« Rats and High-Fructose Corn Syrup |
Main
| Drugs And Their Starting Points »
March 23, 2010
We Don't Know Beans About Biotin
Posted by Derek
You know, you'd think that we'd understand the way things bind to proteins well enough to be able to explain why biotin sticks so very, very tightly to avidins. That's one of the most impressive binding events in all of biology, short of pushing electrons and forming a solid chemical bond - biotin's stuck in there at femtomolar levels. It's so strong and so reliable that this interaction is the basis for untold numbers of laboratory and commercial assays - just hang a biotin off one thing, expose it to something else that has an avidin (most often streptavidin) coated on it, and it'll stick, or else something is Very Wrong. So we have that all figured out.
Wrong. Turns out that there's a substantial literature given to arguing about just why this binding is so tight. One group holds out for hydrophobic interactions (which seems rather weird to me, considering that biotin's rather polar by most standards). Another group has a hydrogen-bonding explanation, which (on the surface) seems more feasible to me. Now a new paper says that the computational methods applied so far can't handle electrostatic factors well, and that those are the real story.
I'm not going to take a strong position on any of these; I'll keep my head down while the computational catapults launch at each other. But it's definitely worth noting that we apparently can't explain the strongest binding site interaction that we know of. It's the sort of thing that we'd all like to be able to generate at will in our med-chem programs, but how can we do that when we don't even know what's causing it?
Comments (12)
+ TrackBacks (0) | Category: Drug Assays | In Silico
POST A COMMENT
- RELATED ENTRIES
- › Why Not Share More Bioactivity Data?
- › An Anticoagulant Antidote
- › Merck's Liptruzet: A Cause For Shame?
- › Your Brain Shifts Gears
- › Total Synthesis in Print
- › Things I Won't Work With: Dimethylcadmium
- › Another Germ Theory Victory - Back Pain?
- › An Update on Deuterium Drugs
1. PharmaHeretic on March 23, 2010 1:51 PM writes...
Link to the last publication does not work.
Permalink to Comment2. PedroS on March 23, 2010 2:18 PM writes...
The paper is
Permalink to Commenthttp://pubs.acs.org/doi/abs/10.1021/ja909575j
3. Wavefunction on March 23, 2010 2:21 PM writes...
Yes, the traditional explanation now is that biotin forms five strong hydrogen bonds by pushing out entropically and enthalpically "unhappy" water molecules from a hydrophobic region. The last link does not work by the way.
Permalink to Comment4. Sili on March 23, 2010 3:28 PM writes...
To the semi-trained eye it makes some sense, the the stronger the bonding, the harder it is to model. Weak interactions allow all sorts of perturbation approaches, but the stronger something binds the further it must be energetically from the (better understood?) precursors.
Permalink to Comment5. Hap on March 23, 2010 3:56 PM writes...
Wouldn't that imply that you are likely to miss lots of tight-binding inhibitors by computational methods though? Unless inhibitors use lots of weak interactions to bind tightly, inhibitors that bind tightly aren't going to be found or understood in silico because the interactions that would be occurring wouldn't be modeled well.
Permalink to Comment6. metaphysician on March 23, 2010 4:11 PM writes...
*cough* Am I the only one who is amused that biotin doesn't contain any tin?
*non-chemist, non-biologist*
Permalink to Comment7. Sili on March 23, 2010 4:15 PM writes...
I'd certainly expect them to be underestimated, and if you push perturbations too far, they do break down.
BUT I'm not a comp chemist. I've only taken the basic three courses in quantum chemistry at my uni, and that's ... closer and closer to a decade ago, so I may very well be utterly and completely off the mark.
Permalink to Comment8. hibob on March 23, 2010 5:34 PM writes...
@#5 Hap "Wouldn't that imply that you are likely to miss lots of tight-binding inhibitors by computational methods though? "
Oddly enough, computational methods miss lots of tight-binding inhibitors. Of course, they also miss most of the other inhibitors, some of the ones they do pick up are modeled as binding at the wrong binding site/ or using wrong orientation... and then there are the false positives! Using full non-linear Poisson-Boltzman equations for a big chunk of protein + solvent + other stuff is still non-trivial, I gather.
Permalink to Comment9. Henning Makholm on March 23, 2010 8:21 PM writes...
#6 methaphysician: Not half as amused as I am by the apparently vital importance of Pb compounds in drug discovery...
*neither*
Permalink to Comment10. Anonymous on March 23, 2010 10:27 PM writes...
Do not believe in those numerical data. Especially those in the supporting file, energies by residue, non-physical at all.
many of them do not make any sense. wonder how it got published by jacs.
Questions on the paper: (1)How big are the errors from each of the term? need to validate it by trying the same approaches for more experiment data, say HIV mutations, lots of data available. (2)Neutral form for the 2nd molecule with guanidine, kidding? (3)PBSA is not realible for Free energy calculation. check the dielectric constant=1 inside protein, 80 in the solvent, realible results from it? (4)the protein structure with or without ligand would be quite different, and contribute a lot to the binding, little info is available from the paper: only took a complex from pdb, and use the protein for another ligand with some MD. Big Q: what's the global minimum structure for the protein without ligand, how much change in structure and energies?
Permalink to Comment11. barry on March 24, 2010 12:36 AM writes...
wow, I casually speak as if I "understand" a binding if I've seen an x-ray diffraction structure of the complex, but nothing about the published co-structure 3FDC sets this one apart as extraordinary
Permalink to Comment12. Pete on March 24, 2010 7:07 AM writes...
I took a quick look at this and found a couple of things that worried me. Apparently in the bound state the cyclic guanidine in 2’-iminobiotin (BTN2) is neutral (i.e. not protonated). I don’t know the pKa of this cyclic guanidine but I would guess that it will high (I seem to recall a figure of 13+ for guanidine itself). This means the free energy that you measure for the binding of BTN2 will reflect the energy cost of converting the cationic (predominant form in solution) guanidinium to the neutral (bound) guanidine. I’m not sure if/how the authors are taking account of this but I suspect that much of a 6 kcal/mol difference in the binding free energies biotin and its imino analogue could be explained by free energy differences between the guanidinium and guanidine forms of BTN2.
The other concern I had was around the SCRF-like calculations. SCRF actually refers to an entire family of solvent models that are available to users of the Gaussian03 electronic structure program but the authors have not indicated which of these their procedure follows. Energies calculated with continuum models like these are sensitive to atomic radii and typically you need to parameterise radii and charge/QM-treatment together since you need larger radii for stronger electrostatics. In addition, nasty things can happen when you’ve got anions (the infamous outlying charge problem) and I am generally wary of physical interpretations of results of calculations like these.
Permalink to Comment