Tag Archives: Undruggable targets

The Gordian knot untied, the undruggable surface drugged

The claim is made that the surface of a coiled coil is flat hence undruggable because there are no nooks and crannies for a drug to fit in.  Well the surface of a coiled coil isn’t really flat but  PNAS vol. 117 pp. 17195 – 17203 ’20 did actually drug one such structure using FKBP12 to help it.

Coiled coil containing proteins are interesting.  There are roughly 3.5 amino acids per turn of each alpha helix, so to get a repeating structure 7 amino acids are required.  They are numbered a through g from amino to carboxy.  Positions a and d are usually hydrophobic amino acids (Leu, Ile, Val, Ala), e and g are usually polar and charged.  The non polar a and d side chains associate with the opposite alpha helix by knobs into holes packing.  Each individuual helix is right handed but the two helices wind around each other with a right handed turn.  Actually it’s a bit more complicated — there are 3.64 amino acids/turn (not 3.5), so to get the number of amino acids down so there are 3.5 per turn (allowing the structure to repeat every 7 amino acids, left handed supercoiling of each helix occurs.  No question, this is an elegant design.  Pauling may have figured out the alpha helix, but I doubt that anyone proposed the coiled coil before it was actually discovered — does anyone out there know?

All intermediate filament proteins (neurofilaments, vimentin, keratin) contain a coiled coil structure.

CEP250 is a large protein (2,442 amino acids) found in the microtubule organizing center (centrosome).  It has a lot of long alpha helices wound around each other.  The picture they show has two alpha helices (each with 24 or so turns) wound around each other — that’s what a coiled coil is.

The authors used a small (12 kiloDalton) widely distributed protein (FKBP12) to accomplish this.  FKBP12 is important because it can complex the immunosuppressants, rapamycin and FK506 allowing them to bind to their targets (mTOR and calcineurin respectively).  Basically FKBP12 forms a torus with the coiled coil inside, with the drug at the hole in the torus where it is held against the coiled coil — clever work.

This isn’t to deny the difficulty of drugging a flat surface which unfortunately characterize most protein protein interfaces.