CELLULAR ELECTROPHVSIOLOGY OF THE HEART 



271 



threshold regenerative process at normal [Ca++]o 

 makes it difficult to understand how decrementless 

 propagation could occur. An explanation of the 

 apparent propagation of early repolarization is 

 suggested by figure 2 1 (cf. 28, figs. 3 and 4). Following 

 the termination of the hyperpolarizing pulses the S 

 at which 8 goes through zero must be G,,,, and so small 

 hyperpolarizing pulses can be used to map out the 

 time course of Seq to the extent that the Seq at any 

 time is unaffected by 8. In figure 21.-I it can be seen 

 that 8,.q is near 8 in early (top) and late (bottom) 

 repolarization but Seq is consideraljly below S in 

 mid-repolarization (middle record fig. 21.-!, top 

 record 21 5). This time course for 8,.q is similar to the 

 one shown in figure 22. Note that in mid-repolariza- 

 tion hyperpolarization speeds the remainder of the 

 repolarization process. Cranefield & Hoffman (28, 

 fig. 8) obtained records of early repolarization at a 

 distance of 5 mm from the polarizing electrode in 

 which nearly all signs of electrotonic current spread 

 had disappeared. The current flowed during early 

 and middle repolarization. There is no sign of electro- 

 tonic spread at 5 mm during the early repolariza- 

 tion where 8pq is near 8 but there is some remnant in 

 mid-repolarization where 8pq has fallen considerably 

 below 8. Thus it seems conceivable that a large 

 hyperpolarizing current terminating in mid-repolari- 

 zation could shorten action potentials for a con- 

 siderable distance Ijecause the 5,,,^ is spontaneously 

 moving in the same direction as the current-induced 

 changes in S and because electrotonic spread is 

 greater during this period due to the reduced slope 

 conductance. As pointed out by Weidmann (125) 

 and Cranefield & Hoffman (28), a 5 mm conduction 

 distance is not great enough to permit a decision 

 between a decrementing and nondecrementing early 

 repolarization. No matter whether early repolari- 

 zation is propagated or not, the conduction speed of 

 normal repolarization is faster than early repolariza- 

 tion and eventually the latter will disappear. 



HYPOTHESES OF REPOLARIZATION. Although Weid- 

 mann's (125) discussion of the ionic mechanism of 

 repolarization in cardiac tissue appeared while the 

 Na+-K+ theory for squid axons was still being for- 

 mulated, he considered the possible modifications of 

 the theory which might explain repolarization in 

 heart. He suggested that either a long delay in the 

 inactivation of gNa or a retardation of the rise in g^ 

 following depolarization might be responsible for the 

 long delay in repolarization. He further ventured 

 (125, p. 234) "that one or other of these processes 

 may be entirely absent, or that one or other may 



occur in two stages." He then concluded that his 

 impedance (slope conductance) data were compatible 

 with a rapid but incomplete inactivation of gNa and a 

 slow rise in gK, but not with a slow inactivation and a 

 rapid rise in gs. Despite the difficulties in interpreting 

 slope conductance measurements, this conclusion 

 seems generally valid provided the implied con- 

 ductance changes are not so voltage-dependent that 

 they give rise to negative G values. 



As mentioned above, Weidmann (130) induced 

 early repolarization of turtle ventricle by increasing 

 [K+]o during the plateau (fig. 19). On the basis of 

 this finding, he suggested that normal repolarization 

 might result from the accumulation of K+ outside 

 the fiber during the prolonged depolarization. From 

 rough calculations, he concluded that existence of 

 this mechanism was not likely unless there is a barrier 

 to ionic diffusion lying a short distance outside the 

 excitable membrane. He calculated that [K+]o 

 would increase about 0.6 niM during the action 

 potential, but he found that a [K+]o of 10 to 30 

 times normal was required to produce appreciable 

 shortening of the duration. A rise this large during 

 activity would require an interstitial space about 20 

 to 60 times smaller than the 28 per cent value used 

 by Weidmann. Recent improvements in electron- 

 micrographic histological techniques have shown 

 that cardiac cells are closely packed, the spacing 

 being only a few hundred Angstroms (6, 91, 93, 94, 

 107, 108). Individual cells are formed into long, 

 thin strands about six cells in diameter. Most of the 

 interstitial space lies between these strands and is 

 occupied by capillaries. Using Weidmann's figures, 

 calculation shows that the interstitial space within 

 the strands is small enough to cause a rise of about 10 

 mvi in [K+]o during the action if all K""" leaving the 

 cells stays in this space. Thus, this mechanism cannot 

 be definitely excluded on histological grounds. 



If an increased [K+]o is the means of repolarization 

 during a normal action potential, then the sequence 

 of events would be somewhat as follows. Depolariza- 

 tion would lead to an increased efflux of K+ owing 

 to the increased driving force, provided that gg 

 either did not decrease too much, remained constant, 

 or increased. This net efflux of K+ would cause a 

 more or less gradual rise in [K+]o. In order to induce 

 early repolarization, this increased interstitial [K+] 

 must increase gK or gci or decrease g^^ or any com- 

 bination of these. An increase in gci is an unlikely 

 explanation since repolarization occurs in Cl~-free 

 media (73). Since an increase in [K+]o increa.ses gK 

 in nerve (65) skeletal muscle (47, 55) and cardiac 



