Pharmacological tunes

What did you do after your PhD?

I was working closely with Jack. By about 1958 I had developed a technique for administering very small amounts of drugs near particular cells. That was really a development from some of the work that Katz had been doing in London, but also Bill Nastuck had started it up in America. It's a matter of having a compound in aqueous solution in a glass pipette, of knowing something about its nature and fixing the pH so that the bit you are interested in is either an anion or a cation, controlling it by means of electrical currents to stop it leaking out, and passing it out when you want it.

I was fortunate that double-barrel electrodes (just two tubes fused together) had been developed by Paul Fatt, who was working with Eccles. They were using these electrodes to record from cells – one barrel to record and the other to pass current through. They were pretty crude electrodes but later we persuaded the glassblower to take a round tube and put a partition in it, and then pull it down so we had a theta glass. By the time we were wanting to look at the pharmacology of single cells we were getting greedy, so eventually an English glassblower made for us five-barrel electrodes: a centre one with four around it. By the mid-'60s we were even more greedy, and it was easy to put six around a single one and have sevens. We developed the idea of cementing another electrode on the outside of this so that we could have a single or a double barrel inside a cell and the six or seven outside to play pharmacological tunes on it and get lots of information.

Exploiting acetylcholine

It was a good period. Other people were investigating the mechanism of synaptic excitation and inhibition, and electron microscopy was developing, so that you could appreciate what was going on at synapses with the morphological machinery; but also neurochemistry was developing. Neurochemistry, to me, is two disciplines. One is the design of organic compounds to affect the nervous system. The other is really neuro-biochemistry, understanding what is going on in the living brain. We didn't make major contributions to that aspect but it was being done elsewhere. It was a matter of refining analytical techniques, of being able to look at the concentrations of particular compounds, like some of the amino acids in a certain area, and then to do this after lesions had been made to particular pathways and see if the amount of a particular substance fell.

We were able to exploit acetylcholine because of experiments which Eccles, Fatt and Koketsu – a Japanese who was visiting the department – had done here in the early '50s. There was a little system in the spinal cord that could have involved acetylcholine. The motor neurone sends its messages to muscle in the periphery with acetylcholine, and just before the motor axons leave the spinal cord a little collateral goes back and tickles up some cells. Paul Fatt had the brilliant idea, 'Well, if acetylcholine is released at the far end, it is likely to be released at this little terminal,' so they went looking for these cells. They were able to show that, if they administered drugs intravenously or intra-arterially, the pharmacology of that synapse was similar to the neuromuscular junction. But unfortunately acetylcholine itself, which is the transmitter, didn't get through the blood-brain barrier, so there was a gap.

Rose Eccles – Jack Eccles's daughter – had come back from doing a PhD in Cambridge and had become interested in this, so we looked at that system first. In retrospect we were very lucky, because it was a ten million to one chance that the system was acetylcholine, but we could play pharmacological cadenzas on that cell, developing our technique.

Fitting excitants into the amino acid collection

David, I know you got very interested in amino acid transmitters, and they have been a major part of your work.

The amino acids came into it when Geoff Watkins joined us. We built on biochemical knowledge obtained in the States, by Gene Roberts and his people, of the possibility that gamma-aminobutyric acid might be a transmitter. It was shown very simply to be an important inhibitory transmitter in Crustacea, and in Crustacea strychnine is not a convulsant but picrotoxin is. (That is another drug that was used clinically for convulsive therapy.) With the Roberts story of this GABA perhaps being a transmitter in the crayfish, and with a lot of GABA in the human brain, we had a beginning, because we knew picrotoxin convulsed cats and people.

But we had great problems there, because we could show that GABA had an inhibitory effect on cells when we squirted it out of our electrodes but we could never get picrotoxin to block it. We were well aware of the problem: because it is not ionised, it doesn't come out of the electrodes. That story had to wait to be developed till the early '70s: after Geoff Watkins had left us and Graham Johnson had arrived, we stumbled upon bicuculline as another convulsant, which was a very effective GABA antagonist. But before that, with Geoff Watkins, we had become interested in all of the amino acids we could lay our hands on. I should think I've got the largest collection of useless amino acids in the world, still in my cupboards.

We were able to show that GABA and glycine and related neutral amino acids – where the acidic group was carboxylic or sulphonic or sulphinic – inhibited cells, and that the dicarboxylic acids related to aspartate and glutamate were excitants. And that was something. The excitatory effect was new. We weren't aware of it, but a Japanese called Hayashi had thrown a massive amount of glutamic acid into the cerebral cortex of dogs some years before and they convulsed. We could have used tap water or potassium chloride or anything, but it was interesting that this was a compound which was known to be in brain and excited cells. We didn't know about this for some time, but we went on looking and collecting amino acids. They were fascinating but we just didn't think they were very important – we were a bit naive, a bit stupid.

I'd thought that cholinergic and adrenergic transmitters were the be-all and end-all.

The work with acetylcholine showed us that that was wrong, in the spinal cord anyway: only this Renshaw cell synapse was cholinergic – using acetylcholine – whereas acetylcholine didn't affect anything else. A major turning point was in '65-'66, when an American group reported that the amount of glycine in the spinal cord fell remarkably if they destroyed a lot of the inhibitory interneurones. This immediately linked glycine up as a possible inhibitory transmitter. They also showed that glycine affected neurones the same way as the inhibitory transmitter. We rapidly confirmed that, but in addition we had strychnine up our sleeve, knowing it should be doing something in that system. It very clearly blocked the effect of glycine.

Using antagonists to investigate transmitters

That put us back into business on gamma-aminobutyric acid, because that led, in another five or six years, to bicuculline and a lot of alkaloids that were not glycine antagonists but were antagonists of GABA. This was useful for our chemical colleagues, because we were able to analyse the glycine receptors and GABA receptors not only by playing tunes on them with a number of glycine and GABA analogues but also through the effects of strychnine-like and bicuculline-like compounds. Some very fascinating chemistry came out of that.

Penicillin, for example, is a GABA antagonist, and it was known a long time ago that in patients where penicillin had been used for, say, a cerebral abscess and had leaked out into the cerebrospinal fluid, this was a convulsant. Fortunately, penicillin doesn't get through the blood-brain barrier, so it doesn't normally have this action.

Focus questions

• What is the difference between an excitatory neurotransmitter and an inhibitory neurotransmitter?

• Why do you think Curtis used the term 'playing pharmacological tunes' to describe the way he investigated neuronal responses?