Dear Dr.***:

I would like to resubmit a revised version of M8-4755, "Tiitle" by Author et al. The new versron incorporates new experiments that we believe answer the concerns of the reviewers. Their major criticism was directed towards the viability of the hepatocytes and the magnitude of their membrane potential. In the original paper, some experiments with viabilities as low as 85% had been included in the analysis. These poorer data, which we believe were due to a widespread problem with Boehringer Mannheim collagenase during early 1988, have now been removed from the analysis, which now includes only experiments with viabilities of >90% as measured by trypan blue exclusion. All the statistical data have therefore been recalculated. Also, as now pointed out on p.5, single-cell imaging of [Ca2.]i Provides additional protection against contamination of results by unhealthy cells, since fura-2 is itself a vital stain that rapidly leaks out once cells are damaged, just as trypan blue leaks in, so that leaky cells are too dim to measure at video camera gains suitable for loaded cells. Also, the field of view was constantly monitored with infrared transmitt~d illumination, and cells that looked sick were also excluded from analysis. For these reasons, single cell analysis is much less affected by the inevitable small percentage of unhealthy cells than experiments on bulk populations are. We have also redone the electrophysiological measurements of resting potential. The values mentioned in the original manuscript, -30 to -40 mV, were obtained with KCl-filled electrodes with 10 - 50 megohm resistances, and are typical of results in the literature (-10 mV to -40 mV) for such methodology. Petzinger and Bigalke (1986) achieved the much more negative membrane potentials demanded by the referees not by improving the hepatocyte isolation procedure but by using much finer-tipped electrodes of 80 - 120 megohm resistances By the same method we have likewise improved our measurements and obtained a mean resting potential of -69.2 mV in 6 cells, with an s.d. of 9 mV. Sample chart recordings are enclosed showing a typical experiment reaching -68 mV as well as the largest potential obtained, -80 to -89 mV with some uncertainty due to baseline drift. By comparison, Fig. 3A of Petzinger & Bigalke (1986) shows a potential of -73 ＋ 8 mV at the external K+ concentration we used, 5.8 mM. We believe this data should satisfy the demand that ,' . Experiments must be performed with hepatocytes with membrane potentials in the -70 mV range", and prove that alternative l) of reviewer I is correct, i.e. that large membrane potentials are obtainable simply with finer microelectrodes without requiring better isolation techniques or substrates for attachment. The new electrophysiological measurements are now incorporated on pp. 6, 9, and 13.

In reply to the concern of reviewer # I about autofluorescence, we now point out more explicitly on p. 6 that autofluorescence was a very small percentage of the total signal from these rather heavily-10aded cells. Moreover, in control experiments where the camera sensitivity was boosted in order to examine the autofluorescence of unloaded cells, the 350:385 nm ratio of the autofiuorescence showed no change (and certainly no oscillation) upon addition of phenylephrine, so that autofluorescence from NAD(P)H cannot be the source of the oscillations observed by fura-2 ratioing.

Finally, we think the concem of reviewer # 2 that the UV excitation may be destroying dihydropyridine drugs is answered by the fact that these drugs at the same or lower concentrations do work as expected in our experimental setup on other tissues known to have L-type voltage-operated Ca2. channels, for example frog sympathetic neurons (Lipscombe et al (1988), Neuron 1, 355-365) or REF-52 fibroblasts (Harootunian et al, ref. 10). In other words, bioassay with susceptible cell types shows that any UV photodestruction is not too severe, perhaps because diffusion from the nonilluminated remainder of the chamber compensates. Also, we did not claim that the drugs were ineffective on hepatocytes, but rather that their effects were inconsistent with simple specificity for closing or opening voltage-gated Ca2. channels. Meanwhile, other colleagues have convinced us that the section on these drugs is the weakest part of the paper, so it has been greatly shortened, moved to the end of the section on effects of depolarization, and all three panels of Fig. 15 have been deleted.

A number of more minor changes have been made, many in response to colleagues' criticisms. Statistical spreads are now stated as standard deviations rather than standard errors of the mean, since even though the latter looks more precise, the former is more appropriate for measurements of the heterogeneity of biological populations that do not necessarily follow a Gaussian distribution. The discussion of the possible reasons for the difference between our results and those of Monck et al (ref. 12) has been reworded at Prof. Williamson's request. Some new references have been added (19, 33, 47), including an abstract (33) that appeared after submission of our original manuscript, agreeing with us that [Ca2.]i oscillations can be observed and analyzed in fura-2-loaded hepatocytes.

We hope that with these revisions and new experiments, the paper is now acceptable for publication in the Journal of ***. I enclose a copy of the original manuscript with changes marked in red, together with two copies of the revised manuscript.

Yours sincerely,