Lysine targeting covalent inhibitors of malarial kinase Pf CLK3

Abstract

Malaria continues to devastate tropical regions of the world, with resistance to frontline drugs on the rise. Kinase inhibition has emerged as a promising novel mechanism of action in the fight against malaria. We previously reported the development of TCMDC-135051 (1), a highly potent, multi-stage inhibitor of Plasmodium falciparum CLK3 (PfCLK3). Building on this work, we subsequently developed the first covalent kinase inhibitor for malaria, selectively targeting a unique cysteine residue. Despite their high potency and selectivity, covalent inhibitors that target cysteine residues are particularly vulnerable to resistance arising from single point mutations of the nucleophilic residue. This work presents a novel strategy targeting the essential kinase catalytic lysine residue which has the potential to evade this resistance mechanism. Using structure based drug design, analogues of TCMDC-135051 (1) targeting Lys394 of PfCLK3 were developed. Four compounds, all harbouring benzaldehyde-based warheads, covalently engaged Lys394 as determined by protein mass spectrometry. These analogues were highly potent against recombinant protein, with good parasiticidal potency and cytotoxicity profiles. These molecules 4, 5, 8, 9 are the first lysine-targeting covalent inhibitors reported for malaria and offer a promising general strategy for future antimalarial drug discovery.

Fig. 1. a, Previously published co-crystal structure of TCMDC-135051 in PfCLK3 (PDB: 8RPC); b, structures of TCMDC-135051, compound 2 and 3; c, and d, molecular docking of compounds 2 and 3 respectively in the co-crystal structure.

Fig. 2. Designed covalent inhibitors 4–13 featuring salicylic aldehyde, ethynyl benzaldehyde, vinyl sulfone and acrylamide warheads.

Scheme 1. Synthesis of aryl warheads as boronic esters. (i) KHF₂, H₂O, MeCN, rt, 24 h; (ii) bis(pinacolato)diboron, KOAc, Pd(dppf)Cl₂·CH₂Cl₂ (10 mol%), 1,4-dioxane, 80 °C, 18 h; (iii) Pd(dppf)Cl₂·CH₂Cl₂, CuI, triisopropylsilylacetylene, NEt₃, 50 °C, 3 h; (iv) trifluoromethanesulfonyl chloride, NEt₃, CH₂Cl₂, 0 °C – rt, 2 h.

Scheme 2. Synthesis of compounds 2, 4, 6 and 8. (i) TsCl, NaH, THF, 0 °C, 2 h; (ii) LDA, I₂, THF, −78 °C, 3 h; (iii) (5-formyl-2-methoxyphenyl)boronic acid, Pd(PPh₃)₄ (10 mol%), Na₂CO₃, 1,4-dioxane, Δ, 12 h; (iv) diethylamine, NaBH(AcO)₃, 1,4-dioxane, rt, 12 h; (v) CH₃OH, KOH, Δ, 18 h; (vi) boronic ester, Pd(dppf)Cl₂·CH₂Cl₂ (5 mol%), Na₂CO₃, 1,4-dioxane, Δ, 0.5 h, μW; (vii) acryloyl chloride, DIPEA, DMF, rt, 2 h. *Obtained using conditions (vi) followed by treatment with CsF in DMF, rt, 0.5 h.

Scheme 3. Synthesis compounds 3, 5, 7 and 9. (i) TsCl, NaH, THF, 0 °C, 2 h; (ii) (5-formyl-2-methoxyphenyl)boronic acid, Pd(PPh₃)₄ (10 mol%), Na₂CO₃, 1,4-dioxane, Δ, 12 h; (iii) diethylamine, NaBH(AcO)₃, 1,4-dioxane, rt, 12 h; (iv) CH₃OH, KOH, reflux, 18 h; (v) NBS, CH₂Cl₂, rt, 2 h; (vi) boronic ester, XPhos Pd G2 (5 mol%), Na₂CO₃, 1,4-dioxane, Δ, 0.5 h, μW; (vii) acryloyl chloride, DIPEA, DMF, rt, 2 h. *Obtained using conditions (vi) followed by treatment with CsF in DMF, rt, 0.5 h.

Fig. 3. Protein mass spectrometry of compounds 2–13 incubated with PfCLK3. a, PfCLK3 apo spectrum b, PfCLK3-compound 4 adduct spectrum. c, Intact mass spectrometry results of compounds 2–9, 12 and 13. For compound 4, PfCLK3 with a C-terminal His-tag was used (sequence given in e). d, List of modified tryptic peptides observed after tryptic digest of compound 4 and 9 adducts. e, Sequences of PfCLK3 constructs used. Modified lysines shown in red.

Fig. 4. a, Lysine interacting moiety of compounds 4 and 6, docked into PfCLK3 (PDB: 8RPC) and overlayed. b, Proposed mechanism of 4's covalent reaction with Lys394, and the inability of 6 to react.

Fig. 5. Potency of compounds 1–9 against recombinant PfCLK3. a, Concentration response curve for the 2,4-substituted azaindoles (2, 4, 6 and 8) b, graph legend. c, Concentration response curve for the 3,5-substituted azaindoles (3, 5, 7 and 9). d, Table of apparent IC₅₀ values all tested compounds. Errors given as standard deviations, N ≥ 3. *Concentration response curve not complete.

Fig. 6. Parasiticidal potency of covalent compounds against Pf3D7 parasites. Errors given as standard deviations, N ≥ 3.

Fig. 7. Selectivity of compounds 1, S1, 4, 5, 8 and 9 against hCLK2 and cytotoxicity against HepG2 cells. Errors given as standard deviations, N ≥ 3.