-437 9 Chin, S Y and Rltchle, J M , Evidence for the presence of potassium channels in the mternode of frog myehnated nerve fibers, J Phystol (London), 322 (1982) 485-501 10 Gregson, N A and Hall, S.M , A quantltatwe analysis of the effects of the lntraneural mlecuon of lysophosphatldyl ehohne, J Cell Set, 13 (1973) 257-277 11 Hernandez-Cruz, A and Munoz-Martmez, E J., Tulhdora (Karwmskla Humboldmana) toxin mainly affects fast conductmg axons, Neuropath Appl Neurobtol, 10 (1984) 11- 24 12 Jones, R E , Heron, J.R, Foster, D H , Snolgar. R S and Mason, R J . Effects of 4-aminopyndme m patients w~th multiple sclerosis, J Neurol Sct , 60 (1983) 353-362 13 Kocs,s, J D , Functional organization of potassmm chan-

nels m normal and pathological mammahan axons. In J.C Zagoren and S Federoff (Eds.), Advances m Cellular Neurobtology, Vol 6, The Node of Ranvter, Academic Press, New York, 1984, pp 183-212 14 Kocsls, J.D., Malenka, R.C. and Waxman, S G., Effects of extracellular potassmm on the excltabdlty of the parallel fibers of the cerebellar cortex, J Phystol (London), 334 (1983) 225-244 15 Kocsls, J D., Preston, R.J. and Targ, E . F , Retrograde impulse acttvlty and horseradish peroxldase tracing of nerve fibers entering neuroma studied in vitro, Exp. Neurol, 85 (1982) 400-412. 16 Kocsls, J D , Rmz, J.A and Waxman, S G , MaturaUon of mammahan myelinated fibers, changes in action potentml characterization following 4-ammopyridme appficat~on, J. Neurophystol, 50 (1983) 449-463. 17 Lundh, H., Ndsson, O and Rosen, I., Treatment of EatonLambert syndrome' 3,4-dtammopyndme and pyndosttgmine, Neurology, 34 (1984) 1324-1330 18 Murray, M.B and Newson-Davls, J , Treatment with oral 4-ammopyndme m d~sorders of neuromuscular transmission, Neurology, 31 (1981) 265-271. 19 Rasmmsky, M., Ephapuc transmission between single nerve fibres m the spinal nerve roots of dystrophic mtce, J. Phystol (London), 305 (1980) 151-169 20 Rltchle, J M., Sodium and potassium channels m regenerating and developing mammalian myehnated nerve, Proc R. Soc Lond Ser. B, 215 (1982) 273-287 21 Ratchle, J M., Rang, H P and Pellegrmo, R , Sodium and potassium channels m demyellnated and remyehnated mammalmn nerve, Nature (London), 294 (1981) 257-259 22 Schauf, C.L and Daws, F A , Impulse conduction in multiple sclerosis, a theoretical basis for modification by temperature and pharmacological agents, J Neurol Neuro-

surg Psychtatr., 37 (1974) 152-161. 23 Sears, T.A. and Bostock, H., Conduction failure in demyelinatlon: is it inevitable 9 In S G Waxman and J.M. Ratchle (Eds.), Demyelinanng Diseases" Basic and Chntcal Electrophysiology, Raven Press, New York, 1981, pp. 359-375. 24 Seltzer, Z. and Devor, M., Ephaptlc transmission m chromcally damaged peripheral nerves, Neurology, 29 (1979) 1061-1064. 25 Sherratt, R.M., Bostock, H. and Sears, T A., Effects of 4-armnopyndme on normal and demyehnated mammalian nerve fibres, Nature (London), 283 (1980) 570-572 26 Shlbasakl, M., McDonald, W.I and Kurolwa, Y., Racial modification of chmcal picture of multiple sclerosis' comparison between British and Japanese patients, J Neurol. Sci, 49 (1980) 253-271 27 Smith, K.J and Hall, S.M., Nerve conducUon during peripheral demyelmatlon and remyelinatlon, J Neurol Sct., 48 (1980) 201-219. 28 Smith, K.J. and McDonald, W I , Spontaneous and evoked electrical d~scharges from central demyehnatmg lesions, J. Neurol. Sci., 55 (1982) 39-47. 29 Smith, M.E., Kocs~s, J.D. and Waxman, S . G , Myelin protein metabohsm m demyelmation and remyehnat~on m the sciatic nerve, Bram Research, 279 (1983) 37-44. 30 Targ, E F and Kocsls, J.D., 4-Ammopyridine leads to restoratton of conduction in demyehnated rat scaatlc nerve, Bram Research, 328 (1985) 358-361. 31 Thesleff, S., Ammopyndmes and synapac transmission, Neurosctence, 5 (1980) 1413-1419 32 Waxman, S G , Kocsls, J D and Eng, D . L , Ligature-reduced injury m peripheral nerve' electrophyslologlcal observatmns on changes m action potenttal characteristics following blockade of potassium conductance, Muscle and Nerve, 8 (1985) 85-92.