Small molecule BLVRB redox inhibitor promotes megakaryocytopoiesis and stress thrombopoiesis in vivo

Figures

• 

  Fig. 1. Computational modeling of BLVRB inhibitors. a Schema of the heme degradation pathway (heme → biliverdin → bilirubin) highlights bifurcated and non-redundant NAD(P)H-dependent biliverdin reductases (BLVRA and BLVRB) and isomeric substrate restrictions. b Self-docking ligand RMSD (root-mean-squared) matrix of 12 BLVRB/NAD(P)H/inhibitor (or substrate) trimolecular complexes [1HE4 (FMN), 1HE5 (Lumichrome), 5OOG (Phloxine B), 5OOH (Erythrosin extra bluish), 7ER7 (Tamibarotene), 7ER8 (Sulfasalazine), 7ER9 (Febuxostat), 7ERA (Olsalazine), 7ERB (Ataluren), 7ERC (Deferasirox), 7ERD (Flunixin meglumin), 7ERE (Pyrantel pamoate)]; each row represents a structure and the column identifies the scoring functions. Values shown are the RMSD (Ǻ) of the best-scoring docked pose to the crystal structure pose, highlighted and depicted using a colorimetric scale for values from 0 to 12 Å. The best docking/scoring schemes were selected using average RMSD values ≤ 2.0 Å, highlighting 7ERA and 7ERC (green rectangles) as docking prototypes (ataluren RMSD is 0.6 for 7ERA and 3.0 for 7ERC). c–e Protein-ligand interactions within the binding site of 7ERA with ataluren (c) highlight hydrogen bonds with HIS153…

• 

  Fig. 2. Structural determinants of BLVRB inhibitors. a Schema outlines the workflow for small molecule screening and metabolic profiling. The scattergram shows the correlation between flavin reductase (FR) and DCPIP reduction (DR) IC50 determinations; ***p < 0.001. b, c Core scaffolds with summary structure/activity relations of diazabicyclo compounds (Series 1, b), or pyrazolopyrimidinones (Series 2, c). For both panels, the core ring is highlighted blue, and critically-modified sites analyzed by SARS are shown (brown, yellow, green; refer to text for details). d, e BLVRB trimolecular structures show BLVRB/NADP (gray/cyan) in complex with BCT1028 (green, d), or BCT2029 (green, e) adjacent to the NADP nicotinamide ring. Both panels feature 2mFo-DFc weighted electron density maps contoured at 1.2σ. Structural elements lining the ligand-binding site and the NADP+ cofactor are labeled, H-bonding and salt-bridging interactions are represented as dashed lines, and ordered water molecules are represented as red spheres. Note that His153 forms a polar interaction with the BCT2029 pyrimidinone carbonyl comparable to that observed in the BLVRB/NADP+/BCT1028 structure, in which His153 forms…

• 

  Fig. 3. BCT1028 metabolic profiles. a PK determinations were assayed in mice (age 9–11 weeks, weight 25–28 g) after single-dose (IV or PO) exposure, followed by serial blood sampling for LC/MS/MS quantification of BCT1028; data represent the mean ± SEM, N = 3 mice/time point. Critical pharmacokinetic parameters (T1/2 (half-life) Cmax (peak concentration), AUC (area under the curve)) are shown (also see Supplementary Fig. 4). b Scattergram shows the activity profile of N = 44 pharmacologic targets using single-point (10 μM) BCT1028 inhibitory screen; data expressed as the mean ± SEM from duplicate determinations (target details provided in Supplementary Table 2). The single target demonstrating >80% inhibition (δ opioid receptor (OPRD1, red circle)) is shown, along with two targets (blue circles) demonstrating modest <50% inhibition (cholecystokinin A receptor (CCKAR)) and 5-hydroxytryptamine receptor 1A (HTR1A).

• 

  Fig. 4. Lineage effects of BCT1028 in erythro/megakaryocyte models. a Heat diagram shows relative expression patterns (RPKM) of heme degradation genes from aggregate NCI60 cell lines (inset delineates 6-member hematopoietic cells); scale bar is shown. The immunoblot (inset, repeated twice) confirms BLVRB expression in K562 and HEL, but not MOLT4 cell lysates (10 μg/lane). BLVRA and actin immunoreactive species are shown as controls. b Dose-dependent DCPIP reduction (DR activity, Δ absorption minute−1) was assayed in K562, HEL, or MOLT4 cell lysates (mean ± SEM, N = 3 independent replicates); p-values using unpaired t-test ** <0.01, *** <0.001. c K562 lysates (6 μg) were incubated with individual inhibitors for 60 min, followed by DR activity assay (mean ± SEM, N = 3 independent replicates); p-values using unpaired t-test. d K562 cells were incubated with varying concentrations of BCT1028 (or v/v-equivalent DMSO as control) for 24 h, washed ×3, and DR activity assays were completed on cellular lysates (6 μg/sample); data represent mean ± SEM, N = 3 independent replicates; p-values **** <0.0001, *** <0.001 using unpaired t-test. e, f K562 cells were supplemented with 1 μM PMA, 10 μM…

• 

  Fig. 5. Hematopoietic and in vivo studies of BLVRB inhibition using BCT1028. a–c CD34+ HSCs (1 × 104/mL) were propagated in stem cell expansion media with thrombopoietin (TPO, 10 ng/mL) supplemented without (DMSO v/v control), or with 10 μM BCT1028, BCT2066, or ataluren, followed by flow cytometric detection of cell-surface CD41a (integrin αIIB, upper panels) or CD61 (β3 integrin, lower panels); box plots (interquartile range, median, and minimum/maximum data range) display the normalized fold-change compared to DMSO controls at each time point (the mean control CD% is shown as reference). Data are from (N = 8 biological and technical replicates; p-value *<0.05, ***<0.001, ****<0.0001 using unpaired t-test or Mann–Whitney test. In (b), a representative scattergram (refer to gating strategy, Supplementary Fig. 7a) highlights differences within the CD41a+ MKs between BCT1028+ and BCT1028− subsets (upper panel), with corresponding histogram (lower panel). c delineates differences in CD41a cell-surface abundance (Days 7 and 10) between BCT1028+ and BCT1028− subsets (upper panel) or time-dependent side scatter (MK cellular complexity, lower panel); data in upper and lower panels represent…

• 

  Scheme 1. Synthesis of phenyl-diazabicyclo derivatives.

• 

  Scheme 2. Synthesis of pyrazolyl-diazobyciclic derivatives.

• 

  Scheme 3. Synthesis of pyrazolopyrimidinone derivatives.