Phosphorylation state-dependent binding kinetics of BTK inhibitor fenebrutinib

Elucidating the selectivity of inhibitors for specific kinase phosphorylation states

Bruton’s tyrosine kinase (BTK) is an attractive target for the treatment of B-cell lymphomas as well as autoimmune diseases.
Three BTK inhibitors, i.e., ibrutinib, acalabrutinib and zanubrutinib, have thus-far been approved by the FDA and each inhibit the kinase through irreversible, covalent binding.
Several reversible inhibitors of BTK (including LOXO-305, ARQ-531 and vecabrutinib) are currently evaluated in clinical trials for different cancer types, while the likewise reversible BTK inhibitor fenebrutinib is evaluated in a phase III trial for multiple sclerosis.
Each of these inhibitors target the ATP binding pocket in the BTK kinase domain.
Despite the difference in ATP affinity between full-length BTK (K_M = 4 μM) and the BTK kinase domain (K_M = 22 μM), similar IC₅₀ values are found for most BTK inhibitors when comparing these proteins in enzymatic assays.
However, fenebrutinib is considerably more potent on the BTK kinase domain compared to full-length BTK, which may be attributed to autophosphorylation of the Y551 residue in the active site of full-length BTK (Dinh et al., 2007), but not the isolated kinase domain.
To study the effect of the phosphorylation state of BTK on the binding of fenebrutinib, surface plasmon resonance experiments were performed using either non-activated BTK or BTK activated by ATP treatment.
Fenebrutinib binds strongly to non-activated BTK with an extended target residence time.
In contrast, the inhibitor rapidly dissociates from activated BTK followed by reaching a plateau, which cannot be fitted properly with the 1:1 binding model.
An alternative model is proposed for the binding of fenebrutinib to activated BTK, which assumes that the enzyme preparation is heterogeneous in its phosphorylation state.
The binding kinetics of fenebrutinib to phosphorylated BTK can be extracted from this model, which shows a more than 100-fold lower affinity compared to non-phosphorylated BTK.
The same difference in binding kinetics is not observed for ARQ-531 and vecabrutinib, suggesting that fenebrutinib makes a unique interaction with Y551, which is in accordance with its crystal structure when bound to BTK (Crawford et al., 2018).

References

Dinh et al. (2007) Activation mechanism and steady state kinetics of bruton’s tyrosine kinase, The Journal of Biological Chemistry, 282 (12):8768-8776.

Crawford et al. (2018) Discovery of GDC-0853: a potent, selective, and noncovalent bruton’s tyrosine kinase inhibitor in early clinical development, Journal of Medicinal Chemistry, 61 (6):2227-2245.

Correlation between BTK inhibitor IC₅₀ values measured on the BTK kinase domain (residues 389–659) versus the full-length BTK enzyme.

Binding curves of fenebrutinib on activated and non-activated BTK fitted with the 1:1 binding model using the Biacore T200 Evaluation software (left), and immunoblot analysis of the phosphorylation state of the BTK preparations (right).

Alternative binding model for fenebrutinib on activated BTK, assuming that the enzyme preparation is heterogeneous in phosphorylation state.

Comparison of the association (k_a) and dissociation (k_d) rates of BTK inhibitors on activated and non-activated BTK. The dotted lines indicate equilibrium dissociation constants (K_D) calculated from the on- and off rates.