Effects of heart failure on brain-type Na+ channels in rabbit ventricular myocytes: Reply

We greatly appreciate the comments of Petitprez and Abriel on our article ‘Effects of heart failure on brain-type Na + channels in rabbit ventricular myocytes’ in a recent edition of Europace. 1 In that article, we presented recordings of quasi-macroscopic Na + currents from rabbit ventricular myocytes in the cell-attached mode using an electrode with a relatively large tip opening to measure from a macropatch on the surface membrane. We reasoned that, due to the differences in localization of the various Na + channel isoforms, 2 a membrane patch from the middle portion of the cell would contain TTX-sensitive brain-type isoforms, whereas a patch from the intercalated disc region would contain the primary cardiac, TTX-resistant isoform, Na V 1.5. This hypothesis seems substantiated by our TTX experiments. Quasi-macroscopic Na + currents from the intercalated disc region were insensitive to 50 nM TTX, whereas quasi-macroscopic Na + currents from middle portions of the cell were virtually absent in the presence 50 nM of TTX. Subsequently, we used quasi-macroscopic currents recordings from middle portions of the cell to characterize the effects of heart failure on brain-type Na + currents. Neither density nor gating properties of brain-type Na + current was altered in response to pressure and volume overload-induced heart failure in rabbit. Petitprez and Abriel address an important and intriguing issue. On the basis of a cylindrical cell geometry used in a computer model of a guinea pig ventricular myocyte, 3 with a length of 100 µm and a diameter of 22 µm, they calculate that the lateral membrane surface area is ∼9 times larger than the intercalated disc surface area. In our study, 1 we found an 8.5 times larger quasi-macroscopic currents in patches from the intercalated disc region, which, as set out by Petitprez and Abriel, would imply that cardiac and brain-type whole-cell Na + currents have similar amplitude. We fully agree with Petitprez and Abriel that this amount of calculated brain-type Na + current contrasts with findings in ventricular myocytes of mouse, dog, and rat, where brain-type Na + currents account for <20% of the whole-cell Na + current [see Verkerk et al. , 1 and primary references cited therein]. Since we used ventricular myocytes of rabbit, species differences may account for the contrasting findings. We have performed pilot experiments in rabbit ventricular myocytes in which we have tested the effects of 100 nM TTX on whole-cell Na + current, using experimental conditions as we described previously. 4 We found that 100 nM TTX never blocked more than 15% of whole-cell Na + current, thereby excluding species differences as a potential explanation. The simplified myocyte morphology as used by Petitprez and Abriel in their calculation, i.e. a cylinder with two intercalated discs (top and bottom of the cylinder) and a length-to-width ratio <5, may also contribute to the discrepancy. In reality, ventricular myocytes are not simple cylinders 5 : intercalated discs can be found also along the lateral sides, 5 the length-to-width ratio is 7–8, 6 and the number of intercalated discs is ∼10 [see Jongsma and Wilders 7 and primary references cited therein]. Although an increase in length-to-width ratio will increase the calculated ratio of brain-type and cardiac Na + current, the higher number of intercalated discs may result in a lower ratio. Unfortunately, despite an extensive literature search, we were unable to find quantitative data on the ratio between intercalated disc surface area and remaining surface area within one species and cell type, which makes a proper calculation impossible. Finally, the discrepancy may also be related to the inevitable underestimation of the amount of cardiac Na + current using macropatch recordings. First, it is easy to imagine that macropatches from the intercalated disc region may not contain intercalated disc surface only, but will be a mixture of ‘real’ intercalated disc (with cardiac Na + channels) and ‘normal’ cell surface (without cardiac Na + channels). Second, we have performed our measurements using close-to-physiological conditions and the activated quasi-macroscopic inward Na + currents will depolarize the myocytes. Although the depolarization with the relatively small brain-type Na + currents is limited, 1 the larger intercalated disc Na + currents may depolarize the cell substantially. Consequently, the driving force for Na + will be reduced in the intercalated disc macropatch recordings, resulting in an underestimation of the intercalated disc Na + current amplitude. A third possible explanation for the underestimation of the cardiac Na + current may come from the total noise in our macropatches. When 50 nM TTX was present, quasi-macroscopic currents as well as single channel activity were virtually absent in macropatches from the middle of the cell. However, we cannot exclude the presence of remaining Na + channels completely. Single Na + channel measurements, especially at physiological temperature, require optimal signal-to-noise ratio, 8 which was definitely not the case in our experiments. Thus, we may have overlooked residual cardiac Na + channels in macropatches from the middle of the cell. This hypothesis seems supported by the presence of cardiac Na + channels along the lateral membrane as observed in some immunohistochemistry studies [see Verkerk et al.1 and Abriel, 9 and primary references cited therein]. In conclusion, the careful comments by Petitprez and Abriel clearly demonstrate that macropatch recordings are not suitable to give an accurate determination of the ratio of ‘cardiac’ and ‘brain-type’ Na + currents in ventricular myocytes. However, these concerns do not affect the main conclusion of our study, i.e. density and gating properties of brain-type Na + current are not altered in response to pressure and volume overload-induced heart failure in rabbit.