SHIP1 Inhibition Increases Immunoregulatory Capacity and Triggers Apoptosis of Hematopoietic Cancer Cells

Using various genetic models, we have shown that SHIP1-deficient hosts are permissive for engraftment of MHC mismatched bone marrow grafts, exhibit reduced graft-versus-host disease (GVHD) posttransplant and delayed rejection of vascularized allogeneic heart grafts (1–5). In addition, SHIP1 deficiency profoundly increases myeloid immunoregulatory (MIR) cell numbers and their function and granulocyte numbers (2, 4, 6–8). These studies suggest that SHIP1 could be targeted to reduce the severity and incidence of deleterious allogeneic T cell responses in bone marrow and organ transplantation (5). SHIP also opposes PI3K signaling at key hematopoietic growth factor receptors such as c-Kit, c-Mpl, Epo-R, and GCSF-R (9), and thus SHIP plays a prominent role in limiting hematopoietic progenitor activity and blood cell production (6, 10).

SHIP1, SHIP2, and PTEN are commonly viewed as opposing the activity of the PI3K/Akt signaling axis that promotes survival of cancer cells and tumors. However, the enzymatic activities of these inositol phosphatases are quite distinct in that the 3′ poly-phosphatase activity of PTEN reverses the PI3K reaction to generate PI(4,5)P2 from PI(3,4,5)P3, whereas the 5′ polyphosphatase activity of SHIP1/2 converts PI(3,4,5)P3 to PI(3,4)P2. This distinction is potentially crucial as it might enable SHIP1/2 and PTEN to have distinctly different effects on Akt signaling. In fact, the pleckstrin homology domain of Akt binds with greater affinity to the SHIP1/2 product PI(3,4)P2 leading to more potent activation of Akt than the direct product of PI3K, PI(3,4,5)P3 (11). Thus, SHIP1, which is expressed in most blood cell malignancies, may contribute to growth and survival of neoplastic cells. Consistent with this hypothesis, PI(3,4)P2 levels are increased in leukemia cells (12) and increased levels of PI(3,4)P2 promote the transformation and tumorigenicity of mouse embryonic fibroblasts (13).

We sought then to identify compounds capable of inhibiting SHIP1 that also have activity in vivo. To date, the molecular structure of SHIP1 has not been determined and thus a rational design approach to develop SHIP1 inhibitors is not feasible. Thus, we pursued a high-throughput screening strategy to identify compounds that can inhibit the enzymatic activity of SHIP1. This screen yielded a SHIP1 selective inhibitor that is capable of increasing MIR cell numbers and function in vivo, enhancing blood cell production in myelosuppressed hosts and promotes apoptosis of blood cancer cells.

C57BL6/J and BALB/C mice were purchased from The Jackson Laboratory (Bar Harbor, ME). The development and ablation of SHIP expression in adult mice using the MxCreSHIP^flox/flox model has been described previously (4, 14). Briefly, MxCreSHIP^flox/flox mice were conditionally deleted for SHIP through the i.p. injection of polyinosinic-polycytidylic acid (poly:I/C) (Sigma-Aldrich, St. Louis, MO). Mice were injected three times with 625 μg poly:I/C on days 1, 4, and 7. SHIP^flox/flox controls were treated in a similar fashion.

A SHIP1 cDNA expression construct was amplified by PCR from the pMIGR1 (SHIP1) vector (15) and then inserted into the pET24b bacterial expression vector at the EcoR1 and Xho1 restriction sites in this vector to create a SHIP1-His tag fusion at the C terminus. His-tagged SHIP1 was then expressed in Escherichia coli Rosetta-Gami 2 (DE3) pLys cells by induction after the addition of 0.5 mM IPTG and purified by Ni-chelating bead chromatography.

A fluorescent polarization (FP) assay that detects the 5′-inositol phosphatase activity of SHIP was developed based on a previously established protocol (16) and was obtained from Echelon Biosciences (Salt Lake City, UT). This assay is based on a competition between PIPs. Mean polarization unit (mPu) values decrease as probe binding to the PI(3,4)P₂ detector is displaced by PI(3,4)P₂ produced by the enzymatic reaction. The assay was performed according to the manufacturer’s protocol, with the exception that the concentration of MgCl₂ was increased to 10 mM to optimize the activity of SHIP1 in this assay. Each enzymatic reaction was performed in a 96-well working plate and added in the following order and concentration: Recombinant SHIP (1 μg/ul) in PBS and 20 mM MgCl₂, 100 μM of each individual compound from the NCI Diversity Set and then PI(3,4,5)P₃ substrate diluted 1:10 in H₂O to a final concentration of 2 μM. The enzymatic reaction was incubated for 30 min at room temperature. The detector, reaction mixture, and probe were then combined according to the manufacturer’s protocol. Malachite Green Assay was also obtained from Echelon Biosciences. Malachite Green forms a colored complex with free phosphate liberated by the SHIP1, SHIP2, or PTEN enzyme reactions. The assay was performed according to manufacturer’s protocol. The 3α-aminocholestane (3AC) was dissolved in 100% ethanol and used at a final concentration of 1 mM in the Malachite Green assay. SHIP2 and PTEN were obtained from Echelon Biosciences. SHIP2 and PTEN were both used at the recommended concentration as determined by the manufacturer.

For in vitro studies, 3AC was suspended in 100% ethanol. For in vivo studies, 3AC was suspended in a 0.3% Klucel/H₂O solution at 11.46 mM. Vehicle-treated mice received a 200 μl injection 0.3% Klucel/H₂O solution. The final concentration of 3AC in the treated mice was 60 μM. For in vivo studies, both vehicle- and compound-treated mice received daily i.p. injections for 7 d prior to analysis or tissue harvest. For analysis of blood cell recovery after radioablation adult C57BL/6J mice (The Jackson Laboratory) received 550 Rads from an X-ray source (RS2000, Rad Source, Suwanee, GA). The irradiated mice then received i.p. injections of 3AC (60 μM) for 7 d. Blood was drawn on 8, 15, and 29 d after irradiation for analysis using a Hemavet 950S automated blood cell analyzer (Drew Scientific, Waterbury, CT). Statistical analyses were performed using Prism (GraphPad, San Diego, CA).

Anti-CD16/32 was incubated with the samples to block Fc receptor binding. Abs used for staining included Gr-1 (RB6-8C5) and Mac-1(M1/70) and were obtained from BD Pharmingen (San Jose, CA). All samples were acquired on a FACSCalibur and analyzed using FlowJo8.7.1. For viable staining, dead cells were excluded from the analysis after cytometer acquisition of data by exclusion of the 7AAD dye.

After RBC lysis, irradiated (2000 Rads) BL6 splenocytes (stimulators) at 8 × 10⁵ cells/well were combined with BALB/c splenocytes (responders) at 4 × 10⁵ cells/well in a one-way MLR assay. All cells were plated in sextuplicate in a 96-well U-bottom plate (Falcon, Becton Dickinson, Franklin Lakes, NJ) containing RPMI 1640 complete medium for 4 d. Cells were then pulsed with 1 μCi [³H]thymidine/well for 18 h. Cells were lysed and high m.w. DNA captured on glass fiber filtermates using an automated cell harvester (Packard Instruments, Meriden, CT). Incorporated [³H]thymidine was quantitated using a Packard TopCount NXT. Specific [³H]thymidine incorporation into genomic DNA was calculated as the average of the mean (± SEM) of sextuplicate wells. For the human MLR studies, human PBMCs to be used as stimulators were first cultured at 4 × 10⁶ cells/ml for 24 h in human IL-15 medium (RPMI medium, 10% FBS, 20 mM HEPES, penicillin/streptomycin, l-glutamine, nonessential amino acids, 20 mM 2ME supplemented with human IL-15 [10 ng/ml]) containing either 9.4 μM 3AC or vehicle. The 3AC- or vehicle-treated stimulator PBMCs (8 × 10⁵ cells) were then mixed with responder PBMCs (4 × 10⁵ cells) from a different donor in 200 μl human IL15 medium. Four days later 1 μCi [³H]thymidine was added to the wells and 18 h later cells were harvested, lysed, and high m.w. DNA captured on glass filtermates. Specific [³H]thymidine incorporation into genomic DNA was calculated as the average of the mean cpm (± SEM) of sextuplicate wells.

KG-1 human myelomonocytic cells and K-562 human leukemia cells were grown in Iscove’s Modified DMEM (Invitrogen, Carlsbad, CA) supplemented with 20% and 10% heat inactivated FCS (Cellgro; Mediatech, Manassas, VA) and 1% of penicillin and streptomycin. KG-1 and K-562 cells were obtained from the American Type Culture Collection (Manassas, VA). C1498 was originally derived from a female C57BL/6J mouse (17) and was obtained from the American Type Culture Collection. C1498 cells were grown in DMEM (Invitrogen) supplemented with 10% FCS (Cellgro; Mediatech), 1% of penicillin and streptomycin. MM cell lines and MG63 cells were a kind gift from Dr. R. van Bezooyen (LUMC, Leiden, The Netherlands), and were propagated in DMEM containing 100 U/ml penicillin, 100 μg/ml streptomycin, and 10% FCS. The PIP species [PI(3,4)P₂-diC16, P-3416, and PI(3,5)P2 diC16, P-3516] were purchased from Echelon Biosciences (www.echelon-inc.com). The PIPs were delivered into cells using the Neomycin Sulfate, P-9C1, PIP carrier according to the manufacturer’s recommendation (18). The shuttle PIP carrier was dissolved with the phosphatidylinositol at a ratio of 1:1 to form the PIP-carrier complex at final concentration of 5, 10, and 15 μm and incubated with 1 × 10⁶ C1498 cells for 1 h before treatment with 3AC (12.5 μM) for 36 h. MTT cell proliferation assay was conducted according to manufacturer's recommendation. The samples were read using an ELISA plate reader at a wavelength of 570 nm. Statistical analyses were performed using Prism (GraphPad).

Cells were counted, centrifuged, and resuspended in Laemmli sample buffer. SDS-PAGE and immunoblotting were performed according to standard procedures. Detection was performed according to the manufacturer’s guidelines (ECL, Pierce, Rockford, IL). Membranes were probed with Abs against Phospho-Akt Ser473, cleaved PARP, cleaved Caspase 3 (Cell Signaling Technology, Beverly, MA), or SHIP1 (Santa Cruz Biotechnology, Santa Cruz, CA). To confirm equal loading, membranes were reprobed with an Ab against β-actin (Santa Cruz Biotechnology).

Cells were treated in duplicate for 36 h with increasing concentrations of 3AC or vehicle. MTT (Sigma-Aldrich) was added at a concentration of 0.5 mg/ml to the cells for 3 h. Formed crystals were dissolved in dimethyl sulfoxide and OD was measured at 570 nm. The OD of compound treated cells was divided by the OD of their vehicle control, and the viability was expressed as a percentage of untreated cells. Results are expressed as mean ± SEM of three individual experiments.

Annexin V/propidium iodide (PI) staining was performed using the Annexin V-FITC apoptosis detection kit from BD Pharmingen (Sharon, MA) per the manufacturer’s instructions. In short, cells were treated with increasing concentrations of 3AC for 36 h. Cells were harvested, washed twice with ice-cold PBS, and stained with Annexin-FITC and PI in binding buffer for 15 min. Fluorescence was determined by flow cytometry (LSR II, Becton Dickinson Medical Systems, Sharon, MA).

Others have previously described a FP assay to detect and quantify SHIP2 activity (15). We adapted this assay to detect the activity of recombinant SHIP1 (Fig. 1A, 1B) and used the FP assay as a rapid screen to detect compounds in chemical libraries that have inhibitory activity against SHIP1. This screen identified 3AC as a potent inhibitor of SHIP1 enzymatic activity (Fig. 1C). Using the FP assay, we find that 3AC exhibits detectable inhibitory activity against 0.1 μg of recombinant SHIP1 at 2 μM and 50% inhibition at 10 μM. To determine whether 3AC exhibits selectivity for SHIP1, we assessed its capacity to inhibit the other major inositol phosphatases in the mammalian cell, SHIP2 and PTEN, both of which also hydrolyze the PI3K product, PI(3,4,5)P₃. We find that 3AC is selective for SHIP1 as it fails to inhibit SHIP2 and PTEN at 1 mM (Fig. 1D). This high degree of selectivity could be important, because loss of PTEN function promotes malignancy in parenchymal and hematopoietic tissues (19–21).

We and others find that expansion of the Mac1⁺Gr1⁺ MIR cell compartment can suppress GVHD (2, 4, 22). Thus, we tested the ability 3AC to expand the MIR compartment in peripheral lymphoid tissues where GVHD and organ transplant rejection are primed. Adult mice were treated for 7 d with a daily i.p. injection of 3AC at 60 μM. We then compared the frequency of MIR cells in the spleens and lymph nodes (LNs) of mice treated with 3AC with that of control mice treated with either vehicle or to unmanipulated mice. We find that the frequency and number of MIR cells is increased significantly in both spleen (Fig. 2A) and LNs (Fig. 2B) following treatment with 3AC. We observed no significant increase in MIR cells with vehicle treatment relative to unmanipulated controls. Thus, consistent with our findings in SHIP^−/− mice (2) and mice genetically ablated for SHIP1 expression during adult physiology (4), inhibition of SHIP1 activity in adult mice expands the MIR cell compartment in peripheral lymphoid tissues.

Increased numbers of MIR cells in peripheral lymphoid tissues of germline SHIP1-deficient mice, and in mice where SHIP1 expression is genetically ablated during adulthood, impairs priming of allogeneic T cell responses (2, 4). The peripheral expansion of the MIR cell compartment promoted by 3AC should then promote an immunosuppressive environment and particularly for priming of allogeneic T cell responses (2). To assess this possibility, we tested splenocytes from H2b mice treated with 3AC for their ability to prime allogeneic T cell responses by completely mismatched H2d responder splenocytes. Their responses were compared with splenocytes from either unmanipulated mice or splenocytes from mice treated with vehicle. As anticipated, splenocytes from mice treated with 3AC are profoundly impaired in their ability to stimulate allogeneic T cell responses (Fig. 2C). Thus, chemical inhibition of SHIP1 enzymatic activity in vivo expands potent immunoregulatory cell populations in peripheral lymphoid tissues and consistent with this expansion these tissues are profoundly compromised for priming of allogeneic T cell responses. We also examined the ability of 3AC to inhibit priming of a human allogeneic T cell response in vitro. We find that pretreatment of human PBMCs significantly reduces their ability to prime allogeneic responses by T cells from an unrelated donor (Fig. 2D). Thus, SHIP1 inhibition significantly reduces priming of human or rodent allogeneic T cell responses.

We have recently shown that SHIP1-deficient mice exhibit increased G-CSF production (14). Consistent with this finding, SHIP^−/− mice are known to have significantly increased granulocyte numbers (6). Thus, we hypothesized that chemical inhibition of SHIP1 might lead to increased production granulocytes, much like recombinant G-CSF administration achieves in cancer patients. To test this, we measured granulocyte levels after a week of daily treatment with 3AC. Treatment with 3AC led to a 4- to 5-fold increase in circulating granulocytes in peripheral blood relative to either vehicle or unmanipulated controls (Fig. 3A). Wright-Giemsa staining of cytosmears prepared from the blood of the treated mice confirmed the granulocyte increase had a neutrophil morphology (data not shown).

SHIP opposes PI3K signaling at several hematopoietic growth factor receptors (9). We considered then that SHIP1 inhibition might be capable of rescuing, or enhancing, recovery of key blood cell components in myelosuppressed hosts. To test this possibility, we sublethally irradiated mice and the following day initiated daily 3AC treatment for 7 d. We then analyzed RBC, neutrophil, and platelet levels at 8, 15, and 29 d postirradiation in both 3AC- and vehicle-treated cohorts. This analysis showed that 3AC treatment promotes a significant increase in RBC numbers in myelosuppressed hosts 8 d after irradiation and as a consequence RBC levels in SHIP1 inhibitor-treated mice do not drop below normal 15 d postirradiation as they did in the vehicle group, as expected for an irradiated host (Fig. 3B). Significant platelet and neutrophil nadirs were observed in both 3AC- and vehicle-treated groups 8 d after irradiation; however, the SHIP1 inhibitor-treated hosts showed faster and more robust recovery of neutrophils (Fig. 3C) and platelets (Fig. 3D). Neutrophil counts were significantly higher in the 3AC-treated hosts at both 15 and 29 d postirradiation (Fig. 3C). Platelet counts returned to normal in the 3AC treated group and were significantly higher 15 d postirradiation, whereas platelet levels were below or at the lower limit of the normal range in the vehicle group at both 15 and 29 d postirradiation (Fig. 3D). Thus, SHIP inhibition is protective for the RBC compartment in radioablated mice, whereas promoting a more rapid and effective recovery of neutrophils and platelets in such hosts.

Helgason et al. first reported that germline SHIP mutant mice develop a severe lung pathology related to a myeloid consolidation of the alveolar spaces (6). We also observe this pathology in our germline SHIP^−/− strain. Moreover, we find that this pathology is routinely observed in MxCreSHIP^flox/flox mice (4) 5 wk after induced genomic ablation of SHIP expression (Fig. 4A, 4B), but not in poly:I/C-treated SHIP^flox/flox controls (Fig. 4C). Intriguingly, SHIP1 inhibition by 3AC treatment for a 7-d period does not cause the pulmonary myeloid infiltration and lung consolidation that is uniformly lethal for adult SHIP1^−/− mice (Fig. 4D–F) (2, 6). Thus, transient, chemical inhibition of SHIP1 does not cause the lung consolidation and pneumonia observed in genomically ablated mice. In addition, we observe no significant morbidity or mortality in mice treated with 3AC for a 1-wk period. Consistent with the general health of 3AC-treated mice, there is no significant weight difference in 3AC-treated mice versus vehicle controls (Fig. 4G). Nor do we detect any evidence of pathology in heart tissue of 3AC-treated mice (Fig. 4H). Thus, transient chemical SHIP1 inhibition can promote some beneficial hematologic and immunoregulatory effects without major toxicity.

We were concerned that SHIP1 inhibitors might potentially enhance the survival and growth of blood cell cancers. However, as discussed previously, SHIP1 inhibition should decrease PI(3,4)P₂ levels and thus in some contexts could decrease activation of downstream effectors of cell survival and proliferation in malignant blood cells. We tested these two possibilities by treating the AML cell line (KG-1) that expresses SHIP1 (Fig. 3A) with 3AC. We find that 3AC decreases the growth and survival of KG-1 cells in vitro, but not osteosarcoma cells that lack SHIP1 expression (Fig. 5A, 5B). Murine C1498 leukemia cells also show comparable sensitivity to 3AC; however, K562 leukemia cells that express both PTEN and SHIP2, but lack expression of SHIP1, are insensitive to 3AC treatment (Fig. 5C). We find that 3AC promotes apoptosis of SHIP-expressing AML cells based on increased cleavage of PARP and Caspase 3, decreased activation of Akt/PKB and increased frequency of AnnexinV⁺ cells (Fig. 5D, 5E). We also find that SHIP1 inhibition reduces both tonic activation of Akt/PKB in human multiple myeloma (MM) cells as well as IGF1-induced activation of this kinase (Fig. 5F). Consistent with this effect on Akt/PKB signaling in MM cells, 3AC is cytotoxic for these cells (Fig. 5G). To provide further confirmation that 3AC is acting on SHIP1 in blood cancer cells, we examined the importance of its product, PI(3,4)P₂, in cancer cell growth and signaling. We tested whether exogenous PI(3,4)P₂ is protective when introduced into SHIP1 inhibitor-treated C1498 leukemia cells. Consistent with 3AC acting as an inhibitor of SHIP1, introduction of PI(3,4)P₂ into leukemic cells is significantly protective in a dose-dependent fashion, whereas introduction of the non-SHIP1 product, PI(3,5)P₃, at equivalent doses did not provide significant protection from 3AC (Fig. 5H). Taken together these results demonstrate 3AC reduces the growth and survival of hematopoietic cancer cells that express SHIP1 by limiting intracellular levels of PI(3,4)P₂.

In this study, we describe the identification and initial in vivo characterization of a small molecule inhibitor of the SHIP1 enzyme. To validate that this compound identified in a solution-based assay for SHIP1 activity is cell permeant and can alter the immune system in a manner comparable to that observed in SHIP1-deficient mice, we tested its ability to expand MIR cells and to consequently inhibit priming of an allogeneic T cell response. We find that chemical inhibition of SHIP1 is capable of both. In addition, chemical inhibition of SHIP1 promotes a profound increase in circulating granulocyte numbers, can facilitate blood cell recovery in myelosuppressed hosts and compromises survival of blood cancer cells.

We show here that administration of a SHIP1 inhibitor can expand immunoregulatory cells in peripheral lymphoid tissues and suppress priming of allogeneic T cell responses. Because allogeneic T cell responses that culminate in GVHD or solid organ graft rejection are primed in peripheral lymphoid tissues (23–25), this study suggests that 3AC, and potentially other SHIP1 selective inhibitors, might be used to limit deleterious T cell responses that mediate GVHD and organ graft rejection. Consistent with this, GVHD is reduced and cardiac graft rejection delayed in adult mice rendered SHIP1-deficient (4, 26). As SHIP1-deficient mice exhibit normal humoral immunity (27, 28) and priming of T cell responses to naive Ags (2), the SHIP1 inhibitor described in this study, and potentially others, may not significantly compromise adaptive immune function. Thus, 3AC may offer a more selective method to dampen deleterious host and donor allogeneic T cell responses without compromising adaptive immune functions necessary to combat opportunistic pathogens that can compromise the recovery and survival of transplant patients receiving state-of-the-art immunosuppressive regimens.

We and others have documented increased Akt signaling and survival in primary NK(1) and myeloid cells (29) isolated from SHIP1^−/− mice. However, there is also an emerging role for the SHIP1/2 product PI(3,4)P₂ in promoting Akt activation (11) and tumorigenicity (13). Thus, via generation of PI(3,4)P₂, SHIP1/2 could in some contexts amplify survival or proliferative signals in neoplastic cells by providing additional docking sites at the plasma membrane for recruitment and activation of PH-domain containing kinases such as Akt. Indeed, PI(3,4)P₂ levels are found to be increased in leukemia cells (12). Consistent with this hypothesis, we find that a SHIP1 selective inhibitor reduces Akt activation and promotes apoptosis of human blood cell cancers that express SHIP1. We further confirmed a role for PI(3,4)P₂ in cancer signaling by showing that introduction of exogenous PI(3,4)P₂ into leukemia cells significantly protects them from SHIP1 inhibition. Thus, SHIP1 inhibition could be considered as an adjunct to other therapeutics to further decrease the survival of hematologic malignancies. There may also be applications for SHIP1/2 inhibitors in nonhematologic cancers as SHIP2 expression is increased in breast cancer and promotes survival signals from epidermal growth factor receptor in these cells (30–32).

Although treatment of mice with a SHIP1 selective inhibitor induced many of the same myeloid phenotypes observed in mice that are germline SHIP1-deficient, we were surprised not to observe myeloid lung consolidation and pneumonia emerge in mice treated with a SHIP inhibitor for 1 wk, whereas this pathology is readily manifest in MxCreSHIP^flox/flox mice rendered SHIP-deficient for a 5-wk period. The lack of lung pathology in the SHIP inhibitor treated hosts could be a consequence of the transient and reversible nature of chemical inhibition, as opposed to the genetic ablation of SHIP that is both sustained and permanent. Alternatively, the difference between chemical and genetic ablation of SHIP1 function could be due to the absence of SHIP1 protein in genomically ablated mice. The absence of SHIP1 protein in the mice may lead to inappropriate activities by other signaling proteins that assume SHIP1’s place in signaling complexes. Such a situation occurs in NK cells from SHIP1^−/− mice. Loss of SHIP1 expression in these cells leads to inappropriate recruitment of the tyrosine phosphatase SHP1 to the 2B4 SLAM family receptor converting this receptor from activating mode to dominant inhibitory mode (33). It is possible then that the myeloid lung consolidation observed in SHIP1^−/− mice also results from inappropriate activity by a signaling protein that fills the void left by the absence of SHIP1 protein. Further analysis of these questions could provide mechanistic insights into the role that SHIP1 plays in alveolar macrophage biology.

Our findings also suggest that SHIP1 inhibitors could potentially offer benefits for myelosuppressed patients. Currently, there are lineage-restricted hematopoietic growth factors, G-CSF and Epo, that are used to effectively treat neutropenic and anemic patients, respectively (34, 35). However, the health of myelosuppressed patients is also frequently threatened by inadequate platelet levels. Unfortunately, a recombinant growth factor, thrombopoietin, that boosts platelet production, has met with unanticipated immune responses in patients that have hampered its clinical development (36). In addition, the substantial cost of combining such lineage-restricted growth factors to boost production of multiple blood cell components poses a significant cost for both patients and modern health care delivery systems, much less for patients in the Third World countries. Our finding that SHIP1 inhibitor treatment can prevent a postradiation RBC nadir and promote faster recovery of neutrophils and platelets after myeloablation suggests the potential for a more cost-effective means to promote and preserve blood cell production in myelodysplastic syndrome, patients recovering from marrow ablative therapies or myelosuppressive infections. Thus, SHIP1 inhibitors represent a novel class of compounds that could find utility in boosting blood cell production, facilitating allogeneic transplantation and treatment of oncologic diseases.

We thank Leina Ibrahim for performing the SHIP ablation studies in MxCreSHIP^flox/flox mice and for harvesting lung and heart tissue. We also thank Elizabeth Gengo and Najwa Khan for performing genotyping of mice for this study.

Disclosures The authors have no financial conflicts of interest.