Gold: I have a question about the K+ currents. In your paper describing modulation of TTX-R Na+ currents in DRG neurons, you also described the inhibition of K+ currents (England et al 1996). Grant Nicol and Michael Vasko reported a similar in£ammatory mediator-induced inhibition of K+ currents, but the time course they reported was relatively slow (Nicol et al 1997). Furthermore, the inhibition of K+ currents that you described was reversed following wash of PGE2. Because PGE2-induced changes in excitability occur relatively quickly and are not very reversible following wash of PGE2, how do these observations fit with inhibition of K+ current as an underlying mechanism of sensitization?
Bevan: I can speculate. As far as the underlying mechanism for a wash, they weren't very rapid reversible effects. If phosphorylation of the K+ channels is involved in this, it will obviously depend on the dynamics of the dephos-phorylation, and the balance between the phosphorylated and dephosphorylated states.
Catterall: It is striking how different and opposite the effects of PKA and PKC phosphorylation are on these Na+ channels from DRG neurons, compared with the type 2A channels that we have studied that are expressed in the brain. In those we have studied, the same signalling pathways (the PKA and PKC pathways) reduce channel activity, but the PKA and PKC act synergistically to do so, just as they do in your experiments to increase channel activity with the SNS channels. It is as if the sign of the regulation is different, but the underlying mechanisms are the same. In the type 2 channel, it is the phosphorylation sites in the loop between domains 1 and 2 that are important, as both Al Goldin's lab and ours have shown. It is as if these two channels are set up to respond to the signalling pathways differently, but may do so by a common mechanism.
Bevan: It will be fascinating to do some of the experiments that you have done involving site-directed mutagenesis to pin down the residues involved in the SNS channel.
Catterall: That story has become even more complicated in more recent experiments. We have found that both membrane potential and PKC influence the functional effects caused by phosphorylating individual PKA sites.
Spruston: Have you compared whether the effect is the same or different in the TTX-S Na+ channels with these cells? It could also be a cell-specific effect as opposed to a channel-specific effect.
Bevan: We studied neonatal rat DRG neurons, and most of the cells we studied seemed to respond to prostanoids. This is not the case in the adult cells, where about 50% of them don't respond. If you take cells that only have TTX-S currents, then there doesn't seem to be much modulation at all by prostanoids. We didn't look at cAMP analogues in those experiments. During the separation of TTX-R and TTX-S, it looks as though there may be an effect by prostanoids on the TTX-S currents in those cells that showed a TTX-R modulation. We didn't study this in detail; it seemed to be much more subtle than the effect on the TTX-R current.
Gold: That is one of the nice things about the DRG neurons: you can study both currents and use one as a control for the other. It wasn't consistent that we would see an inhibition of the TTX-S current, but we did see an inhibition as a general rule. This was consistent with the brain-type channels. While PGE2 would increase TTX-R it would decrease the TTX-S. It didn't seem to be cell specific.
Bevan: It is a rather nice system where there are opposite effects in the same cell by activating the same pathways.
Goldin: Are the effects reversible, and with a similar time course?
Bevan: In our experiments we didn't study the cells for a long time. In some of them we saw the currents begin to decrease. We were always worried about long-term rundown in the currents. They are reasonably robust but after an hour I wouldn't be confident we weren't getting effects completely unrelated to previous drug treatment.
Bean: How do you choose the cells in doing these experiments? You said these were from neonates.
Bevan: Ours were from neonates and Michael Gold's were from adults. For ours, we didn't use any special criteria to select the cells.
Bean: One hears so much about the heterogeneity of sensory neurons. Is this an issue with the adult cells?
Gold: They are heterogeneous. Early on we had broken them into categories on the basis of cell size, responsiveness to capsaicin and IB4 binding. In general it doesn't seem to fall out in terms of which population will be modulated and which won't by any of those categories. Our interpretation of that is that it had to do more with how the receptor for the prostaglandin is distributed. The receptor didn't seem to be distributed with any a priori category we used. With respect to cell size, my feeling is that it is much less predictive than some people have been proposing. There is a lot of weight put on the functional significance of cell size. There is a loose correlation that smaller cell bodies give rise to more slowly conducting axons. I think this is reflective of target innervation. This correlation holds up fairly well with cutaneous afferents, but in the visceral afferents that we have recently been studying, the correlation with cell body size seems to fall down. It should be noted that Reese Scroggs has implemented a system for categorizing DRG neurons and he has reported that specific groups of cells respond to particular inflammatory mediators (Cardenas et al 1997).
Waxman: Bruce Bean is raising a very important point. Cutaneous afferents express large TTX-R currents, whereas muscle afferents show lower levels. Then we have the TTX-S fast current which is probably the composite of the products of several different TTX-S channels. The reflex solution is to put them into heterologous expression systems and Lori Isom addressed very clearly the problems we run into there. We need to triangulate between studying native systems and the various expression systems.
Catterall: However, the PKC effect is exactly the same in the transfected system as it is in hippocampal neurons.
Goldin: You said you saw variability in the decrease of the TTX-S currents. This would make sense, because 1.1 is decreased by PKA, whereas 1.6 does not have the critical second PKA site, and as predicted from this absence it does not show a decrease with PKA. If there are variable proportions of 1 and 6, you would see variable amounts of decreased current.
Strichartz: I have some concern about the steady state levels of phosphorylation being very dependent on the activity of phosphatases. I wonder whether these experiments should be repeated with perforated patch or cell-attached patches, because of the possibility of a much larger or smaller effect as a result of the inevitable perfusion that occurs. Does anyone know anything about the effects of the intracellular media that we choose? They are pretty much the same.
Gold: I think that is a good point. Using a cell-attached patch we have seen larger modulation. The other issue that also adds to this is the complication to the story with PKA and PKC when you move back to in vivo models. For instance, people who have been trying to inhibit PGE2-induced hyperalgaesia with a PKC antagonist have been unable to do so. This is associated with what was shown in terms of the in vitro data, although there are inflammatory mediators that do act through a PKC pathway. Our subsequent data would suggest that this reflects the level of resting Ca2+ in the cells, which changes depending on how you prepare your cells. If we do Ca2+ measurements on cells prepared with one sort of enzyme versus another and look at neurons in a window of 4-6 h versus 18—20 h, the resting Ca2+ levels change. Consequently, my guess is that resting PKC activity and phosphorylation is going to change. This will impact what is seen in terms of subsequent modulation of Na+ currents.
Strichartz: What is the highest resting Ca2+ level that you see?
Gold: If I recall correctly, about 200 nM.
Bean: It is extremely difficult to do Na+ channel voltage clamping with perforated patch recordings. It is hard to get the series resistance down low enough to get a good clamp.
Raman: In whole cell recordings from Purkinje cells over time, the inactivation curve tends to shift to more negative voltages.
Strichartz: That is a universal phenomenon, by the way.
Horn: Not so much in two microelectrodes in oocytes.
Strichartz: It seems to occur when you start mucking up the cytoplasm.
Horn: In perforated patches it tends to be relatively stable.
Strichartz: Interestingly enough, in old experiments that Jim Fox did in node of Ranvier, if he set the holding potential very negative (to about—110 mV) he found that the mid-point for inactivation shifted much more slowly over the same period than if the membrane were held at—80 mV (Fox 1976). So it seems that this 'drift' does depend on the state of the Na+ channel. If you can remove slow inactivation, you don't see this drift in the fast inactivation parameter.
Horn: In our experiments we usually use holding potentials of —140 or —150 mV, and we still see those shifts. There isn't much slow inactivation at those holding potentials.
Wood: I was very interested in the HSV data. I was rather surprised at these dramatic effects, because one normally associates HSV with a latent infection, with very low level viral protein synthesis. Does this in vitro model reflect that form of latency, or is it a high multiplicity of infection model?
Bevan: It is a high multiplicity of infection. We have 5 pfu per DRG.
Wood: Did the cells survive long-term?
Bevan: No. We were interested in trying to induce latency and then looking at recovery from latency, but in vitro there are no reliable methods of doing that.
Noebels: Are these replication-defective HSVs being used as vectors for gene transfer?
Bevan: Nina Storey studied a large number of vectors that are potentially going to be used therapeutically, and none of them have effects on any of the currents.
Cummins: We tried the amplicon system and it seemed to have major effects on the Na+ currents in both DRG and cortical neurons
Bevan: It depends on which vectors you are using. The ones we looked at were those developed by David Latchman and intended ultimately for therapeutic use.
Strichartz: We were doing some experiments on neuroblastoma cells to try to understand local anaesthetic-induced neurotoxity, which was a clinical problem a while ago. Anaesthesiologists use obscenely high concentrations of local anaesthetics because the efficiency of transport into the peripheral nerves is so low; 200 mM is not unusual. We applied 200 mM lidocaine and saw that although some of the cells died, many didn't. When lidocaine was removed after an hour, all the cells that didn't have crenulated membranes and that we could get a whole seal on, had almost no current in them. However, over the next 10 h the current grew back to normal (of course we were not clamping any one cell for 10 h). We thought that this might be due to the independently observed increase in intracellular Ca2+ that is produced by local anaesthetics, and which I think you have also seen in DRG. I wondered whether there might not be Ca2+ going through your Na+ channels and producing those changes, and it is not Na+ per se, but an elevation of intracellular Ca2+.
Bevan: We can't rule that out, because we haven't checked the Ca2+ levels. One can get a similar down-regulation of the Na+ current in these cells by short-term treatment with veratradine.
Meisler: I wanted to ask about the brefeldin experiment you used as a control. Does that prevent turnover or delivery of the endogenous proteins as well? If so, does that say anything about the turnover of channels?
Bevan: It will block all proteins being exported.
Meisler: It had no effect on channel activities in 24 h.
Bevan: That would fit in with cycloheximide experiments, where over that period we see no appreciable loss of Na+ currents. Whether preexisting channels can be cycling in and out is another question.
Catterall: Francois Couraud's lab in Marseilles has done similar experiments on cultured brain neurons from early embryos, and finds a similar Na+ dependent internalization and down-regulation of Na+ channels. However, this is not observed in more mature cultures of neurons. Do you see the same correlation between the age of the neurons and the ability to down-regulate Na+ channels?
Bevan: All the HSV experiments were done on adult neurons. On the basis of their results we were expecting not to see an effect of veratradine in the adult animals, but in fact we clearly do see this. It occurs within about 5 h, which is very similar to the time course they see for their cortical neurons from young rats. At one stage they speculated that the stability changed: as the animals aged, the veratradine effect was lost. They speculated this was due to an increase in the synthesis of b subunits. I don't know whether there is any further evidence for or against that hypothesis.
Catterall: There's a temporal correlation between expression of b subunits and loss of down-regulation. Over the same time frame the Na+ channels also change from being mostly type 3 to mostly type 1 or 2. Couraud and colleagues found that it is Na+ and not Ca2+ that is the mediator. Other than this, I don't think the mechanism is further developed.
Was this article helpful?