(D) Dendrogram showing single point inhibition for 319 kinases for PKC412. to allow improved kinase inhibitor design of clinical brokers with enhanced efficacy and reduced toxicity. DOI: http://dx.doi.org/10.7554/eLife.03445.001 Thus, normal mature hematopoietic populations can be maintained in the context of either Flt3 or Kit inhibition alone but not dual Flt3/Kit inhibition (Bershtein et al., 2006). This synthetic lethal toxicity relationship between FLT3 and KIT for maintaining normal hematopoietic populations may explain the adverse side effects of the current kinase targeted drugs in clinical development. In a recent single agent Phase II trial, PKC412 failed to achieve a single complete remission (CR). When combined with cytotoxic brokers PKC412 showed some promise, achieving a 25% CR rate, but responses were primarily incomplete recovery of peripheral blood counts (CRi, 20%) with over 90% of patients developing grade 3/4 myelosuppression (Strati et al., 2014). While AC220 monotherapy impressively exhibited a 50% CR rate in a Phase II trial, these consisted primarily of CRi (45%) with few real CRs with complete recovery of blood counts (Cortes et al., 2013), correlating with the comparable potency of these brokers for both FLT3 and KIT. A recent study showed increased selectivity of the clinical agent crenolanib for FLT3 over KIT IOWH032 and reinforced the correlation between target inhibition, and anti-target avoidance (Dar et al., 2012), which lead to lowered toxicity towards normal hematopoiesis (Galanis et al., 2014). However, the potency of crenolanib for KIT remains too high (IC50 = 67 nM for p-KIT inhibition in TF-1 cells; 65% inhibition at 100 nM, in vitro) (Galanis et al., 2014). This is likely insufficient to fully minimize clinically relevant myelosuppression, as a recent interim analysis reported only a 17% (3/18 patients) composite CR rate in AML patients, with 2/3 of these responders achieving only CRi (Collins et al., 2014). These findings highlight the need for new clinical candidates IOWH032 that better minimize KIT and other Class III RTK inhibition. While avoiding inhibition of the presumed anti-target, KIT, is one chemical challenge toward inhibitor design, the emergence of on-target resistance is another clinical challenge. We (Smith et al., 2012) and others (Wodicka et al., 2010) have identified the acquisition of secondary FLT3 kinase domain name (KD) mutations that cause drug resistance as another limitation of current clinically active FLT3 inhibitors. Mutations at the activation loop residue D835 are particularly clinically problematic. These mutations are proposed to bias the kinase toward the CD36 constitutively conformation by disrupting a hydrogen bond from D835 to S838, and thus limit the efficacy of Type II inhibitors such as AC220. We have recently proposed that a Type I inhibitor, which binds to the active kinase conformation, would circumvent these mutations that confer resistance to AC220 (Smith et al., 2012). New small molecule therapies have been reported to bypass these particular mutations, including crenolanib (Galanis et al., 2014), a Type I inhibitor (Lee et al., 2014; Smith et al., 2014), but the CR rate of crenolanib remains modest (Collins et al., 2014). Moreover, it is likely that a repertoire of drugs will be necessary to combat emerging resistance. We propose herein a solution to the FLT3/KIT selectivity problem designed to avoid myelosuppression and also retain potency against drug-resistant mutations. The IOWH032 staurosporine scaffold has been utilized pharmacologically for 30 years, and staurosporine analogs have been proven to be potent FLT3 inhibitors (PKC412, CEP701) (Strati et al., 2014), though clinical activity of these compounds has been modest, perhaps caused by lack of potent FLT3 inhibition due to dose-limiting toxicity in vivo. The lactam ring C7 position remains virtually unexplored for modulating selectivity (Wood et al., 1999; Bishop et al., 2000; Heidel et al., 2005). We recently reported that C7-substituted staurosporine analogs, we term staralogs, are potent and selective inhibitors of engineered analog-sensitive (AS) kinases (Lopez et al., 2013). For example, when C7 (R1) equals isobutyl (Star 12), AS Src kinase is usually potently inhibited but WT kinases remain unaffected. However, we also observed that Star 12, in a panel of 319 kinases, IOWH032 weakly inhibits only one WT kinase, FLT3 (57% inhibition at 1 M; KIT, CSF1R, PDGFR/ all inhibited 10%). Thus, the C7-alkyl group of Star 12 may allow for weak but selective inhibition of FLT3 over the anti-target KIT, which contributes to myelosuppression when FLT3 is also inhibited. Although substitution of an isobutyl group at C7.