598 J. Phy8iol. (1960), 151,pp. 598-607 With3plate8 and4text-ftgure8 Printedin GreatBritain SUPERSENSITIVITY OF SKELETAL MUSCLE PRODUCED BY BOTULINUM TOXIN BY S. THESLEFF From the Department ofPharmacology, University ofLund, Sweden (Received 3 February 1960) In a chronically denervated mammalian skeletal muscle the entire membrane becomes sensitive to applied acetylcholine (ACh). About 4daysafterdenervation, thesizeoftheACh-sensitiveareaattheend-plate starts to increase and a few days later covers the larger part or all ofthe muscle membrane (Axelsson & Thesleff, 1959). The conversion ofthe membrane, following denervation, into an ACh- sensitive surface might be due to the absence of some chemical influence exerted when the motor innervation is intact. The purpose ofthe present investigation was to see whether or not the release ofthe chemical trans- mitteragent mightprovidesuchaninfluence. Usewas madeofbotulinum toxin, whichisconsideredtopreventrelease ofAChfrom cholinergicnerve terminals (Burgen, Dickens & Zatman, 1949; Brooks, 1956). This mode of action of the toxin was confirmed, and it will be shown that when the transmitter output was reduced or abolished the size ofthe receptor area in muscle started to increase in a manner which was identical with that observed after denervation. METHODS Unless otherwise stated the experiments were made on the isolated tenuissimus muscle of the cat. The sensitivity ofindividual muscle fibres to ACh was determined by ionto- phoretic release ofthe drug from the tip ofa micropipette as described by del Castillo & Katz (1955). When the tip of the pipette was close to the receptor structure, the ACh releasedbyacurrentpulseof10msecdurationproducedatransientmembranedepolariza- tionofafewmillivoltsamplitude. Thispotential change wasrecordedwithaconventional capillary micro-electrode inserted into the muscle fibre close to the point of drug appli- cation. Fordetailsoftheexperimentalset-up,techniquesfordrugapplicationandrecording see Axelsson &Thesleff(1959). The end-plate region ofindividual musclefibres waslocated bypursuingfine superficial nervetwigsandbytheappearanceofminiatureend-platepotentials(m.e.p.p.s) witharapid time course. The mean frequency of m.e.p.p.s was calculated fromrecordingsmadeover severalminutesonmovingfilmorwithanink-writer. Thepartofthe musclemembraneat whichACh,whenreleasedfromamicropipette, producedadepolarizationwithalatencyof less than 10msec was considered sensitive to the drug. Its length was measured with a binoculardissectingmicroscopeandaneyepiecescaleat80x magnification. SUPERSENSITIVITY OF SKELETAL MUSCLE 599 A powdered preparation ofCl. botulinum toxin typeAwithamouseLD50of005ug/kg wasused. 1mgofthetoxinwasdissolvedin1ml.ofsterilephosphatebtifferasdescribedby Ambache (1949). Furtherdilutionsofthetoxinweremadefromthisstocksolutionimmedi- atelybefore use. Afreshstocksolutionwaspreparedforeach day'sexperiment. Thetoxin,inamountsiangingfrom0.01to15 yg,waseitherappliedtotheexposedsurface ofthetenuissimus muscle orinjected individed amountsintothe musculature ofthehind leg. With these doses and modes of administration the action of the toxin was confined mainlytothesiteofapplicationandgeneralizedintoxicationswereusuallyavoided.Twoto four weeks after the administration ofthe toxin the tenuissimus muscle was removed in pentobarbitone anaesthesia. In experiments on frog (Rana temporaria) the toxin was injectedundertheskin oftheventralsurface ofthethigh. Forexaminationoftheultrastructure ofmotornerveterminals anumber ofbotulinum- intoxicated and control muscles were removed and fixed at 40 C for 1hr in 1% osmium tetroxidesolutionbufferedwithveronal-acetatetoaboutpH7.5,accordingtotheprocedure describedbyPalade (1952). Afterfixationthetissuewasdehydrated inethanolandsmall pieces of tissue containing end-plates were cut out. The specimens were stained in 1% phosphotungstic acid in absolute alcohol and embedded in 'Araldite' according to the methodofGlauert & Glauert (1958). RESULTS The ultrastrudure of motor nerve terminals in botulinum-intoxicated muscles. It has previously been found by R. Thies (personal communi- cation) in acute experiments on isolated serratus anterior muscles ofthe guinea-pig, that the paralysis produced by bath application ofbotulinum toxin was not accompanied by structural changes in the presynaptic vesicles. It seemed possible, nevertheless, that chronic intoxication might inducedegenerative changesinthenerveendingssimilartothosedescribed by Birks, Katz & Miledi (1960) during Wallerian degeneration. However, no structural abnormalities were foundwiththe present technique inend- platesfromparalysedfrogandcatmuscles, evenafterperiodsof3-4weeks. Examples are shown inPls. 1-3, takenfrom afrog's sartorius 13 days and from a cat's tenuissimus 27 days after botulinum injection. During the experimentsneitherofthesemuscles showedany m.e.p.p.s or responses to nerve stimulation, and the cat muscle had become supersensitive to ACh all along its length. If one compares the micrographs with those from normal end-plates (for frog muscle, see Birks, Huxley & Katz, 1960; for mammalian muscle; seeReger, 1958, andAndersson-Cedergren, 1959), the internalstructureofthenerveendingsandtherelationsoftheirmembranes to the muscle fibre and its junctional folds appear to be quite unchanged, nor is there any difference in the size and spatial distribution ofthe pre- synaptic vesicles. Thepresent methodwouldnot, ofcourse, reveal changesofamuchmore minute nature. Thus to demonstrate thelocation ofthe toxinmolecules or to show up slight changes in the structural detail of the terminal nerve membrane would require a more powerful technique. 600 S.THESLEFF Effects ofbotulinum toxinontransmitterrelease. As described by Guyton & MacDonald (1947) it was observed that the neuromuscular block pro- duced by a single dose ofbotulinum toxin reached its maximum in about 5 days and thereafter remained at a constant level for a period of several months. In amounts exceeding 1 ug the toxin completely abolished transmitter release from motor nerve terminals in the tenuissimus muscle of the cat. In such muscles spontaneous m.e.p.p.s were normally absent and end- plate potentials (e.p.p.s) were not recorded following nerve stimulation. I I I I II/ I I I I 1-1 I I I I / I I III I /I I' 1'.1 .1 I 2 mV 'TI,Wwlr III LL1111. I I I I ltil I I I I I I I I I I I t A A 100 msec Text-fig. 1. The insertion of the tip of the micro-electrode into the end-plate region of a single muscle fibre gave rise to the burst of m.e.p.p.s shown in the upperinkandcathode-rayoscilloscoperecordings.Twominuteslatertheelectrical activityattheend-platehadalmostsubsided (lowerrecord). Therecordsarefrom anexperimentmadeonthetenuissimusmuscleofacatwhich 3weekspreviously had received an injection of 2jig ofbotulinum toxin into the hind-leg. Neuro- muscular transmission was completely blocked and transmitter release was not observed in any other end-plate. Calibration of ink recordings: upper record, 3squares = 1mVand 1-5sec; lowerrecord, 3 squares = 1 mV and 15sec. However, the insertion of the micro-electrode occasionally gave rise to a short-lasting burst of m.e.p.p.s at a high frequency of discharge (Text- fig. 1). Apparently this was due to mechanical injury to nerve terminals causedbythetipoftheelectrodewhenitpenetratedthemusclemembrane in the end-plate region. With smaller quantities of the toxin (0 05-1 tg) m.e.p.p.s with a fre- quency of discharge about 100 times less than in a normal muscle were observed (mean intervals were of the order of 10-100 sec instead of 0-1-1 sec). In confirmation ofearlier investigations (Brooks, 1956) it was foundthat the amplitude andtime course ofthese m.e.p.p.s were roughly the same as those recorded at a normal end-plate. In these muscles nerve stimulation was either ineffective orproduced an e.p.p. ofa fewmillivolts amplitude.AtsucharelativelylightstateofintoxicationtheadditionofCa2+ tothebathingfluidinaconcentration oftwicenormalincreasedthetrans- SUPERSENSITIVITY OF SKELETAL MUSCLE 601 mitteroutputbyanerve stimulus (Text-fig. 2). Tetraethylammonium ina concentrationof1 mmhadasimilarfacilitatory action (cf. Koketsu, 1958). Katz & Miledi (1959) made the interesting observation that in de- nervated frog muscle spontaneous subthreshold potential changes re- appew at the end-plate region several days after cessation of activity. This renewed activity resembled the normal m.e.p.p.s except that the average frequency of discharge was much lower and that the amplitude distribution was wider. Probably end-plate potentials which occur in denervated frog muscle are produced by quanta of ACh released by the Schwann cell, as suggested by Birks, Katz & Miledi (1959). I2 mV 5 msec Text-fig. 2. The addition ofa twice normal concentration ofCa2+ to the bathiing fluid resulted in an e.p.p. in response to a nerve stimulus (right-hand record). Before the increase in the external [Ca2+], nerve stimulation had no effect (left- hand record). Since the origin of ACh release in a chronically denervated junction presumably differs from that in an innervated one it was ofinterest to see whether botulinum toxin would also prevent this type of spontaneous activity. In frogs kept at room temperature the sartorius muscle was de- nervatedonone sideand 5 dayslater 2 ,ug ofbotulinum toxinwasinjected into each leg. This amount of toxin produced a complete paralysis of all leg muscles. Two weeks later the innervated as well as the denervated sartorius muscle wasremoved, andboth wereexaminedforthepresence of spontaneous electrical activity at the end-plate region. As shown in Text-fig. 3, aslowrate ofspontaneous discharge was observedatend-plate regionsinadenervatedandbotulinum-treatedmuscle. Inthemusclewith an intact nerve, however, the same amount of toxin had completely 602 S. THESLEFF abolished m.e.p.p.s. Spontaneous AChrelease in a chronically denervated junction appears consequently to be less affected by the action of botu- linum toxin than the ACh release which in the innervated end-plate originates from motor nerve terminals. ACh-sensitivity of botulinum-intoxicated muscles. When transmitter release is blocked by the use of botulinum toxin the fibres ofthe tenuis- simus muscle become sensitive to applied ACh along their entire length. --F -1 '11 -1 l 1 I I. i J f A XI 1Sa I-I. I o i 50 msec Text-fig. 3. Spontaneous subthreshold activity in a chronically denervated and botulinum-intoxicated frog end-plate (for explanation see text). Calibration of upperrecord: 3squares = 1mVand 15sec. One to two weeks aftertheadministration ofthetoxin, AChreleased from the tip ofamicropipette produces depolarizationswitharapidtimecourse wherever it is applied to the muscle membrane. The whole surface ofthe muscle fibre becomes as sensitive to ACh as the end-plate region, which maintains itsoriginalresponsiveness tothe drug (Text-fig. 4). Theuniform sensitivity of the muscle membrane to ACh is similar to that which has previously been observed in chronically denervated mammalian muscles (Axelsson & Thesleff, 1959). As in a chronically denervated tenuissimus muscle, ACh produces in a botulinum-intoxicated one a graded and 'electrically silent' contracture. With a small quantity ofthetoxin (0.01-0.05 [kg) itispossible to obtain tenuissimus muscles in which the ACh-sensitive surface of individual musclefibresvaries insize. Thisisillustrated inText-fig. 4. In somefibres the membrane was uniformly sensitive to ACh over at least 3 mm, as is shownintheupperrecords. Inanothermusclefibre (middlerecords)about SUPERSENSITIVITY OF SKELETAL MUSCLE 603 2 mm was sensitive to applied ACh, with the highest sensitivity at the end-plate. In a third fibre (lower records) only the end-plate region responded to ACh. When spontaneous m.e.p.p.s were completely absent or when they appeared at a very low rate (< 01/sec) the whole fibre was invariably sensitive to ACh (Text-fig. 4, Table 1). With a higher rate of m.e.p.p. dEischarge E(01-1-0/sEec) the reEceptor suErface wasEenlargeEd as compoared to 2mV f * 03/sec -_ - 100 c/sec 2-4/sec 0 05 1.0 1 5 2-0 25 30 mm End-plate region Text-fig. 4. In a tenuissimus muscle, intoxicated 3 weeks earlier with a small amount of botulinum toxin, muscle fibres were observed in which the ACh- sensitivesurfacevariedinsize.ThesensitivityofthemembranetoAChwastested by iontophoretic micro-application ofthe drug. The membrane potential ofthe fibre is recorded in the upper tracing of each record and the current passing through the pipette in the lower tracing. The fibre used for the upper records wasuniformlysensitivetoappliedAChoveradistanceofatleast 1-5mm at each sideoftheend-plate. ThesizeoftheACh-sensitive surfaceintwootherfibreswas smaller (middle and lower records). The frequencies at which m.e.p.p.s occurred are shownbythe figures tothe right ofthe records. normal, whereas at a frequency above 1-0/sec only the end-plate was sensitivetoappliedACh. Table 1 showstherateofm.e.p.p.'sandthemean length of the ACh-sensitive surface in fifty muscle fibres from seven tenuissimus muscles intoxicated 3-4 weeks previously with botulinum toxin. (For technical reasons distances longer than 2 mm could not be measured accurately in single fibres.) The results shown in the table 604 S. THESLEFF indicate that, despiteindividualvariations, arelationexistsbetweentrans- mitter release and the extent to which a muscle fibre becomes sensitive to applied ACh beyond the end-plate region. TABLE 1. Therelationship betweenthe frequencyofspontaneous m.e.p.p.s and the length ofthe part ofthe muscle fibre whichwas sensitive to applied ACh. The experimentswvere madeonseventenuiimusmusclesintoxicated3-4weekspreviouslywithbotulinumtoxin. Figureswithinbrackets denotethe range ofmeasurements Meanlength of Frequency of No. of ACh-sensitive area m.e.p.p.s (sec-1) fibres (mm) 0.01-0.1 16 2-0 (2.0) 0'1-l*0 8 1*6 (1-0-2*0) > L*0 26 0-8 (0.3-2.0) DISCUSSION Two observations are ofparticular interest inthe present investigation: first, the finding that botulinum toxin prevents transmitter release from motornerveterminalswithoutalteringtheirultrastructure, and, secondly, that this action causes a change in the chemical sensitivity ofthe muscle similar to that produced by chronic denervation. It is characteristic ofbotulinum poisoning that there is a lack oftrans- mitter release from cholinergic nerves, while in other respects nerve and muscle are unaffected by the toxin. The mechanism of this action is not understood, but it is conceivably localized to the nerve terminals (Guyton & MacDonald, 1947; Burgenetal. 1949; Brooks, 1954, 1956). Theresultsof thepresentinvestigation donot explainthemodeofactionofthetoxinbut they suggest, by excluding the possibility of a morphological injury to presynaptic structures, that it is the ACh mechanism in nerve endings which is affected. The number and size ofpresynaptic 'vesicles', supposed to contain the quanta ofAChreleasedbythenerve (cf. delCastillo & Katz, 1956), appear to be normal in botulinum-treated muscles. That the terminals still con- tained packets of multimolecular quanta of ACh was disclosed when a mechanical injury to the end-plate produced a burst of m.e.p.p.s. ACh formation seems to be unaffected by the toxin. Burgen et al. (1949) and Stevenson & Girvin (1953) have shown that botulinum toxin does not interfere with choline acetylase systems, and in the present investigation it was observed that spontaneous ACh release from a chronically de- nervated amphibian junction was unimpaired by the toxin. The results of the present and previous investigations thus indicate that the toxin has a selective mode of action and that its point of attack is the mechanism responsible for the release ofthe chemical transmitter agent from cholin- ergic nerves. SUPERSENSITIVITY OF SKELETAL MUSCLE 605 In chronically denervated muscles the whole membrane becomes sensi- tive to applied ACh. This increase in size ofthe receptor surface can, as was shown byAxelsson & Thesleff (1959), account forthe supersensitivity ofdenervated muscles to ACh and other chemical substances. In botulinum-intoxicated muscles the spreading of the ACh-sensitive area ofthe end-plate to the entire membrane occurs in a similar manner and with about the same time couarse as in denervated muscles. Conse- quently,-it is likely that the cause ofthe receptor change is the same in bothinstances. Areasonable assumption is thatindenervated-muscles, as in botulinum-treated ones, lack of transmitter release and nQt nerve de- generation is the cause which initiates the increase in size ofthe receptor surface. That supersensitivity following denervation is caused by the dis- appearance oftransmitter release from the nerve has also been proposed by Burn & Rand (1959), who studied smooth muscles deprived of their sympathetic innervation. Of interest was the observation that a diminished transmitter release, i.e. a lowrate ofm.e.p.p. discharge, produced an enlargement ofthe ACh- sensitive area at the end-plate. This suggests that the size ofthe chemo- sensitiveregion ina muscle fibreisvariableandisregulatedbytheamount and frequency oftransmitter released from the motor nerve. There' ate no'clues as to the nature ofthe mechanism by which trans- mitterreleasemightaffectthesizeofthereceptorsurfaceinamusclefibre. A possibility is that the permeability increase produced in the end-plate membrane-bythereleasedAChallowsachemical agent'toenterthem'uscle fibre and that the presence ofthis substance prevents ACh receptors from beingformedoutsidetheend-plateregion. Alternatively, thepermeability increasemaypermittheefflux ofasubstance, formedinthemyogenic part ofthe end-plate, whose presence inside the fibre induces the formation of ACh receptors. This latter possibility could explain the spatial'sequence ofthe change in size ofthe receptor surface following denervation and re- innervati'on. This hypothesis is speculative but has at least the advantage that it can be experimentally tested. For example, it might be possible to determinewhetherlong-lasting curarizationproduces changes inmuscle similar to denervation. SUMMARY 1. In mammalian skeletal muscle intoxicated by botulinum the entire muscle membrane becomes sensitive to applied acetylcholine. One to two weeks afterthe administration ofthetoxinthewhole surfaceofthemuscle is uniformly sensitive to acetylcholine. The spread ofacetylcholine sensi- tivity from the end-plate to the whole membrane occurs in a similar manner and with about the same time course as in a chronically de- nervated muscle. 606 S. THESLEFF 2. Botulinum toxin prevents transmitter release from motor nerve terminals without altering their ultrastructure. The present results indicate that the toxin acts by blocking the mechanism responsible for transmitter release from cholinergic nerve endings. 3. It is suggested that transmitter release from motor nerve terminals determines the size of the acetylcholine-sensitive area in the post-junc- tional membrane and that lack of transmitter agent and not nerve degeneration is responsible for initiating the process which leads to a high and uniform chemo-sensitivity in chronically denervated or botulinum- treated muscles. Thestudyoftheultrastructure ofmotornerveterminalswasmadebyProfessorB.Katz, Department of Biophysics, University College London, who has kindly permitted me to publishhisresults.TheexpensesofthisinvestigationwereaidedbygrantsfromTheMuscular DystrophyAssociationsofAmerica,Inc.andtheAirResearchandDevelopmentCommand, United States Air Force, through its European Office. I am indebted to DrJ. Keppie, of theMicrobiologicalResearchStation, Porton,forageneroussupplyofbotulinumtoxinand toxoid. Unfailing technical assistance wasprovidedbyMiss E. Adler. REFERENCES AMBACE,N. (1949). TheperipheralactionofCl. botulinumtoxin. J.Phy8iol. 108,127-141. ANDERssON-CEDERGREN, E. (1959). Ultrastructure ofmotor end-plate and sarcoplasmic components ofmouseskeletalmusclefibre. J. Ultrastructure Res. Suppl. 1, 1-191. AXELSSON,J.&THESLErFF,S.(1959). Astudyofsupersensitivityindenervatedmammalian skeletal muscle. J.Physiol. 147, 178-193. BIRKs, R., HUXLEY, H. E. & KATZ, B. (1960). The fine structure ofthe neuromuscular junction ofthefrog. J. Phy8iol. 150, 134-144. BIMKS, R., KATZ, B. & MIEDI, R. (1959). Electron-microscopic observations on de- generatingnerve-muscle junctions ofthefrog. J.Physiol. 146, 45-46P. BIRKS, R., KATZ, B. & MILEDI, R. (1960). Physiological and structural changes at the amphibian myoneural junction, in the course ofnerve degeneration. J. Phy8iol. 150, 145-168. BRooKs,V.B. (1954). Theactionofbotulinumtoxinonmotor-nervefilaments. J.Physiol. 123, 501-515. BROOKS,V.B.(1956). Anintracellularstudyoftheactionofrepetitivenervevolleysandof botulinum toxin onminiature end-plate potentials. J. Phy8iol. 134, 264-277. BURGEEN, A. S. V., DICKENS, F. & ZATMAN, L. J. (1949). The action ofbotulinum toxin on theneuro-muscular junction. J. Physiol. 109, 10-24. BURN, J. H. &RAND; M. J. (1959). The cause ofthe supersensitivity ofsmoothmuscleto noradrenaline aftersympathetic degeneration. J. Phy8iol. 147, 135-143. DEL CASTILLO, J. & KATZ, B. (1955). On the localization of acetylcholine receptors. J. Phy8iol. 128, 157-181. DEL CASTILLO, J. & KATZ, B. (1956). Biophysical aspects ofneuro-muscular transmission. Progr. Biophys. 6, 121-170. GLAUE;RT, A.M. & GLAUERT, R.H. (1958). Aralditeasanembeddingmediumforelectron microscopy. J. biophys. biochem. Cytol. 4, 191-194. GUYTON, A. C. & MACDONALD, M.A. (1947). Physiologyofbotulinustoxin. Arch.Neurol. Psychiat., Chicago, 57, 578-592. KATZ,B.&MILEDI,R. (1959). Spontaneoussubthresholdactivityatdenervatedamphibian end-plates. J. Physiol. 146, 44-45P. KOKETSU, K. (1958). Action of tetraethylammonium chloride on neuromuscular trans- mission infrogs. Amer. J. Physiol. 193, 213-218. The Journal of Physiology, Vol. 151, No. 3 Plate 1 ra5. 'Kr' I.II't*,t, . P%''., 4#.4 A Mi' kQ.,1 e-#L1-...aI k XS,.,# - ~~~_'~~.1~~. a *,4...1 1 F zt 7 :. lix f im *'MA 4' .. ,j '0. 0t,,,0..4-F..1Af%t aI Li,.4I. ,I S.W,.,, s: !,. .i -' N Io , I I I .:. -. A.. S. THESLEFF (Facingp. 606)