Taming the induced folding of drug-targeted kinases

Artificial Intelligence Teaches Drugs to Target Proteins by Tackling the Induced Folding Problem

“…For example, proteins such as transcription factors typically possess only induced structures; that is, their fold cannot prevail in an "apo" (separated) form, 7 and antibodies usually present antigeninduced conformational multiplicity, 8 while proteins which constitute drug targets often have a variety of "holo" (withincomplex) forms, depending on the ligand/drug they associate with. 9 The latter context is the focus of this work and may be illustrated by the structural diversity of the stress-responsive molecular chaperone Hsp90, 10 a cancer target that adopts different structures depending on the drug−target complex (Figure 1). This type of induced folding diversity should prompt rational drug designers to focus on inferring the target's IFE, rather than a unique 3D structure, probably only suitable for autonomous folders, 11 forming nonobligatory complexes.…”

“…TARGET A PROTEIN Drug-based targeted therapy, aimed at blocking specific dysfunctional proteins, often faces a major obstacle due to induced folding, a hard-to-predict phenomenon that often generates unexpected and undesirable cross-reactivity, while making the intended target elusive to molecular recognition. 9 We advocate that DL can steer drug design to achieve therapeutic impact by controlling the induced folding in the target protein. The way DL may teach the drug to target the protein is apparent as we focus on reworking the anticancer drug imatinib 18 into the prototype WBZ_4 9,13 (Figure 3).…”

“…However, we found an exception to the steric hindrance rule in JNK1/2, which has no detectable affinity for imatinib but is a target for WBZ_4. 9 JNK stands out in the kinome-wide IFE inference because of its significant induced folding diversity in the targeted region and its susceptibility to getting wrapped by the extra methyl group not present in the imatinib scaffold, as shown in Figure 3. In fact, the nanomolar affinity of WBZ_4 against JNK1 has been experimentally documented, 9 making this compound potentially impactful on ovarian cancer, where JNK inhibition has shown clinical relevance, at a variance with imatinib, which is not active in that therapeutic context.…”

Intrinsic protein disorder in human pathways

“…6 Disorder is important to key cellular processes involving nucleic acid and protein binding such as transcription regulation 7 and cell signaling, 8–11 IDRs are also implicated in pathological conditions including cancer, diabetes, and heart disease, 12,13 and can be potential drug targets. 14,15 …”

Pharmaceutical Industry at the Post-Genomic Junction

“…More than ever, the lead in the pharmaceutical industry depends on the ability to harness innovative research. In this regard, wrapping [2][3][4][5][6][7][8][9][10][11][12][13][14][15], a basic concept stemming from fundamental bio molecular research, may well hold potential to broaden the technological base of the industry and enable a vigorous recovery.…”