Friday, June 19, 2009

Drugs, Receptors, and the Brain

My last couple of posts were pretty heavy, certainly for me, so I'm going to drop back to a slightly lighter level, and comment on something I've noticed, both in the literature and in blogs, about neurotransmitters and their pharmacology. Let me start with a few posts from Scicurious at Neurotopia, whose posts on this subject I try to always catch: All of these posts are discussing the action of neurotransmitters such as dopamine and serotonin, which tend to modify the action of other neurotransmitters rather than directly cause voltage changes in the post-synaptic cell.

A tendency I notice is to assume that a neurotransmitter such as dopamine has a particular "meaning", such as that more dopamine in the brain means you're experiencing something you like, or that dopamine is part of the "reward" system in the brain. We can see this in the last link, which discusses "Two types of dopamine neuron distinctly convey positive and negative motivational signals" by Masayuki Matsumoto and Okihide Hikosaka. Here, it's shown that only a subset of neurons that deliver dopamine to their synapses are entirely reward-motivated, while others seem to respond more to unexpected stimuli, whether desirable or otherwise.

IMO this is a somewhat simplistic view. The first thing we have to remember is that there are many receptors for each of these neurotransmitters, with different actions. The second is that, as Matsumoto and Hikosaka have demonstrated, different populations of neurons that deliver the same transmitter "mean" different things by their message. Let's start with that. ...

Suppose you have a population of cells that deliver dopamine when some significant class of event has occurred. They deliver their message to each region of the brain that needs this message, and the cells in each region modify their behavior in response. However, there are many types of cells in each region, and each may have to modify its behavior in a different way. Moreover, even when the same modification is being made, cells may use different receptors to trigger their changes.

Now, many regions in the brain will have to receive many different messages from different populations of dopaminergic neurons. Different cells within those regions will have synapses receiving dopamine from different populations of neurons, and many different synapses, receiving different messages, will probably use the same receptors.

But what do drugs do? They affect the activity of one or more specific type of receptor, either making them more or less sensitive, or "turning them on" without reference to whether the neurotransmitter is present at all. Their actions are, at best, specific to one type of receptor, and some, such as reuptake inhibitors, are independent of which receptors are used, although they may be specific to one type of reuptake transporter. Still, it's hard to imagine that there's any specific correlation between which receptor(s) are present, and which transporter is used. But as we saw above, there will be many different cells, with different functions, responding to different messages, all using the same receptors. And they'll all be affected.

What this means is that any drug used to affect the activity of these neurotransmitters is going to be like a shotgun, or maybe a claymore mine in its effect. It's hardly likely to only affect whatever specific issue you are trying to treat.


  1. Excellent point. Dopamine really has no meaning as such. But since humans think in terms of narratives, protagonists, and villains, we try to cram poor dopamine into a particular characterisation.

    There are no good words to describe the simultaneous activity of 100 billion neurons. By the time you get them out, the whole picture has changed. Several times.

  2. Great post! I think it is hard for even people in the field to keep this in mind. I tend to think of it as neurotransmitters being messengers, and the interpretation of the message is different depending on where the messenger hits. But even that is a simplistic view.

  3. Thanks for the comment, al fin. (And the plug earlier.)

    As for "describ[ing] the simultaneous activity of 100 billion neurons", I'd guess you need a landscape in n dimensions, or, rather, one for each region/area of the brain. Each nerve cell in one population (within layer, within area/region, etc.) samples a specific location in that space, and sends a scalar signal every so often (represented by instantaneous firing rate). It certainly seems to work that way in the early visual (V1-V4) areas.

    Of course, it gets much more complex when you take membrane-level calculations into account.

    As for narratives, I've been meaning to do something on that sometime, but with everything about cells/brains, I haven't had time.

  4. Thanks for the comment Sci...

    Of course, science has to create simplistic models, else it wouldn't be able to get anywhere modeling at all (see Kuhn, etc.)

    I just tend to see it as (part of) my function to point out where these models are tending to make people miss things they should (IMO) be looking at.

  5. This comment has been removed by the author.

  6. I really loved your blog, but it appears that you have put a lot more work into it. I will keep your blog in my twitter list so I can come back and see it again when it has some new information. Here is a link that I think you will find interesting.