S T R U C T U R A L B I O L O G Y

S C I E N T I F I C H I G H L I G H T S

3 6 H I G H L I G H T S 2 0 2 1 I

Exploiting a transient protein conformation for generation of selective ligands of FKBP51

FK506-binding protein 51 (FKBP51) has emerged as a promising drug target for stress-related disorders, obesity-induced diabetes and chronic pain. The development of selective ligands is challenging because of the high similarity to other members of the FKBP family. X-ray crystallography data revealed that such ligands exploit a transient protein conformation.

The generation of compounds that exhibit a high affinity towards the target protein but discriminate against close homologues is a challenging task in chemical biology and medicinal chemistry. High selectivity is needed to unambiguously interrogate the target and to reduce side effects of a potential drug. A promising strategy to develop selective ligands is to exploit differences in protein dynamics [1].

A prominent example for such a problem is the family of FK506-binding proteins (FKBPs), which share a highly similar binding pocket in their peptidyl-prolyl isomerase (PPIase) domains, to which the immunosuppressant drugs FK506 and rapamycin bind (Figure 21). The larger homolog FKBP51 is a regulator of glucocorticoid receptor signalling, and has emerged as a potential drug target for stress-related disorders, obesity-induced diabetes and chronic pain. Due to their divergent biological functions, FKBP52, FKBP12 and FKBP12.6 are considered as anti-

targets. For instance, FKBP52 acts as a biological counter- player to FKBP51, and FKBP52 deficiency impairs sexual development and female fertility. FKBP12 and FKBP12.6 are cofactors of the ryanodine and TGFb/ALK receptors and play important roles in cardiac function.

The FKBP51-FKBP52 selectivity issue was solved by the development of ligands of the SAFit class (Selective Antagonists of FKBP51 by induced fit), which bind 10 000-fold stronger to FKBP51 compared to FKBP52 [2]. Interestingly, the selectivity of these ligands is due to binding to a minor conformation of FKBP51, which is highly disfavoured in FKBP52. However, ligands of the SAFit class still bind to FKBP12 and FKBP12.6 with substantial affinities, requiring optimisation of these ligands. Macrocyclisation, which is often used for improving drug-like properties for large compounds [3], was used to modify the ligands. Surprisingly, using different biochemical and biophysical assays, it was found that the generated ligands were selective against both FKBP52 and the off-targets FKBP12 and FKBP12.6 [4]. This unexpected selectivity profile was confirmed in cell culture, indicating that this ligand series is indeed the first compound class that robustly discriminates between FKBP51 and FKBP12/FKBP12.6.

In order to investigate the molecular basis for the unexpected selectivity profile of the macrocyclic SAFit analogues, the crystal structures of the complexes were solved using data collected at beamlines ID30B and ID23-1, revealing the molecular details of the conformational changes occurring upon binding at atomic resolution (Figure 22).

Fig. 21: Superimposition of FKBP12 (red, PDB-ID: 1FKJ), FKBP12.6 (blue, PDB-ID: 5HKG) and the FK1 domains of FKBP51 (grey, PDB-ID: 3O5R) and FKBP52 (magenta, PDB-ID: 4LAX), in complex with FK506 or rapamycin (not shown for clarity). Key amino acids of the binding pocket are shown in stick representation.

Fig. 22: Crystal structure of the FK1 domain of FKBP51 (light blue, PDB-ID: 7AOU) in complex with the macrocyclic SAFit analogue 13d (dark green sticks) superimposed on the co-crystal structure of FKBP51 (light green, PDB-ID: 4TW7) bound to the SAFit analogue iFit4 (not shown for clarity). The beta strand that is rearranged upon binding of 13d is highlighted in cyan and green, respectively, and the key moiety of 13d responsible for the conformational changes is highlighted in magenta.