Inhibition of the TGF-beta receptor I kinase promotes hematopoiesis in MDS

Abstract

MDS is characterized by ineffective hematopoiesis that leads to peripheral cytopenias. Development of effective treatments has been impeded by limited insight into pathogenic pathways governing dysplastic growth of hematopoietic progenitors. We demonstrate that smad2, a downstream mediator of transforming growth factor-beta (TGF-beta) receptor I kinase (TBRI) activation, is constitutively activated in MDS bone marrow (BM) precursors and is overexpressed in gene expression profiles of MDS CD34(+) cells, providing direct evidence of overactivation of TGF-beta pathway in this disease. Suppression of the TGF-beta signaling by lentiviral shRNA-mediated down-regulation of TBRI leads to in vitro enhancement of hematopoiesis in MDS progenitors. Pharmacologic inhibition of TBRI (alk5) kinase by a small molecule inhibitor, SD-208, inhibits smad2 activation in hematopoietic progenitors, suppresses TGF-beta-mediated gene activation in BM stromal cells, and reverses TGF-beta-mediated cell-cycle arrest in BM CD34(+) cells. Furthermore, SD-208 treatment alleviates anemia and stimulates hematopoiesis in vivo in a novel murine model of bone marrow failure generated by constitutive hepatic expression of TGF-beta1. Moreover, in vitro pharmacologic inhibition of TBRI kinase leads to enhancement of hematopoiesis in varied morphologic MDS subtypes. These data directly implicate TGF-beta signaling in the pathobiology of ineffective hematopoiesis and identify TBRI as a potential therapeutic target in low-risk MDS.

Figure 1

smad2 is activated in MDS. Bone marrow (BM) biopsies from patients with MDS and controls with non-MDS causes of cytopenias were fixed and immunostained with antibody against phospho-smad2 (A). Histologic examination reveals more intense staining in MDS samples. Eight representative samples of MDS and controls are shown in panel A. The quantification of p-smad2 staining was analyzed by counting the total number of positively stained cells (B) and by measuring intensity of the positively stained cells (C) in 5 hot fields (which is defined as area of high density of p-p38 staining) and aided by Image Pro Plus (Nikon, 400×). Two-tailed t test shows significantly higher smad2 activation/field in MDS samples. Differences in smad2 expression were also evaluated in normalized meta-analysis of MDS CD34⁺ (69 cases)– and normal CD34⁺ (57 cases)–cell–derived gene expression microarray studies. Smad2 gene expression was significantly up-regulated in low-grade MDS bone marrow CD34⁺ cells (2-tailed t test) (D). Error bars represent SEM.

Figure 2

Down-regulation of TGF beta receptor I (TBRI) can inhibit smad2 activation in hematopoietic cells and stimulate MDS hematopoiesis. GFP-expressing lentiviral-based shRNA against TBRI (A) was used to knock down TBRI in hematopoietic cells. qPCR shows decrease in TBRI mRNA expression after lentiviral shRNA-TBRI infection in bone marrow stromal (HS-5) and leukemia cells (K562) compared with scrambled control (B). K562 cells with stable expression of TBRI-shRNA lentivirus show decreased smad2 phosphorylation after TGF stimulation (C). Primary CD34⁺ progenitors were electroporated with GFP coexpressing anti–TBRI-shRNA construct and sorted after 48 hours. GFP-positive cells were grown in methylcellulose with cytokines, and erythroid colonies were counted after 14 days. TBRI-shRNA–transfected progenitors were less inhibited by TGF-β compared with cells transfected with scrambled control shRNA (31% colonies/control vs 16% colonies/control). Expressed as means (± SEM) of 4 independent experiments (P = .03, 2-tailed t test) (D). CD34⁺ cells transfected with anti–TBRI-shRNA also formed bigger colonies in the presence of TGF-β1 compared with controls (E; Nikon, 40×). Primary bone marrow–derived mononuclear cells from 5 patients with MDS were transfected with shRNA targeting TBRI and control, and equal numbers of cells (for each individual patient) were grown in methylcellulose with cytokines. Erythroid (BFU-E) and myeloid (CFU-GM) colonies were counted after 14 days of culture and demonstrated an increase after anti–TBRI-shRNA transfection (significance between means calculated by 2-tailed t test) (F).

Figure 3

SD-208 is an inhibitor of TGF-β signaling in hematopoietic cells. Leukemic cells (K562 and KG-1) and primary hematopoietic progenitors at the colony-forming unit–erythroid stage of maturation (CFU-E) were treated with TGF-β1 (20 ng/mL) in the presence and absence of SD-208 (.5 μM) and assessed for smad2 phosphorylation by immunoblotting. SD-208 pretreament (1 hour) led to attenuation of activation/phosphorylation of smad2 (A-C). Bone marrow stroma–derived cells (HS-5) were transfected with plasmids expressing smad binding 3TP-luciferase and β-galactisidose (transfection control) and stimulated with TGF-β1 in the presence and absence of SD-208 (dose .5 μM). TGF-β1–induced control-normalized luciferase activity was potently inhibited by SD-208. A kinase-null mutant of TGF-β receptor I (TBRI-KR) was used a positive control (D). Error bars represent SEM.

Figure 4

TBRI inhibition can inhibit TGF-β–mediated cell-cycle arrest of CD34⁺ cells. Equal numbers of BM CD34⁺ cells were grown in the presence of SCF, TPO, and FLT-3L and were pretreated with DMSO or 0.5 μM SD-208 for 1 hour before TGF-β1 (0.5 ng/mL final concentration) was added. On days 3, 5, and 7, cell aliquots were taken and viable cell concentration was determined using Guava ViaCount. The experiment was repeated at least 3 times (using multiple donors of CD34⁺ cells) and means plus or minus SEM is shown (A). CD34⁺ cells were treated with DMSO or SD-208 in the presence or absence of TGF-β1 for 7 days. Cell-cycle distribution of CD34⁺ cells (gated with PE-conjugated CD34 antibody) was determined on day 7 using the APC BrdU Flow Kit and the LSR-II flow cytometer (BD Biosciences; B; representative sample, C). Error bars represent SEM. CD34⁺ cells were treated with DMSO or SD-208 in the presence or absence of TGF-β1 as described above. After 24 hours, cDNA was prepared and hybridized on a cDNA microarray. Selected cell-cycle progression genes that were down-regulated by TGF-β by 2-fold were validated at the protein level by Western blotting. CD34⁺ cells were treated with DMSO or SD-208 in the presence or absence of TGF-β1 as described above. Cells were collected at the indicated time points and lysed in radioimmunoprecipitation assay (RIPA) buffer. Equal protein was separated on a 10% Bis-Tris sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) gel and transferred to nitrocellulose membrane and immunoblotted with the antibodies (D). GAPDH levels were used as protein loading controls, and p-smad2 was used as positive control for TGF-β1 stimulation.

Figure 5

SD-208 can improve anemia in a murine model of TGF-β1–driven bone marrow failure. Mice transgenic for alb/TGF-β were killed at 8 weeks of age and their bone marrows were stained for histology (hematoxylin and eosin and reticulin stain). Transgenic mice demonstrated dysplastic micromegakaryocytes and patchy fibrosis. (A) Blood counts were analyzed at 3 weeks by Advia (Bayer) counter, and alb/TGF⁺ transgenic mice were found to be significantly anemic compared with WT controls (n = 4; means ± SEM; Nikon, 40×) (B). alb/TGF⁺ mice were treated with either SD-208 (30 mg/kg per day) or vehicle (placebo, daily) by gastric lavage for 14 days. Blood counts were done on the 14th day and revealed a significant rise in hematocrit after SD-208 treatment (C). The mice were killed and bone marrow cells were plated in methylcellulose with Epo (for BFU-E colonies) and IL-3, IL-6, and SCF (for CFU-GM colonies). SD-208 treatment led to a significant increase in both erythroid and myeloid colony-forming potential compared with placebo (n = 8; means ± SEM; 2-tailed t test; D).

Figure 6

TBRI inhibition stimulates hematopoiesis in MDS. MDS bone marrow–derived MNCs from 15 patients were plated in methylcellulose and cytokines in the presence and absence of TBRI inhibitor SD-208 (100 nM and 200 nM). Colonies were scored at day 14 and results were expressed as means (± SEM) of 15 independent experiments.