272 



HANDBOOK OF PHYSIOLOGY 



CIRCULATION I 



muscle (9, 75), an increased gK is the most likely cause 

 ol the early repolarization. If it is assumed that Pk 

 is proportional to [K+]„ and that 8 is given by the 

 Goldman equation (equation 6), then the S,[K+]o 

 curve has a minimum (most negative) \'alue at an 

 intermediate [K+]o. Either an increase or decrease of 

 [K+]o from this value causes a depolarization. The 

 higher Pn:, is, the higher the minimum value of S. 

 Thus, during the early plateau, the increasing [K+]o 

 could cause repolarization up to the value of [K+]o 

 at which S is minimum. Any further accumulation of 

 K+ would depolarize. The maximum repolarization 

 would be considerably less negatix'e than fir and the 

 remainder of the repolarization process would require 

 the reduction of [K+]„ to normal. This could be 

 accomplished by enhanced Na+-K+ pumping and by 

 diffusion of K+ into the larger interstitial spaces. 



This hypothesis explains the following findings 

 concerning the nature of repolarization, a) The 

 conductance changes of repolarization are non- 

 regenerative, b) The third phases of action potentials 

 of different lengths are superimposable. c) The 

 potential changes produced by large depolarizing 

 currents applied to Purkinje fil)ers bathed in Na+-free 

 media tend to fall off after about 100 msec (75). 



On the other hand, the following objections can 

 be raised, a) At first thought, the ts,tAP relationship 

 could be explained by assuming that [K+]o is slightly 

 above normal immediately following repolarization 

 so that less time would be required for [K+]„ to reach 

 the same value during an immediately succeeding 

 action potential. However, if this were the case, the 

 conductance immediately following repolarization 

 would be elevated and this does not accord with 

 experiment (iii, 125). /)) Hutter & Noble (75) 

 found that depolarization decreases membrane 

 conductance of Purkinje fibers in Na+-free media. 

 They also found that increasing [K+]o increased 

 conductance and so it appears that depolarization 

 causes at most a small increase in interstitial [K+]. 

 c) The time course of S,.,, (fig. 22) during the late third 

 phase indicates that Sk is not greatly different from 

 normal. However, the calculation of 8,,, is not accurate 

 enough to give this argument much weight, d) Since 

 this repolarization hypothesis requires increased 

 turnover of K+ and enhanced active K+ transport, 

 the teleological argument for economical operation 

 makes the hypothesis less likely, e) Brady (personal 

 communication, i960) has recently repeated VVeid- 

 mann's (130) experiments. He found that [K+]o 

 must be raised at least 20 times during the plateau to 

 produce early repolarization in turtle ventricle. The 



early repolarizing action has a latency of 3 to 4 sec 

 as compared with a latency of about o. 1 5 sec for a 

 depolarizing action when the high [K+]o solution 

 was perfu.sed through the coronaries during diastole. 

 The extremely long latency for induction of early 

 repolarization implies an even more indirect mecha- 

 nism than an induced change in gK. 



Hoffman & Cranefield (68) have advocated the 

 hypothesis that g^ is inversely proportional to the 

 driving force on K+ as an explanation of repolariza- 

 tion — i.e., depolarization decreases gK- This, however, 

 is a regenerative system and hence must be discounted 

 on the basis of present evidence. Shanes (106) suggests 

 that the inward mo\'ements of Na+ and the outward 

 movements of K+ interact when the two fluxes are of 

 the same order of magnitude. For example, such 

 interaction could occur if Na"*" and K+ moved 

 through the membrane in file in the same pores (65). 

 This ingenious mechanism is a concrete model 

 whereby an increase in [K+Jo could lead to a decrease 

 in gN.T. Such a decrease in g^^ could account for the 

 early repolarization induced by an increase of [K+]o 

 during the plateau. Shanes also suggests that gK 

 decreases continuously throughout the slow diastolic 

 depolarization, the succeeding rapid upstroke of the 

 action potential and the plateau phase, and then 

 increases rapidly at the time of rapid repolarization. 

 He does not specify if these changes in gK are de- 

 pendent on time, voltage, or both. If they are voltage- 

 dependent, the third phase is regenerative and thus 

 suspect. If they are time-dependent, the factors that 

 make gK oscillate in time in this manner are not clear. 



Coraboeuf et al. (24) have measured G during the 

 action potential of the guinea pig \entricle. Their 

 results are closely similar to Weidmann's observations 

 (125) on Purkinje fibers except that G > Gr during 

 the action potential. Coraboeuf and his co-workers 

 suggested three possible mechanisms of repolarization. 

 The most intriguing one, attributed to Hodgkin, is 

 that K+ can use the Na+ channels to a limited extent 

 and so gNi, and gK, initially high, fall simultaneously, 

 the \oltage staying constant as long as gK is much 

 larger than its resting value. As inactivation pro- 

 ceeds, gNa and gK would fall in parallel until gK 

 approaches its resting level. Thereafter gK gNa 

 would rise rapidly and cause a rapid repolarization 

 which would be aided by a voltage-dependent 

 regenerative decrease in the remaining gNa and 

 slowed by a similar decrease in gK- It appears that 

 this ingenious scheme suffers from the usual defect 

 that G goes negative during fast repolarization. 



Brady & Woodbury (4) have recently proposed a 



