Approved FLT3 drugs are kinetic and conformation selective

Exploring activation state-dependent kinetics to discover novel drugs

Avoid missing promising compounds using KINETICfinder®. Our assays will help you to:

  • Design drugs with optimal binding kinetics and improved efficacy, selectivity or higher barrier to resistance development. 
  • Explore systematically activated and non-activated kinases to identify novel chemotypes and binding modes.
  • Use phosphorylation state-dependent binding kinetics to interpret structural data.

Approved FLT3 drugs are kinetic and conformation selective

The FLT3 receptor tyrosine kinase consists of an extracellular domain, a transmembrane region and a cytoplasmic juxtamembrane (JM) region and kinase domain.

  • JM domain acts as an autoinhibitory loop that interacts with the catalytic domain (JM-in), stabilizing FLT3 in an inactive, autoinhibited state in which the activation loop is closed (DFG-out).
  • JM domain phosphorylation leads to the non-autoinhibited state (JM-out) where the DFG motif is unlocked and able to switch from the inactive (DFG-out) to the open activated state (DFG-in).1

Internal tandem duplication (ITD) mutations within the JM domain contribute to the majority of FLT3 activating mutations in acute myelogenous leukemia (AML).

  • These mutations disrupt the JM autoinhibitory activity, resulting in constitutive FLT3 activation.
  • Type I (sunitinib, midostaurin, gilteritinib) and type II inhibitors (foretinib) can target ITD mutations.
  • Midostaurin and gilteritinib are the only FDA approved inhibitors for the treatment of AML.2

Our data establish that the activation states of FLT3 can have a large impact on inhibitor affinity and binding kinetics, conferring to FDA-approved inhibitors optimal clinical efficacy and safety profile.

Clinically effective drugs display long-lasting inhibition of FLT3-ITD

All inhibitors bind to both FLT3-ITD and non-autoinhibited FLT3 forms.

  • Midostaurin and sunitinib have similar kinetic profiles and affinities for the activating mutation and non-autoinhibited state.
  • Gilteritinib is 60 times more potent with FLT3-ITD due to a substantial increase in the residence time (30X).
  • Type II foretinib has 4-fold lower affinity for the activating mutation caused by a shorter residence time. Our results confirm the preferential binding of foretinib to the inactive (DFG-out) non-autoinhibited state.
  • Midostaurin and gilteritinib are long-lasting inhibitors (>60 min) of FLT3-ITD, a driver mutation in AML.
CompoundOn‑rate (M‑1s‑1)Residence time (min)Kd (nM)
Sunitinib 1.2×10740.3
Midostaurin1.0×1046127
Gilteritinib3.5×106620.1
Foretinib4.3×1043312
Table 1. FLT3-ITD: Association rate, residence time and affinity values of sunitinib, midostaurin, gilteritinib and foretinib for the activating mutation.
CompoundOn‑rate (M‑1s‑1)Residence time (min)Kd (nM)
Sunitinib 7.9×10670.3
Midostaurin1.9×1044221
Gilteritinib1.7×10624.7
Foretinib5.4×1041112.8
Table 2. Non-autoinhibited FLT3: Association rate, residence time and affinity values of sunitinib, midostaurin, gilteritinib and foretinib for the non-autoinhibited state.

Figure 1. Kinetic curves of gilteritinib for the constitutive active, non-autoinhibited and autoinhibited states of FLT3 obtained with KINETICfinder® showing different binding kinetics depending on the conformation of FLT3.

JM domain substantially impacts binding kinetics and affinity of gilteritinib and midostaurin

Binding affinity is consistently reduced by the insertion of the JM domain into the active site of FLT3 because of a significant reduction in the association rate.

  • Sunitinib binds to all the activation sates with high affinity. However, the JM domain slightly interferes with sunitinib binding, suggesting steric hindrances.
  • Our results reveal that rearrangement of the JM domain in the autoinhibited conformation is required for midostaurin and gilteritinib binding. In fact, these inhibitors have 100- and 20,000-fold greater affinity for the activating mutation.
  • Midostaurin and gilteritinib present exquisite selectivity for the fully active state (JM-out, DFG-in) relative to the autoinhibited state, minimizing undesired effects on normal cells.
Figure 2. Comparison of the association (kon) and dissociation (koff) rates of sunitinib, foretinib, midostaurin and gilteritinib on activated and autoinhibited FLT3 obtained with KINETICfinder®. The dashed lines refer to the Kd values.

These data exhibit that the prolonged interaction of midostaurin and gilteritinib with the FLT3-ITD conformation along with the high selectivity relative to the autoinhibited FLT3 conformation, contribute to their clinical success in the treatment of AML.

References

  1. Griffith J. et al. (2004) The structural basis for autoinhibition of FLT3 by the juxtamembrane domain. Mol Cell 30;13(2):169-78.
  2. Stanchina M. et al (2020) Advances in Acute Myeloid Leukemia: Recently Approved Therapies and Drugs in Development. Cancers (Basel) 12(11):3225.