(A) Peptide-IP/LC-MS/MS was used to identify proteins dephosphorylated upon gefitinib treatment. HL-60 cells were treated with vehicle (DMSO) or gefitinib at 10 μM for 10 minutes. Phosphotyrosine peptides were identified in Syk, Cbl, and Vav1 that were dephosphorylated with gefitinib. The LC-MS/MS extracted ion currents (XIC) for all identified pTyr peptide precursor ions belonging to an individual protein were summed and used to calculate differential expression ratios in DMSO- and gefitinib-treated cells. The specific phosphopeptides and pTyr sites observed are shown in Table S1. CPY01,2,3,5 are control pTyr peptides spiked into each sample at the same concentration. Positive ratios indicate higher expression for DMSO treatment, negative ratios indicate higher expression for gefitinib treatment. Ratios of 13 and −13 indicate that phosphopeptides for a protein were only detected in one of the treatments.
(B) High-throughput RNAi Screen Performance. Five scoring methods were used to identify hairpins inducing a myeloid differentiation signature (Summed Score, Weighted Summed Score, Naïve Bayes, K-Nearest Neighbor, and Support Vector Machine). The majority of shRNAs did not score by any method. The SYK-specific shRNA (Syk_1) scored in all methods.

(A) Knockdown performance of shRNAs targeting SYK in HL-60 and U937 cell lines. Transcript levels were measured at 5 days post-infection using real-time PCR and are reported for each hairpin (Syk_1, Syk_7, and Syk_10) relative to the transcript level of a luciferase shRNA control (Luc). Error bars depict mean ±standard deviation (SD) across three replicates.
(B) HL-60 and U937 cells were infected with shRNAs targeting SYK and luciferase (Luc) and differentiation evaluated 7 days post-infection. A 32-gene differentiation signature was quantified by the LMA/bead-based approach and a Weighted Summed Score (Differentiation Score) calculated for all genes. Error bars denote mean ±SD of 16 replicates. The two SYK-specific shRNAs that induced the greatest degree of gene suppression also induced the highest Differentiation Score.
(C) May Grunwald Giemsa staining of HL-60 and U937 cells 10 and 12 days post-infection, respectively, with shRNAs targeting SYK demonstrates cellular differentiation when compared to a luciferase shRNA control. Images were acquired with an Olympus BX41 microscope, 1000X magnification under oil, and Qcapture software. The scale bar equals 25 μm.
(D) FACS analysis was performed with FITC and PE-labeled antibodies for CD11b and CD14, respectively. Six days post-infection with shRNAs targeting SYK, HL-60 cells were positive for single stained CD11b and CD14 and double staining compared to a luciferase control.

(A) Western immunoblot of HL-60 cells treated with vehicle, gefitinib, or R406 for one hour. Lysates were immunoprecipitated with anti-Syk antibody and blotted with antibody to p-Syk (Y525/526). Treatment with gefitinib and R406 dephosphorylates Syk at Y525/526. Total Syk was evaluated for a loading control.
(B) HL-60 and U937 cells were treated for three days with R406 and DMSO and a Weighted Summed Score (Differentiation Score) determined. Error bars denote the mean ±SD of 16 replicates. A dose response was seen in both cell lines. High-dose DMSO is known to induce differentiation in HL-60; however, the Differentiation Score is higher for each corresponding dose of R406.
(C) May Grunwald Giemsa staining of HL-60 and U937 cells treated for three days with DMSO or R406. R406 induces myeloid maturation in both cell lines as characterized by nuclear condensation and cytoplasmic ruffling. Images were acquired with an Olympus BX41 microscope, 1000X magnification under oil, and Qcapture software. The scale bar equals 25 μm.
(D) FACS analysis was performed with FITC- and PE-labeled antibodies for CD11b and CD14, respectively. HL-60 cells were treated for three days with vehicle or R406. R406 induced both CD11b and CD14 stained double positive cells and single stained positive CD11b consistent with myeloid maturation.
(E) HL-60 cells were treated in triplicate for three days with vehicle, ATRA, or R406. The percentage of NBT-positive cells and mean ±SD are shown. Both ATRA and R406 reduced NBT consistent with the functional maturity of HL-60 cells.

(A) Western immunoblot of AML cell lines reveals MOLM-14 and KG-1 cell lines to be highly phosphorylated at the p-Syk (Y525/526) enzymatic site.
(B) MOLM-14 and KG-1 cells were treated with vehicle or R406 for 6 days. Cells were stained with annexin V-FITC and propidium iodine (PI) and evaluated by flow cytometry. R406 treatment induced increased annexin V positive cells in a dose responsive manner consistent with apoptosis.
(C) MOLM-14 and KG-1 cells were infected with shRNAs targeting SYK and luciferase. After 8 days of infection, cell viability was evaluated at days 0, 3, and 6 using an ATP-based assay. Genetic loss of Syk resulted in dramatic decrease in proliferation. Error bars depict mean ±SD across ratios of 6 replicate measurements at each time point relative to the 6 replicate measurements at time zero.
(D) Ectopic expression of TEL-Syk immune to the SYK-directed shRNA Syk_1 in MOLM-14 cells rescues the effects of Syk knockdown on cell viability. Cell viability was evaluated 4 days after infection at days 0, 3, and 6 with an ATP-based assay. Error bars depict mean ±SD across ratios of 8 replicate measurements at each time point relative to the 8 replicates at time zero.
(E) May Grunwald Giemsa staining of KG-1 cells 12 days post infection with shRNAs targeting SYK versus a luciferase shRNA control demonstrate evidence of macrophage-like differentiation. Images were acquired with an Olympus BX41 microscope, 1000X magnification under oil, and Qcapture software. The scale bar equals 25 μm.
(F) KG-1 cells were treated with R406 versus DMSO and the 32-gene differentiation signature measured at 24 hours. R406 induces a myeloid differentiation signature as measured by a Weighted Summed Score (Differentiation Score). Error bars depict mean ±SD across 8 replicates.

(A) The ability of KG-1 to form colonies in methylcellulose with R406 and multiple shRNAs directed against SYK was assessed. Both chemical and genetic inhibition of Syk abrogate colony formation in KG-1. The scale bar equals 1 mm.
(B) KG-1 luciferase positive xenografts were established in NOG mice. Mice were treated with placebo (n=4), food impregnated with 5g R788 / kg AIN-76A food (n=5), and food impregnated with 8g/kg AIN-76A food (n=3) for 8 days. Mice treated with R788 had a significant difference in tumor burden compared to those treated with placebo. Serial in vivo BLI was used to assess disease burden, and data plotted as the mean ±SEM for each group. One way ANOVA analysis with Tukey post-test was used to determine the significance of all pairwise comparisons.
(C) A syngeneic, luciferase positive mouse model of MLL-AF9 AML was established in C57BL/6:Tyr^C/C mice. Mice were treated with placebo (n=10) or food impregnated with 3g R788 / kg AIN-76 food (n=10) for 8 days. Serial in vivo BLI was used to assess disease burden and plotted as mean ±SEM. Statistical significance was calculated using Student’s t-test.

(A) Fresh primary patient AML cells were treated in vitro with vehicle versus R406 and Syk phosphorylation at Y525/526 was assessed by Western immunoblotting. In all four samples Syk is phosphorylated at the kinase activity site, and this phosphorylation is inhibited by R406.
(B) Primary AML blasts were treated with DMSO, ATRA, or R406 for three days and the 32-gene differentiation signature evaluated. Error bars denote mean ±SD across 8 replicates per condition. R406 induces comparable or greater levels of differentiation to ATRA. May Grunwald Giemsa staining of ATRA- and R406-treated primary AML at four days reveals myeloid differentiation with lobulation of nuclei in the APL sample and macrophage-like differentiation in Patients 12 and 13. Images were acquired with an Olympus BX41 microscope, 1000X magnification under oil, and Qcapture software. The scale bar equals 25 μm.
(C) Primary patient AML cells were treated in quadruplicate with R406 in a two-fold dilution series. At day three, cell viability was evaluated using an ATP-based assay. Values for the IC₅₀ are depicted with error bars denoting the 95% confidence interval for the SEM for the four chemical replicates relative to the four control replicates. At day three the 32-gene differentiation signature was measured. Samples are scored as differentiating if there was a 1.5-fold change in the Differentiation Score across at least 2 doses with p < 0.05 by one-tailed t-test assuming two samples with unequal variance.

(A) A primary AML (Patient 11) orthotopic xenograft was established in NOG mice. Six days after initiation of treatment with food containing 5g R788/kg AIN-76 rodent diet, the AML blast percentage, measured by percentage of human specific CD45+ cells in peripheral blood, continued to increase in the placebo-treated mice but decreased in the R788-treated mice.
(B) Histopathology studies of bone marrow in R788-treated versus placebo-treated animals reveal near resolution of AML infiltration with areas of necrosis and areas of recovering marrow identified. Hematoxylin and eosin staining of spleen reveals areas of leukemic infiltration and extramedullary hematopoiesis in placebo-treated mice and near resolution of AML infiltration in the R788-treated animals. Images were acquired with an Olympus BX41 microscope, 1000X magnification under oil, and Qcapture software. The scale bar equals 25 μm.
(C) The ability of normal CD34 myeloid progenitor cells and primary patient AML blasts to form colonies in methylcellulose with R406 was assessed. Colony formation was assessed in duplicate and colonies per 4 cm² counted and displayed relative to control cells. Error bars depict mean ± SD across 4 ratios of dose response to vehicle. Primary patient AML blasts were more sensitive to the effects of R406 than were normal myeloid progenitors.