Positive allosteric modifiers — exciting (and humbling)

Not all drugs are agonists or antagonists. If you’re unsure what this means — have a look at https://luysii.wordpress.com/2010/08/29/some-basic-pharmacology-for-the-college-student/  which should give you all the background you need.  For a nice discussion of allosteric modifiers is which is (1) more leisurely (2) filled with nice pictures see http://www.scientificamerican.com/article.cfm?id=a-biochemical-way-to-reduce.  It’s in the August 2009 issue.

The benzodiazepines (xanax, librium, valium, etc. etc.) and the barbiturates bind to a protein receptor for gamma amino butyric acid (the GABA[A] receptor).  Unlike D-Tc which binds to the place acetyl choline wants to bind on its receptor and is thus a competitive antagonist, these drugs bind to a second site on the GABA[A] receptor, causing a conformational change, which makes the receptor more responsive to its natural ligand (gamma amino butyric acid).   This conformational change is called an allosteric effect and the place on the protein where these drugs bind is called an allosteric site.  Allosteric effects can be either positive, as in the example above, or negative.

A meatball classification of protein receptors on the surface of brain cells which control their behavior breaks them into two categories (1) ion channels (2) proteins which cross the neuronal membrane 7 times.  Ion channels are the easiest to understand.  Something binds to them and they open.  The channels don’t let everything in, just particular ions.  Thus one type of receptor for acetyl choline lets in sodium (not potassium, not calcium) which causes the cell to fire impulses.  The GABA[A] receptor (for gamma amino butyric acid) lets in chloride ions (and little else) which inhibits the neuron carrying it from firing.  (This is why the benzodiazepines and barbiturates are anticonvulsants).

The second class of receptor has the most drugs for it (L-DOPA, LSD, morphine and its fellow narcotics, etc. etc. ).  We’d love to know more about their structure, but the fact that they pass through the membrane 7 times means that the parts going through the membrane (which is basically all lipid) can’t be removed from the membrane without destroying the actual 3 dimensional shape of the protein (which is what drug chemists, biochemists, etc. etc.) are interested in.  All sorts of tricks have been applied, but even as of 7/10 the structures of only 5 of them were known (Nature vol. 466 pp. 544 ’10).

Now we’ve got a lot of drugs which bind to the 7 transmembrane class (7TM from now on).  All of these drugs bind to 7TMs which themselves bind rather small molecules (dopamine, norepinephrine, serotonin etc. etc.).  But some 7TM receptors bind peptides — the most famous is met-enkephalin, the brain’s endogenous morphine.  Many of these peptides bound are large — somatostatin (14 amino acids), neuropeptide Y (36 amino acids), galanin (29 amino acids).  The list goes on.

Making drugs which interact with these 7TM polypeptide receptors is easy in cell culture (where you just throw stuff at the receptor and watch what happens). Getting them into the brain is another matter.  Something called the blood brain barrier (aka BBB, to be discussed in another post) keeps everything out of the brain with a molecular mass of more than 400 daltons (about 4 amino acids).  Aside: that’s why I’ve always found statements about blood levels of endorphins and enkephalins rather hilarious — you can give them intravenously, but they don’t get into the brain.  While exercise may make you feel good, and cause a rise in ‘endolphins’  (see Meryl Streep in Postcards from the Edge) in your blood, the first doesn’t follow from the second.

So if you want a drug which affects these 7TM polypeptide receptors in the brain, it isn’t going to be a peptide.  How do you find such drugs? Certainly not by our understanding of protein structure and dynamics.  Just make a lot of different compounds (combinatorial library) and throw them at cells containing the receptor you’re interested in.  This is how galnon (a low molecular weight galanin receptor agonist) was found.

I’ve not really talked about all the cool things that polypeptide ligands for 7TM receptors do in the brain, but they make neuropharmacologists drool with desire.  Galanin looks to be something in the brain which stops convulsions. Galnon has anticonvulsant activity.  Galanin also decreases morphine withdrawal signs (all this in animals).  NeuropeptideY is thought to be important in appetite. The list goes on and on.

Galnon doesn’t bind to the galanin receptor very well.  The equilibrium constant is in the microMolar range (most useful drugs have equilibrium constants in the nanoMolar range).   That’s why [Proc. Natl. Acad. Sci. vol. 107 pp. 14943 – 14944, 15229, 15234 ’10 ] is so exciting.  A low molecular weight compound (CYM2503)  binds to one of the galanin receptors (we have 3) and increases the effects of galanin when it binds.  By itself it has no effect on the receptor.  So it is a positive allosteric modifier.  Like galnon, CYM2503  has anticonvulsant activity.  The fact that it doesn’t do anything to the receptor by itself makes it likely that it will have relatively few side effects.  The experienced drug chemists reading this are undoubtedly snickering at this point.

Now for the humbling part:  Think how far away we are in our understanding of protein structure from designing such a drug.  Here’s what you’d need to be able to predict.  (1) the conformation of the protein in its normal state { by the way one of the 3 galanin receptors contains 370 amino acids } (2) the conformation of the protein as you’d like to have it { so it was more or less active } (3) the actual location of the allosteric site { to my knowledge NO allosteric site in any protein has ever been predicted a priori — correct me if I’m wrong } (4) a drug which would bind to the site and cause the desired allosteric shift.   We are light years away from this.  It certainly is a fit task for smart people to work on.  While organic chemistry does all sorts of elegant and marvelous tricks with smaller molecules (mostly under 1000 Daltons from the examples in Clayden), it isn’t sophisticated or advanced enough to tackle this sort of thing.  We’ve certainly come a long way in my lifetime, but we’ve much, much farther to go.


Advertisements
Post a comment or leave a trackback: Trackback URL.

Comments

  • Yggdrasil  On September 9, 2010 at 11:56 am

    Another problem with crystallizing GPCRs (i.e. 7TM receptors) is that these types of receptors seem to be very conformationally flexible which makes it difficult to get crystals that diffract well (conformational heterogeneity in a crystal will degrade the information from the high resolution diffaction spots). Indeed, the first GPCR crystallized was rhodopsin, which is basically locked into a particular conformation by the bound retinol cofactor. Similarly, for the β2-adrenergic receptors, the Kolbilka lab had to lock the receptor into a certain conformation by either binding an antibody to it or replacing one of the intracellular loops with a protein that crystallizes well.

    Because antagonists bind to these receptors with very high affinities (and very low dissociation rates) and likely lock the receptors into a single (inactive) conformation, they are likely to be very useful tools for the structural biologists who are trying to tackle other GPCR structures.

  • Wavefunction  On September 10, 2010 at 9:49 am

    Also worth noting that all the GPCRs that are out until now have been crystallized in inactive states with inverse agonists/antagonists. The agonist-bound active state is the really interesting one.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: