Rational anti-psychotics

On The Blob, I've had rather a lot to say about side-effects. Not my side-effects: I don't do drugs except maybe 2 aspirins in every 3 years. But I'm a member of the between generation which is caught between looking after aged parents and coming on stream to be available for grand-children. Both those categories eat significantly at the trough of medicaments: the former because their immune system is running out of steam, the latter because theirs isn't up to speed yet; therefore they both need a bit of pharmaceutical help. Being inveterately curious, I read the small print of all the drugs I dispense to Pat the Salt and wonder at the diversity of unwonted effects that potentially come with a therapeutic. The key problem is that all the natural molecules which control our metabolism and immune system - all the hormones, cytokines and neurotransmitters - are double- and triple-jobbing. Heck, dopamine and adrenalin don't even know if they are neurotransmitters or hormones. In fact of course they are both, in the former case acting across a synapse 20nm wide, in the latter the action occurs up to 2 m from the site of production. In both cases, they only work if and because there is a receptor with which to get up close and personal.

The receptors are often as much or more interesting than the "ligand" which binds to them. Many receptors, including all 1,000 olfactory receptors up our noses, are GPCRs [G-protein-coupled receptors] which all consist of 7 transmembrane domains which loop back and forth through the cell-membrane. If you think 7 stubby pencils making a hollow tube you won't be far wrong. Al though the basic structure of each GPCR is the same, the devil is in the detail: each one is subtly different and therefore has a peculiar and particular affinity for a specific molecule. One olfactory receptor will respond to patchouli while another will sing to vanilla. Over the eons of evolution, the small signaling molecules have accumulated a variety of different effects. Dopamine, for example is a key piece in the puzzle that is Parkinson's disease. Without it, the subtle contraction of muscles that keeps a hand steady enough to drink a glass of water all falls apart and the hand yaws and bucks all over the slop trying desperately to control things. And we've also met dopamine controlling the subtle growth of the eyeball, so that the retina is perfectly in focus with the lens. One of the consequences of this multi-tasking is that there are, not one, but five dopamine receptors. They are all clearly descended from an original protein but have evolved each in its own direction to better serve one of the dopamine functions.

Here's a nice little News & Views by Sibley and Sei from NIH about a recently revealed crystal structure of a drug and its dock. All that intro about the manifold action of dopamine is relevant here because the drugs investigated work by sitting into a dopamine receptor and preventing the normal occupant from getting access. No access, no dopamine-triggered response. This is just bonzer for conditions where over-sensitivity to dopamine is the issue. Schizophrenia can be thought of as anti-Parkinson's only in the sense that one is about too much dopamine activity and the other is about too little. Clinicians and drug-designers are pretty much agreed that drugs against Scz are ideally tailored to fit D2R, the dopamine-2-receptor. The picture at the top of the film strip above, shows Risperidone sitting into the cleft crafted by evolution to fit dopamine. The other two panels show two different anti-Scz drugs Eticlopride and Nemonopride docked into two of the other dopamine receptors. If you go only from the pictures it looks like Eticlopride and Nemonopride lie across the groove while Risperidone  stands up-and-down. Sibley and Sei make much of how well Risperidone fits D2R the optimum receptor noting particular adjacent amino acids. That's what rational drug design is all about, you tweak your product in an evidence-based fashion so that it works better. But we know so little about the dynamic processes of ligand-receptor interaction, that a static 3-D snapshot only seems to give clues to function.

Right at the end Sibley and Sei note that we don't have the counterpoint data: Risperidone squeezed into D3R and D4R and Eticlopride and Nemonopride doing their best in D2R. Without those controls much of what they claim is, well, moot.

Johnson & Johnson, whose subsidiary Janssen developed Risperidone  in the early 1990s, has been shelling out Big Bucks as compensation for marketing the drug for conditions beyond those for which it is licensed. It's a dog eat drug world out there as each Pharma sales-force tries to reach the threshold for a year-end bonus.