In article <5vpg45$100i at f1n1.spenet.wfu.edu> John Mihic, jmihic at bgsm.edu
>Matt, now look what you did. Here I was contentedly wasting my time preparing
>a lecture and writing grant proposals, when you have to come along and distract
>me with your analysis of the Eghbali et al. paper.
Oops. Sorry, John ;-)
>They state that they controlled for injection technique by injecting GABA with
>DMSO (vehicle to dissolve diazepam, for the uninitiated). And in Fig. 2, they
>show a gradual increase in conductance by diazepam using the inside-out patch
>technique and reversal of the diazepam effect by flumazenil.
You're right. They did say that. But they didn't show it or give any
stats. Publishing in Nature limits the available space, fair enough. But
that makes it harder, rather than easier, to evaluate the data. A lot of
other people have studied benzo effects without seeing any conductance
increase whatsoever, as you pointed out before. I frankly didn't find
their discussion of that particular matter very compelling.
>I think the flumazenil reversal (Fig. 2B) is more convincing than anything they
>could have done at the GABA-A site.
Hmm. I guess what I was getting at is that if you can show that whatever
you are looking at is affected by doing something at the GABA-A site,
then by definition, it involves GABA-A receptors. If you *can't* show
that, it amounts to proving that it's *not* a GABA-A receptor. If you
*can* but you *don't*, then how should your data be interpreted? You
probably know more about this than I, but aren't there non-GABA-A
receptor benzo binding sites (especially at micromolar conc)? Like, isn't
there some well known mitochondrial effect of benzos (not that that would
have any direct bearing on this topic)? So the flumazenil expt shows that
they're dealing with a benzo site, but that's not exaclty the same as
proving that they're dealing with GABA-A receptors. Or is it?
>>The authors state they used low concentrations of GABA in these studies.
>Benzodiazepine potentiation is not observed when high concentrations of GABA
>are used. However, if conductance is changed by benzos, wouldn't you expect to
>still see enhancement at saturating GABA concentrations?
That appears to be the concensus, that you don't see peak potentiation of
macroscopic currents under saturating GABA concentrations. This has led a
number of people to conclude that the kinetic effect of the benzos is to
increase just the microscopic affinity (specifically the forward binding
rate) without changing the gating. This probably isn't strictly true
because there is recent evidence (from Andy Randall and others) that the
intrinsic gating (e.g., macroscopic desensitization during long
saturating GABA pulses) is also affected. But to answer your question,
yes, I would expect to see enhancement at saturating concentrations if
there were a 2-10 fold conductance change.
*However*, the authors had their own explanation for this discrepancy,
which was to suggest that the conductance change itself only happens at
low concentrations. They further suggest that part of the prolongation of
the IPSC may be due high conductance channels emerging during the decay,
as receptors slide down into lower occupancy (i.e. when the GABA
concentration in the cleft falls and they unbind GABA molecules).
Overall, this picture is quite consistent with their multiple-occupancy
multiple-pore scenario, which is good. Still, it needs more direct
evidence I think.
I guess I've started to sound like I'm harping on the problems of this
paper. If so, I apologize. Like I said, we should try to be open minded.
But this result is so jarringly unfamiliar that I can't help but look for
alternative explanations (is that so WRONG?). How about you, John? Got
any benzo/conductance expts running these days? Not me, baby.
>Ahh, the sweet smell of scepticism. Neil indeed trained you well.
Yeah, he's a pharmacologist *and* he's British: scepticism is second
nature. He actually teaches a graduate course in scepticism at U. of C.