curtis cole 1942 ann p3

8/3/2019 Curtis Cole 1942 Ann p3

1/10

3IEltBRANE RESTING A N D ACTION POTENTIALSF R O N THE SQUID GIANT AXON

H O W A R D J. CURTIS AND KENNETH S. COLEUcpart t t irr i t of Plr yn io logy , Th e Joliits H o p k i n s UniceTeity Bchool of Medic in e , Bo l t i tn or e ,Maryland; t h e D e p a r t m e n t of P l i p i o l o g y , Col lege of P h y s i c i a n s and Surgeons, ColzrsibiaU n i u w s i t y , N ew Y o r k ; and the 2 l a , . i n t B i o l o g i c a l L a b o r a t o r y , W o o d s H o l e, Y a . w a c l i t i s e t t s

F O U R FIGURES

IRTRODUCTIONThe resting aiid action potentials of nerve and iriuscle have usually

been measured between a illed-end and the region being investi-gated. These potentials would be the actual potential differences acrossthe membrane i f there were no external current flow and if the mem-brane potential at the killed-end had been reduced t o zero, but suchconditioiis a re difficult to obtain experimentally. Th e potential dr opdeveloped by current flow in the transition region between the activean d inactive portions reduces tlie measured resting potential below theiiiembraiie re st ing potential and tlie short-circuiting in the neighborhoodof the active electrode reduces the measured action potential si nd ar ly ,Cole and C urt is ( 38 ) . Also, the impulse may contribute to the nieasurcdaction potential tis it is extinguished in the transition region. Themembrane potential of the injured region can certainly be eliininatedbut there is n o direct measurement to indicate it has been reduced tozero.

Tlie introduction of tlie squid giant axon prep ara tion by Young ( 3 6 ) ,iiiade it possible to considerably reduce the er ro rs in the killed-endtecliriique of measurement. B y working with well-cleaned fibers sur-rounded by as little sea-water a s possible, Cole an d Hodgkin ( XI), andWebb and Young (40),obtained action potentials of over i 0 niilli-volts and Steinbach ( 40) , recorded r estin g o r in ju ry potentials a s liiglias 65 millivolts. It became possible to avoid most of the difficulties ofthe killed-eiid77 technique by inser ting a micro-electrode inside thisaxon aiid measuring these potentials directly across the active cellmembrane. This w a s done at Plymouth by Hodgkin and Huxley ( 3 9 ) ,with a very small silver-silver chloride electrode apparently in directcontact with thc axoplasm; and a t Ko ods Hole by Curtis and Cole ( $ O ) ,with a micropipct filled with KCI. At both laboratories action poteii-

135JOURNAL or c> . i .Lw AN D rnxxU(.&mmPnrsioi OGI. VOI . . 19. so. 2

4il tI l . , 1942


2/10

136 H. J. CURTIS AND 6. . C OLEtials as large a s 90 millivolts were found. Hodgkin and Huxley furtherreported measurements of the resting potential of about 50 millivolts.W e could not obtain any reliable resting potential measurements be-cause of electrode difficulties.

This finding of an action potential larger than the resting potentialwas quite unexpected because it showed that the potential differenceacross the membrane was actually reversed during the passage of theimpulse. I n these experiments the electrical characteristics of themicro-electrodes and the amplifiers were seen to be unusually iniportaittfor an accurate record of the action potential. Also the contributionof the silver-silver chloride electrode to the resting potential mighthave been rather large because of the low chloride ion concentrationof the axoplasm. Since the membrane potential reversal depended uponthe difference between these two measurements, it seemed advisableto attempt to improve the technique for each.

I n conjunction with these measurements, preliminary data were takenof the effects on the membrane action and resting potentials of chang-ing the concentration of several of the ions in the external medium,particularly potassium.

METHODThe material used was the giant axon from the hindmost stellarnerve of the North Atlantic squid, Loligo pealii. The method of dissec-

tion and teasing was essentially the same as previously described(Cole and Curtis, '39). The nieasuring cell was similar to the one usedpreviously (Curtis and Cole, '40) and is shown in figure 1. The axonfitted snugly in a trough 540 1 square cut in the top of a sheet of poly-styrene, with the tied ends of the fiber in the enlarged portions at Aand B. The trough was filled with sea-water and covered with a corerslip. The micro-needles were about 15mm. long and about 4 O p indiameter, and were pulled st raight from the glass tubing which formedand shank of the needle. This shank had an outside diameter of nearly540 ~1 and slid easily in the trough on both sides of C. After beingpulled the needles were filled with KC1 isosmotic with sea-water andstored in this solution.

With the axon in place, the needle was placed in the end of the troughat C and pushed along until its point encountered the axon where itentered the enlargement at B. A quick push usually sufficed t o pierccthe cell wall, and the needle was then pushed along until its tip wasjust opposite the capillary of the E electrode. As soon a s the axon wasplaced in the cell fresh sea-water was siphoned ill through one of thetubes at A and out the tube at B and a constant circulation maintained.


3/10

MEMBRANE POTICXTIALS 137The electrode wells at E, F, (3, and H contained silver-silver chloride

electrodes and were kept filled with sea-water. Small glass capillaries,filled with sea-water, joined these electrode wells with the trough. TheD electrodes were used for stimulation, and the action potential couldbe recorded between any pair of the other electrodes. A vaseline sealwas placed around the needle between B and C, so that the H electrodecould only respond to potential changes near the tip of the needle.The other electrodes were also well insulated.

Fig. 1 Schematic drawing of the cell used for ineasuiing membrane action a id restingpotentials. A clean a xon was placed in the trough w ith its ends in tlie enlargement at Aand B. The micropipet, filled with KCl, was placed in the trough with its tip in th e B eii-largement, and, after piercing the ccll wall, was puslied dong to the position shown. Allchambers were then filled with sea-water, and covered with a g1a.w plate. Fres h soa-waterwas siphoned in t he tubes a t A and out tlic tuhe at B. The axon wa(l stimulated at tlie Delectrodes, and membrane poteutials measured between the E and H electrodes.

Since the resistance of the needles was 5 t o 20 megohms, small ex-ternal capacities introduced relatively large errors in the size and shapeof the action potential. These were minimized by an impedancechanger placed close to the measuring cell. There remained, however,the capacity of the needle itself, and a variable inductance in theamplifier was used to compensate for its effect. Before an experimentwas started, the needle to be used was placed in the measuring celland the inductance set for the optimum response t o a potential suddenlyapplied to the needle tip. In this way, needle action potentials wererecorded with a response time of 10 microseconds or less.

The amplifier was direct coupled and drew an input current of lessthan lo- amperes and therefore resting potentials could be measureddirectly on the face of the cathode ray oscillograph tube. However, itwas found that the small silver-silver chloride electrodes which werenecessary for action potential measurements did not maintain as con-


4/10

138 H. J. CURTIS AND K. S . COLEstant a potential as was desired, so a pair of large silver-silver chlorideelectrodes was used. These electrodes were sealed in glass tubes filledwith sea-water and made contact with the electrode w7ells by means ofwicks. They maintained a constant potential t o within 0.5 millivolt formany hours. A string soaked in sea-water was placed between theE and H electrode wells to short-circuit the re sting potential and estab-lish the zero for the res ting potential measurements. This was donebefore and after each resting potential measurement.When the medium sur rounding the fiber mas changed the fluid wassiphoned in through one of the tubes at A and out the other, thusquickly flushing out this chamber, while the flow out the tube B con-tinued. In this way it was usually possible to change the fluid surround-ing the entire fiber in less than 15 seconds.In order to vary the potassium ion concentration, solutions isosmoticwith sea-water were made in which the calcium ion concentration wasthe same as that of sea-water, and potassium was substituted for sodiuiii,or vice versa. All solutions were buffered with phosphate buffers a tpH 8.03, the p H of Woods Hole sea-water. I n making a potassium coii-centratioii curve, a sea-water value was taken, then the solution waschanged and kept circulating until the potential became constant(usually 2 to 3 minutes), and the eolution immediately changed backto sea-water. All effects repor ted here were entirely reversible. Whenthe resting or action potentials did not r eturn to the ir original valuesaf te r the fiber was returned to sea-water, the results were discarded.Results obtained from axons in which the potentials did not remainconstant for at least an hour after impalement were discarded. It wasfurther required that the action potential recorded between the F andG electrodes be substantially the same as that measured between theE: and G electrodes at the time the measurements were made. In all,fifteen axons fulfilled these requirements. Seventeen others showedsubstantially the same characteristics but were not sufficiently constantand reproducible.

RESULTSThc restiiig potentials of these fifteen axoils varied from 46 to 59 milli-volts, and averaged 51 millivolts. The potential of any given ax011woulcl remain remarkably constant and one axon maintained a coilstantpotential for more than 8 hours. Furthermore the potentials ~ o u l d

not fall appreciably for at least half an hour after the axon becaiiieinexcitable. No correlation was observed between the magnitude ofthe resting potential and the length of survival.


5/10

MEMBRANE POTENTL4LS 139The eirect on the resting potential of changing the potassiuiii conccn-

tration while the calcium concentration w-as maintained constant isshown in figure 2 where the resting potential is plotted against thepotassium ion concentration f o r a typical experiment. All curves takenshowed the same f o r m ; they tended to become parallel t o the concentra-tion axis at low values of concentration, and to approach a straight lineat high concentrations. The slope of the line a t high concentrations isabout 50 millivolts for a tenfold change in concentration. It will benoted that the potential is zero at a concentration of about 18 timesthe normal, and is reversed in sign by about 15 millivolts at 40 times(isosmotic KC1). The same effect of potassium was observed whenboth calcium and magnesium concentrations were held constant.

-Z O P

I 1W 10KI0.1

L C UM POTASSUM CWCLNTSATPNQ

F i g . 2 Resting poteirtiul in mil l ivo l ts cs. potassium Concentration of the surrounding fluid.Tlic concentrat ion scale is in mdtipka of the potassium concentration of sea-water, 13 i d h -~uolar, slid is logaritlrmie. A t high potnasium conrentrntions the curve is I traight l ine,th e slope of which is nenrly that of the potassium elcc!rodc. I n the physiological rnnge ofconcentrat ions the potent ial i s ncarly independent of th e conerntrstion.

Action potentials recorded with tlic needle electrode are somewbatdiphasic and tend t o be oscillatory (Curtis and Cole, '40; Hodgkiri aiiclHuxley, '39) with the first positive phase about 15% of the spike lieight.The maximum negative variation from the resting potential (spikeheight) varied from 77 t o 168 millivolts, and averaged 108 millivolts.A typical action potential record is shown in figure 3.

Action potentials were also recorded as a function of potassiuiii coil-centration, and the spike height is plotted in figure 4 as a function ofthe potassium concentration. The fibers ceased to conduct at a potas-sium. concentration between 2.5 and 6 times normal. N o suddcxi cessa-


6/10

140 H. J. CURTIS AND Ii. S. COLE

Fig. 3 Membrane action potential of the squid axon. The tw o horizontal traca aro 511iiiillivolts apart. The reating potential in this axon was 58 millivolta. Thus the upper Iiori-zontal line approximately represents zero potential difference acrolul the membrane, the lowrrline the resting potential (outeide ponitive) and th e action potential, sta rtin g from the reetingpotential, swings to 110 millivolts (outside negative) . Time interval s a t t,lie bottom arcs0.2 msee.

DRELATIVE POTASsluM CONCENTRATION

Fig. 4 Spike height iu millivolts c.7. relative potassium concentration as in figure 2. Theaction and resting potentials both st ar t t o decrease rapidly at about twice the normal potm-sium concentration.


7/10

M E M B R A N E P O T E N T I A L S 141tioii of the action potential was observed, although when the potentialbecame less than a millivolt it was d a c u l t to distinguish the all ornone response from the electrotonic potential. On being returned tosea-water, the action potential returned to its normal height evenafter 3 minutes immersion in isosmotic KC1.

Some preliminary observations were made on the effects of changingthe other ions in the circulating fluid. Changing the calcium ion con-centration from zero to 4 times normal produced practically no effecton the resting potential. Removing all ions by circulating isosmoticdextrose increased the potential only slightly ( 3 t o 5 millivolts) higherthan it was raised by removal of potassium alone. Likewise, the heightof the action potential was not appreciably affected by these procedures.Thus in B conducting fiber it was found that the action potential mightvary from over 150millivolts to nearly zero under various conditions ofionic environment, deteriorations, etc., but that the resting potentialwould reniain relatively constant in the neighborhood of 50 millivolts.

DISCUSSIONThe entire process of excision, teasing, and impaling these fibers would

not be expected t o be conducive t o long survival. The various techniqueswere all improved over the previous summer but probably the mostimportant factor was the circulation of fresh sea-water past the axonin the cell. Some of these fibers were probably no more abnormal thanother excised fibers. T h e process of impalement did not measurablychange the height of the action potential as recorded by means of aninactive end before and after impalement. An excised fiber must bein very good condition before it is possible to elicit spontaneousactivity by lowering the calcium ion concentration but this phenomenonwas demonstrated with a few of these impaled fibers. We thereforefeel that the process of impalement produced no drastic change in theaxoii as a whole. The improved condition of the fibers and the in-creased fidelity of the electrical equipment are primarily responsiblefor the larger action potentials observed.

It is rather surprising that the range of the observed resting poten-tials is not greater than 13 millivolts, but on the other hand the con-stancy of the resting potential in each axon during the periods ofobservation suggests that the resting potentials of the axons in situmay have had a similar variation.

The reasons for believing that the potentials recorded here are truemembrane potentials have been discussed (Curtis and Cole, '40). HoM--ever, there is uiidoubtedlp some error in the absolute values of the


8/10

142 H. J. CURTIS AXD Ii. S. C O L Eresting potentials caused by liquid juiictioii potentials. These poten-tials are between (1) he axoplasm and tlie KC1 of the micropipet, and( 2 ) the KC1 of the micropipet and the sea-water communicating withits silver-silver chloride electrode, H of figure 1. The second of thesecan be quite accurately calculated (MacInnes, '38) , and amounts to4.5 millivolts. The first cannot be accurately calculated since the conipo-sition of axoplasm is not known but a n estimate can be made. Usingtlie chemical analysis of axoplasm hy Bear and Schmitt ( ' 39 ) , themobilities of tlie cations and 25% of the anions may be assumed to beknown. The remaining 7570 of the onions are unknown, but let usassume they are monovalent, and have average niobilities such as tomake the known conductivity of axoplasm (Cole and Hodgkin, '39)approximately correct. These values may then be substituted in tlieequation of Henderson ( '07), which gives 6 millivolts. This ra lue isprobably of tlie correct order of magnitude. The total correction is thenthe sum of these two, o r about 10 millivolts. Since these potentials ar eof opposite sign to the resting potential, the true resting potentialsare larger than the measured values by th is amount, and we concludethat the average value of the resti ng potential is about 6 1 millivolts.

These liquid junction potentials will change as the potassiuiii con-centration of the surrounding fluid changes, but assuming tliat tliecomposition of the axoplasm is unchanged, the inaximnni change in thejunction potentials would be 4.5 millivolts. All tliese estimates of tlicjunction potentials are rather uncertain. For this reason it seemsadvisable to repor t only the measured values of potential. The abovecalculations were included only to indicate that the corrections mhiclimust ultimately be made to these nicasurements will probably be i ~ l a -tir ely small.

Thus during tlie passtige of an impulse the membrane potcntiul ismomentarily rerersed in sign, so that the outside may bc as mucli as110 millivolts negative with respect to the inside. This fact throwsdoubt on the simple explanation of the action potential as a passivrdepolarizatioii of the membrane or abolition of the resting potential.IVith the further observatioiis of wide variability in the size of tlieaction potential with little if any change of the resting potential, it isreasonable to suppose that a separate mechanism is responsible forthe production of each. Thus the resting potential may be a n electricalmeasure of the encrm made avaiiable by metabolism and the actionpotential an index of the ability of the membrane to utilize this energyf o r propagation.


9/10

MEMBRANE POTENTIALS 143However, -there may be an explanation of this phenomenon on tlie

basis of a passive depolarization. A membrane inductance has beenobserved, (Cole and Baker, '41) in this fiber of 0.2 henries per cm."and this, in conjunction with the membrane capacity of 1 microfaradper cm.2 (Curtis and Cole, '38) forms a resonant circuit. It has beenpossible t o explain several phenomena of peripheral nerve on the basisof an equivalent membrane circuit involving capacity, resistance, andinductance (Cole, '41). The explanation of the present phenomenon interms of this equivalent circuit i s not available, but it seems possiblethat a coniplete solution of the problem on the basis of the cable equa-tions may yield an adequate explanation.

Thc curve of figure 2 indicates that the potassium concentration dif-ference between axoplasm and sea-water is not a complete explanationof the resting potential, but undoubtedly is an important factor. Athigh potassium concentrations the curve is approximately u straightline. The limiting slope of 50 millivolts for a ten-fold concentrationchange is near the value which it would theoretically have if the mem-brane were permeable to potassium alone, i.e., 58 millivolts. Thcmeasured potential is zero f o r about a twenty-fold increase in potassiumconcentration above normal, or at about the point where the potassiumconcentration is the same inside as outside (Bear and Schmitt, '39).However, the curves flattened out in the neighborhood of the normalconcentration and below, so it appears that some other and perhapsmore important factor is operative. It seems improbable that thisother factor could be some of the other ions normally present, sinceremoval of all ions from the surrounding fluid by the addition of isos-niotic dextrose caused no more change in potential than was causedby the removal of potassium alone. The fact that the size of the actionpotential falls sharply in the concentration range where tlie slope oftlie resting potential curve is changing most rapidly (figs. 2 and 4 )is undoubtedly very important in this connection.

S UM M ARYThe action and resting potentials from the membrane of the squid

giant axon have been measured between one electrode inside and anotherelectrode outside the axon. The resting potential of these axons variedfrom 46 to 59 millivolts and averaged 51 millivolts, but these potentialsare probably about 10 millivolts too low because of liquid junctionpotentials. Action potentials varied froxu 77 to 168 millivolts andaveraged 108 millivolts. Thus during the passage of an impulse theiiiembrane potential is moinentarily reversed in sign so that the out-


10/10

1.44 H. J. CURTIS AND Xi. S. COLEside may be as much as 110 millivolts negative with respect to the inside.The observed membrane inductance may explain this reversal.

The resting potential was measured as a function of the potassiumion concentration of the fluid surrounding the axon. It was found thatin the region of the normal potassium concentration there was a rela-tively small change of potential with concentration, but a t higher con-centrations the potential fell sharply, reached zero at about 20 timesnormal concentration, and was about 15 millivolts negative at 40 times(isosmotic KC1). The spike height was very sensitive to increases inpotassium concentration, and w as reduced to only a few millivolts bya concentration which left the resting potential practically unchanged.

L I T E R A T U R E C I T E DBEM,R. S., AND F. 0. SCHYITT E l ec tro l y te s i n t he axop l a sm s of t he g i an t nerve f ibersCOLE, K. S. 1941 Reet i f i ca t ion and inductance in the squid g i an t axon . J. Gem Pllysiol . .COLE, K. S., A N D R . F. BAXFR1941 Longi tudina l impedance of t h e s q u i d g i a n t ason.COLE,K. S., AND H. J. CURTIR 1938 E l e c t r i c i m p d a n c e of Nitel la during nct ivi tg. J. Gcn.

1039 Elec t r i c impedance of th e Rquid gi an t taxon d urin g act ivi ty. J. Gen. Physiol . ,COLE,K. S., AND A. L.HOWKIN M em brane and pro toplasmic res i s t ance in the squid

g i an t axon . J. Gen. Physiol., vol. 22, p. 671.CURTIE,,H. ., A ND K . 6. COLE T ransve ree i m pedanec of th e squid g i an t axon . J . Geii.Physiol . , vol. 21, p. 757.-~ 940 M em brane ac ti on po t en ti a la f ro m t he squ i d g i an t itxoii. J. Ccll . and Comp.Physiol . , vol. 15, p. 147.HENDERSON,. 1907 Z u r Thermodynamik der F luss igkei t ske t t en . Z. Physik. Chem., vol.

HOWKIN,A . L., A.ND A. F. HUXLEY Act ion poten t i a l s recorded f rom ins ide a nerveM A C I N N E ~ ,. A. 1938 P r i nc i p l e s of elect rochemist ry. Reinhold Pu bl ish ing Co., New York .STEINBACR,. B. Chemica l and concent ra t ion poten t in l s i n the g i an t f ibers of squidnerves. J. Cell . and Comp. Physiol . , vol. 15, p. 373.WEBB,D. ., AND J. Z. YO UN G Elect ro lyte conte nt an d act ion p3teii ti :i l of t h e g i a n tnerve f ibers o f Loligo. J. Physiol. , vol. 98, p. 299.YOUNO,J. Z. 1936 S t r u c t u r e of ne rve fihera and synapses i n som e i nve r t eb ra t e s . ColdS p r i n g H a r b o r S y m p o s i a on Qua nt i t a t ive Bio logy , vc l . 4, p. 1.

1939of t he squid . J. Cell . and Comp. Physiol . , vol. 14, p. 205.vol. 25, p. 29.J. Gen. Physiol. , vol. 24, p. 771.Physiol. , vol. 22, p. 37.

VOI. 22, p. 649. 19391938

59, p. 118.1939

fiber. Natur e , vo l. 144, p . 710.1940

1940

curtis cole 1942 ann p3

Documents