KIT receptor gain-of-function in hematopoiesis enhances stem cell self-renewal and promotes progenitor cell expansion

Abstract

The KIT receptor tyrosine kinase has important roles in hematopoiesis. We have recently produced a mouse model for imatinib resistant gastrointestinal stromal tumor (GIST) carrying the Kit^V558Δ and Kit^T669I (human KIT^T670I) mutations found in imatinib-resistant GIST. The Kit^{V558Δ;T669I/+} mice developed microcytic erythrocytosis with an increase in erythroid progenitor numbers, a phenotype previously seen only in mouse models of polycythemia vera (PV) with alterations in Epo or Jak2. Significantly, the increased hematocrit observed in Kit^{V558Δ;T669I/+} mice normalized upon splenectomy. In accordance with increased erythroid progenitors, myeloerythroid progenitor numbers were also elevated in the Kit^{V558Δ;T669I/+} mice. Hematopoietic stem cell (HSC) numbers in the bone marrow (BM) of Kit^{V558Δ;T669I/+} mice were unchanged in comparison to wild-type mice. However, increased HSC numbers were observed in fetal livers and the spleen and peripheral blood of adult Kit^{V558Δ;T669I/+} mice. Importantly, HSC from Kit^{V558Δ;T669I/+} BM had a competitive advantage over wild-type HSC. In response to 5-fluorouracil treatment elevated numbers of dividing Lin^-Sca⁺ cells were found in the Kit^{V558Δ;T669I/+} BM compared to wild-type. Our study demonstrates that signaling from the Kit^{V558Δ;T669I/+} receptor has important consequences in hematopoiesis enhancing HSC self-renewal and resulting in increased erythropoiesis.

INTRODUCTION

The HSC compartment in the bone marrow (BM) maintains steady state hematopoietic numbers by self-renewal and differentiation into progenitors. HSC, based on their self-renewing abilities, have been distinguished into long-term hematopoietic stem cells (LT-HSC), short-term hematopoietic stem cells (ST-HSC) and finally multipotent progenitor cells (Morrison and Weissman, 1994). The receptor tyrosine kinase KIT and its cognate ligand KITL have long been known to be critical components of the bone marrow niche. LT-HSC and ST-HSC from Kit loss-of-function mutations such as W^v, W⁴¹, W⁴² in either homozygous or heterozygous conditions are ineffective at reconstituting bone marrow in irradiated recipients indicating an important role for KIT in stem cell self-renewal and survival (Sharma et al., 2007).

Apart from its role in stem cell function, the KIT-KITL pair is also known to have a role in regulating adult definitive erythropoiesis. The mutations in the Kit (W) and KitL (Sl) genes have provided important insights into the role of KIT in erythropoiesis. Mice with partial loss-of-function mutations in either gene, are viable but have macrocytic anemia, while more severe mutations lead to embryonic lethality due to lack of erythropoiesis (Besmer, 1991; Nocka et al., 1989; Russell, 1979). KIT signaling is especially important in the proliferation and survival of erythroid progenitors, specifically at the BFU-E stage (Muta et al., 1995; Nocka et al., 1990).

While much is known about hematopoietic phenotypes of Kit deficient mutant mice, relatively little information is known about the consequences of Kit gain-of-function mutations (Besmer, 1997). Somatic activation loop Kit mutations such as Kit^D816V are observed in human mastocytosis, AML and seminoma. BAC transgenic mice carrying the Kit^D816V mutation, die perinatally as a result of strong hematopoietic phenotypes (Gerbaulet et al., 2011). In contrast human patients with somatic or familial GIST most often carry juxtamembrane domain mutations and as in Kit^V558Δ/+ mice only weak hematopoietic phenotypes are observed (Sommer et al., 2003). Recently, we engineered mice bearing in addition to the Kit^V558Δ mutation a “gatekeeper” Kit^T669I second site mutation (human T670I), Kit^{V558Δ;T669I/+}, as observed in imatinib resistant human GIST (Bosbach et al., 2012). The Kit^{V558Δ;T669I/+} mice developed small cecal GIST lesions with pronounced interstitial cell of Cajal hyperplasia in the stomach and colon. Interestingly, these mice showed increased hematocrit and BFU-E numbers with features reminiscent of mice with polycythemia vera (PV) and other myeloproliferative neoplasms (Bosbach et al., 2012).

In the present study we have investigated the consequences of the Kit^V558Δ;T669I gain-of-function mutation on hematopoiesis in mutant mice. We have assessed erythroid progenitor numbers by flow cytometry, analyzed whether splenectomy could normalize hematocrit numbers in Kit^{V558Δ;T669I/+} mice and then investigated whether the mutation resulted in the expansion of myeloid progenitors and stem cells. We also investigated if stem cell function is altered in the Kit^{V558Δ;T669I/+} mice by transplantation assays and analysis of stem cell proliferation. Our data shows that the Kit^V558Δ;T669I mutation results in expansion of erythroid and myeloid progenitors and in particular enhances HSC self-renewal.

MATERIALS AND METHODS

Experimental animals

The Kit^{V558Δ;T669I/+} mice on the C57BL/6J background have been described previously and genotyped according to the published protocol (Bosbach et al., 2012). Congenic B6.SJL-Ptprc^aPep3^b/BoyJ (CD45.1) mice were purchased from the Jackson Laboratory. Peripheral blood was collected from the retro-orbital cavity or the tail vein using EDTA-treated glass capillaries and blood parameters were analyzed using Hemavet 950 (Drew Scientific). All animal procedures were approved by the Institutional Animal Care and Use Committee of Memorial Sloan-Kettering Cancer Center.

Progenitor Cell Analysis

Colony forming assay: Unfractionated bone marrow (BM) (50,000/plate) or spleen (100,000/plate) cells from at least three mice were plated in 1 ml IMDM medium containing 1.2% (wt/vol) methylcellulose, 30% fetal calf serum, 2 mmol/l glutamine, 0.1 mmol/l 2-mercaptoethanol, and 4 mmol/l hemin with cytokines (EPO 6U/ml, IL3 100 ng/ml and KitL 100 ng/ml). Cultures were maintained in triplicate at 37°C in humidified 5% CO2. Granulocyte-monocyte (GM) and granulocyte-erythrocyte-monocyte-megakaryocyte (GEMM) colonies were scored after 7-10 days of culture. Colonies were scored using Nikon Eclipse TE 200 and images were taken using Nikon Eclipse Ti-S.

CFU-S: C57BL/6 mice used as recipients were irradiated (10 Gy: 5 Gy and 5 Gy, 3hr apart) on the day of transplantation and injected with 100,000 wild-type or Kit^{V558Δ;T669I/+} BM cells. Mice were euthanized at the end of 11 days or when they appeared very morbid and spleens were immersed in Bouin's solution for analysis.

Flow Cytometry

Single cell suspensions from BM or spleen were prepared by crushing the tissues in a mortar and pestle in phosphate buffered saline (PBS) containing 2% fetal bovine serum (FBS) and straining the cell mix through a 40 μm cell strainer. Nucleated cells were counted in methylene blue with 1% acetic acid (Gibco) using a hemacytometer. Peripheral blood was collected by retro-orbital or tail vein bleeding and subjected to osmotic lysis using ammonium chloride buffer (Stem Cell Technologies). Wild-type and Kit^{V558Δ;T669I/+} embryos were obtained by crossing Kit^{V558Δ;T669I/+} males with C57BL/6 females. Fetal livers (FL) were excised from the embryos at dpc 14.5. A single-cell suspension of FL cells was obtained by gently mashing the FL on a 40 μm cell strainer in PBS containing 2% FBS. BM, spleen and FL cells were stained in PBS buffer containing 2% FBS and the complete list of antibodies used is provided in supplementary information. DAPI (Invitrogen) or 7AAD (BD Biosciences) was used to electronically gate live cells. Erythroid progenitors were analyzed by staining with Ter119 and CD71 antibodies as described previously (Liu et al., 2006). HSC in BM, spleen and peripheral blood were stained with the following lineage antibodies: CD3, CD4, CD8, B220, Ter119, Gr1 and Mac1. LT-HSC were defined as Lin^-Kit⁺Sca⁺CD34^-CD150⁺. FL-HSC were stained using a similar lineage cocktail with the omission of Mac1. FL-HSC were gated as Lin^-Kit⁺Sca⁺CD48^-CD150⁺. Myeloid progenitors were stained using the lineage cocktail of LT-HSC in addition to IL7Ralpha, IgM and CD19. The progenitors were gated as megakaryocyte/erythrocyte progenitors (MEP): Lin-Kit⁺Sca^-CD34^-FcγR^-, common myeloid progenitors (CMP): Lin^-Kit⁺Sca^-CD34^lowFcγR^low, and granulocyte/macrophage progenitors (GMP): Lin^-Kit⁺Sca^-CD34^highFcγR^high. CD45 isotype analysis was done using CD45.1 PE/APC/APCCy7 and CD45.2 FITC/APC/PerCP5.5 antibodies. Cells were analyzed using FACSCalibur (BD Biosciences) or CyAn ADP (Beckman Coulter) and data was analyzed using FlowJo software (TreeStar, OR).

Cell cycle analysis

Cell cycle analysis of stem cells and erythroid progenitors was performed using Ki67-FITC (BD Biosciences) and DAPI according to the manufacturer's instructions. Briefly, cells were stained with biotinylated lineage antibodies as described for LT-HSC staining in addition to Kit (APC), Sca (PECy7) and CD34 (PE). Cells were fixed, and processed for Ki67-FITC and DAPI staining according to manufacturer's instructions. Gating for Ki67 was established by staining cells with lineage, Kit, Sca and CD34 antibodies in addition to the isotype control antibody provided by the manufacturer. Cell cycle entry of Lin^-Kit⁺Sca⁺ and Lin^-Kit⁺Sca⁺CD34^- cells was also determined by 5-bromo-2-deoxyuridine (BrdU) incorporation in conjunction with DAPI. Mice were injected intraperitoneally twice with BrdU 0.5mg/6g body weight, at 6 hr interval and euthanized 18 hr after the last injection. BrdU incorporation was detected using APC-anti-BrdU kit as per the manufacturer's instructions (BD Biosciences). Briefly, cells were stained with lineage antibodies-biotin (Streptavidin-APCCy7), Kit (PE), Sca (PECy7), CD34 (FITC), followed by fixation and staining with APC-anti-BrdU and DAPI. Cells were analyzed using CyAn ADP (Beckman Coulter). Cells positive for BrdU were gated as described earlier (Amrani et al., 2011; Challen et al., 2009).

Transplantation

Competitive repopulation assay was performed for in vivo analysis of stem cell function. Wild-type or Kit^{V558Δ;T669I/+} BM cells were mixed in 1:1 ratio with age (8-10 week) and sex matched CD45.1 BM cells. CD45.1 female recipient mice were irradiated (10 Gy: 5 Gy and 5 Gy, 3 hr apart) on the day of transplantation and injected with 1 million BM cell mixture. In instances of secondary transplantation, primary transplanted mice were euthanized after 16 weeks, BM cells were pooled from at least three mice and indicated numbers of cells (2 million or 5 million) were injected into irradiated CD45.1 recipients. To investigate if the spleen had functional HSC, 2 million whole spleen cells from wild-type or Kit^{V558Δ;T669I/+} mice were mixed with 200,000 CD45.1 BM competitor cells and transplanted into CD45.1 recipient mice. Reconstitution of donor (CD45.2) myeloid and lymphoid cells was analyzed by flow cytometry in the peripheral blood and wherever indicated in the BM and spleens of transplanted mice at 16 weeks post-transplantation. All transplanted mice were kept on Sulphatrim diet for at least 4 weeks. Percent chimerism was defined as (%CD45.2 donor cells)(100)/(%CD45.2 donor+%CD45.1 competitor cells) (Harrison et al., 1993; Morita et al., 2011).

Splenectomy

Wild-type and Kit^{V558Δ;T669I/+} mice were anesthetized, the splenic vessel was tied up and the spleen was surgically excised. Animals received buprenorphine for management of pain following surgery. Peripheral blood parameters were analyzed two weeks prior to the surgery to obtain pre-splenectomy values. After splenectomy mice were allowed to recover for 2 weeks and peripheral blood parameters were analyzed every week for up to 8 weeks. At the end of 8 weeks mice were euthanized and the BM was analyzed for erythroid progenitors by flow cytometry.

5-Fluorouracil treatment

5-Fluorouracil (5-FU, Sigma) 150 mg/kg body weight was administered to mice once intraperitoneally. Peripheral blood was drawn at regular intervals by retro-orbital bleeding to measure leucocyte number and hematocrit using Hemavet 950 (Drew Scientific). Mice were injected with BrdU as described above and cells harvested from BM were used for cell cycle analysis by flow cytometry using APC-anti-BrdU (BD Biosciences) and DAPI.

Statistical Analysis

Comparison between two groups was done by unpaired Students t-test analysis using GraphPad Prism (version 5.0). Statistical significance was achieved when P<0.05 indicated as following in the text *P<0.05, **P<0.01, ***P<0.001, P=NS (non-significant).

RESULTS

Increased erythroid generation in Kit^{V558Δ;T669I/+} mice and effect of splenectomy

Previously, we showed that the Kit^{V558Δ;T669I/+} mice have cecal GIST lesions, gastric and colonic ICC hyperplasia and have increased hematocrit values. In addition, the spleen cellularity is increased by 1.5 fold in mutant mice. In accordance with the increased hematocrit BFU-E progenitors in the BM and the spleen were also elevated in the Kit^{V558Δ;T669I/+} mice, BFU-E in Kit^{V558Δ;T669I/+} mice similar to their wild-type counterparts were EPO-dependent but their growth was more sensitive to KITL than wild-type progenitors (Bosbach et al., 2012). In order to gain insight into the various stages of erythroid maturation in wild-type and mutant mice, flow cytometric analysis of erythroid progenitors was performed as described previously (Liu et al., 2006). Different stages of erythroid maturation were classified on the basis of Ter119 and CD71 expression in conjunction with size of the cells (defined by forward scatter (FSC) property) (Fig. 1C). Proerythroblasts or R1 progenitors are defined as Ter119^loCD71^hi. As the erythroblasts mature discrete populations can be identified based upon gradual decline in CD71 expression and FSC as A:Ter119^hiCD71^hiFSC^hi, B:Ter119^hi,CD71^medFSC^lo and C:Ter119^hiCD71^loFSC^lo. Our results show that R1 progenitor frequency was higher in the mutant BM and significantly increased in the mutant spleen compared to wild-type (Fig 1A). Total R1 progenitor numbers were also significantly increased in the Kit^{V558Δ;T669I/+} BM and spleen (Fig 1B). The total numbers of the erythroblast subsets A and B were also significantly increased in the mutant spleen compared to wild-type (Fig 1B). In accordance with the increased R1 progenitors, spleen cytospins of Kit^{V558Δ;T669I/+} mice showed increased number of erythrocytes (data not shown). To assess if increase in R1 progenitor numbers was due to increased proliferation, cell cycle status of these cells was analyzed by Ki67 and DAPI staining followed by flow cytometry. In accordance with increased erythroid numbers in the Kit^{V558Δ;T669I/+} spleen, we found significantly increased frequency of R1 progenitors in S/G2/M phase of the cell cycle in the Kit^{V558Δ;T669I/+} spleen (Fig 1D). The cell cycle status in the mutant BM however was not significantly different compared to wild-type (Fig 1D). Given the increased number of erythroid progenitors in the spleen we sought to ascertain the role of the spleen in the development of increased hematocrit in the Kit^{V558Δ;T669I/+} mice, by performing splenectomy on 10 wk old wild-type and Kit^{V558Δ;T669I/+} mutant mice. Post-surgery, hematocrit values dropped in both the wild-type (55.92 ± 2.1 to 43.24 ± 2.38) and Kit^{V558Δ;T669I/+} (84.48 ± 2.32 to 52.88 ± 3.28, P<0.0001) mice, although the reduction was significantly greater in the Kit^{V558Δ;T669I/+} mice. Interestingly, the reduced post-surgery hematocrit values of Kit^{V558Δ;T669I/+} mice were maintained until the time point of last measurement, namely 8 weeks (Fig 1E). This is in contrast to the JAK2^V617F-GFP BM transplantation model of PV, where splenectomy caused a gradual decrease in hematocrit levels and only by 7 weeks normal wild-type hematocrit values were obtained (Mo et al., 2009). Erythroid progenitors assessed in the BM of splenectomized wild-type and Kit^{V558Δ;T669I/+} mice had similar numbers of R1 progenitors at 8 wk post-splenectomy (Fig 1F). The splenectomy results collectively suggest that the spleen contributes significantly toward elevating erythroid cell numbers in the Kit^{V558Δ;T669I/+} mice. Together our results show significantly elevated erythropoiesis in the Kit^{V558Δ;T669I/+} mice which is normalized by splenectomy. This may suggest a critical role for a combination of cell intrinsic property and the spleen microenvironment in the expansion of the erythroid compartment in the Kit^{V558Δ;T669I/+} mice.

Figure 1. Kit^{V558Δ;T669I/+} mice develop erythrocytosis which is remedied by splenectomy.

Total BM and spleen (SP) cells were prepared for staining with Ter119 and CD71 antibodies for flow cytometry (A-B) Graphical representation of flow cytometry results show increased frequency and total Ter119⁺CD71⁺ cells, especially R1 erythroid progenitor population in the Kit^{V558Δ;T669I/+} (GTK) BM and spleen. Frequency and total numbers of A and B populations are also significantly increased in the mutant spleen compared to wild-type (WT) (n=4-9). (C) Representative flow cytometric histograms of bone marrow and spleen cells show cells labeled with Ter119 and CD71 antibodies. The frequency of R1 and Ter119hi populations in total BM or spleen cells is depicted. Ter119hi population was further analyzed into A, B and C subsets and numbers indicate frequency of cells within total BM or spleen cells. (D) BM and spleen cells were stained with cell surface markers Ter119 and CD71 and processed for Ki67-DAPI staining. Erythroid R1 progenitor cycling measured by Ki67 flow cytometry in the BM and spleen revealed increased cycling of progenitors in the Kit^{V558Δ;T669I/+} spleen compared to wild-type but no change was observed in the BM of the two groups (n=3-5). (E) In order to assess the contribution of the spleen toward increased erythropoiesis in the Kit^{V558Δ;T669I/+}, wild-type and Kit^{V558Δ;T669I/+} mice were splenectomized. Hematocrit (HCT) levels two weeks post-surgery were significantly reduced in Kit^{V558Δ;T669I/+} splenectomized mice compared to pre-surgical values, while wild-type hematocrit levels did not decrease significantly (n=5) (F) Flow cytometric analysis of BM from wild-type and Kit^{V558Δ;T669I/+} by Ter119 and CD71 staining at 8 weeks post-splenectomy showed similar numbers of erythroid R1 progenitors in the BM of Kit^{V558Δ;T669I/+} compared to wild-type splenectomized mice (n=3). Data is representative of at least two independent experiments. n-indicates number of mice per genotype. Values are mean±SEM.

Myeloid progenitor expansion in the Kit^{V558Δ;T669I/+} mice

The increased number of erythroid progenitors in the Kit^{V558Δ;T669I/+} mice led us to investigate if myeloid subsets were also increased in these mice. Flow cytometric analysis of granulocyte and monocyte-macrophage populations in the BM and spleen revealed increased frequency and total cell numbers of both these cell types in the Kit^{V558Δ;T669I/+} mice (Fig 2A-C). In order to assess whether homeostasis was also altered upstream in the myeloid hierarchy, a detailed flow cytometric analysis of myeloid progenitors namely CMP, GMP and MEP was carried out as described previously (Akashi et al., 2000). In the BM of wild-type and Kit^{V558Δ;T669I/+} mice, while the frequency and total numbers of CMP and GMP populations did not differ significantly, the MEP population in the Kit^{V558Δ;T669I/+} BM was significantly increased compared to wild-type in terms of both frequency and total numbers (Fig 2D,E,H). In the Kit^{V558Δ;T669I/+} spleen the frequency of all three progenitor types was significantly increased compared to wild-type (Fig 2F,H). A significant elevation in total CMP and MEP numbers was also observed in the Kit^{V558Δ;T669I/+} spleen (Fig 2G). In order to physiologically assess the expansion of some of these myeloid populations, in-vitro colony assays were performed. GM colonies were increased twofold in the Kit^{V558Δ;T669I/+} BM and several fold increased in the Kit^{V558Δ;T669I/+} spleen. In addition, GEMM colonies were 12 fold greater in the Kit^{V558Δ;T669I/+} spleen compared to wild-type reiterating the elevated expansion of myeloerythroid progenitors in the spleens of these mice (table S1). Reconstitution assays such as CFU-S, also provide a measure of in-vivo progenitor function. It has been described previously that CFU-S day 8 correspond to MEPs, CFU-S day 9 to CMP and CFU-S day 12 to a mixture of MPP and CMP (Morrison and Weissman, 1994; Na Nakorn et al., 2002; Sharma et al., 2007). Transplantation of BM cells from wild-type or Kit^{V558Δ;T669I/+} mice into irradiated C57BL/6 mice, showed no difference in day 11 CFU-S colonies between wild-type and Kit^{V558Δ;T669I/+} (wild-type: 8.25 ± 4.11; Kit^{V558Δ;T669I/+}: 10 ± 3.36; P=NS, n=5) which is in accordance with unchanged CMP numbers in the BM of wild-type and mutant as assessed by flow cytometry (compare Fig 2D,E). Investigation of the lymphoid lineages by flow cytometry interestingly revealed reduced frequency of B220⁺, CD4⁺ and CD8⁺ cells in the Kit^{V558Δ;T669I/+} BM and spleen compared to wild-type (Fig 3A,D), however respective total numbers were not statistically different (Fig 3B,E). In summary, our results show that the Kit gain-of-function mutation affects the myeloid lineage and results in significant expansion of myeloerythroid progenitors in the Kit^{V558Δ;T669I/+} mice.

Figure 2. Expansion of the myeloid lineage in the adult BM and spleen of Kit^{V558Δ;T669I/+} mice.

Cells from the BM and spleen were stained with Gr1 and Mac1 antibodies and analyzed by flow cytometry. (A) Analysis of Mac1⁺-Gr1⁺ population in the BM and spleen showed significantly increased frequency of granulocyte and monocyte populations in Kit^{V558Δ;T669I/+} mice compared to wild-type (n=4-6). (B) Total cell numbers of granulocyte and monocyte cells is significantly increased in the Kit^{V558Δ;T669I/+} spleen compared to wild-type (n=4-6). (C) Representative staining profiles show frequency of granulocytes and monocytes in total BM and spleen. (D,E) BM cells were stained against lineage markers and Lin^-Kit⁺Sca^- cells were distinguished into CMP, GMP and MEP subsets according to their CD34 and FcγRII/III surface expression. Flow cytometric analysis showed increased frequency of MEP in the mutant BM compared to wild-type. Total MEP numbers in the mutant BM were increased 5-fold over wild-type (n=3-4). (F,G) Spleen cells from wild-type and mutant stained for myeloid progenitor markers showed increased frequency of CMP, GMP and MEP subsets in the Kit^{V558Δ;T669I/+} spleen. While total CMP numbers were significantly increased in the mutant spleen, MEP numbers were remarkably (20-fold) higher in Kit^{V558Δ;T669I/+} compared to wild-type (n=3-4). (H) Representative flow cytometry plot for myeloerythroid progenitors. Lin^-Sca^-Kit⁺ (LSK)cells were divided into GMP, CMP and MEP subsets. Numbers indicate frequency of cells in total BM and spleen. Data is representative of at least two independent experiments. Values are mean±SEM. n-indicates number of mice.

Figure 3. Lymphopoiesis in Kit^{V558Δ;T669I/+} mice.

(A-B) BM and spleen cells were stained with B220 antibody and analyzed by flow cytometry. The frequency of B220⁺ cells was significantly reduced in the Kit^{V558Δ;T669I/+} BM and spleen compared to WT (n=3-5), however total cell numbers were not significantly reduced compared to WT (n=3-5). (C) Representative staining profile shows frequency of B220⁺ cells in BM and spleen. (D,E) The frequency of CD4⁺ and CD8⁺ cells were reduced in the Kit^{V558Δ;T669I/+} BM and spleen, however total cell numbers were not significantly reduced compared to WT (n=3-5). (F) Representative staining profile shows frequency of CD4⁺ and CD8⁺ subsets in BM and spleen. Data is representative of at least two independent experiments. Values are mean±SEM. n-indicates number of mice.

Kit^{V558Δ;T669I/+} mutation results in increased stem cell numbers at extramedullary sites and in the fetal liver

We first sought to determine whether in addition to the expansion of MEPs in the Kit^{V558Δ;T669I/+} mice, there was a change in the HSC-enriched Lin^-Sca⁺Kit⁺ (LSK) population. Interestingly, flow cytometric analysis revealed no significant differences in LSK frequency or numbers in the wild-type and Kit^{V558Δ;T669I/+} BM (Fig. 4A-B). LSK frequency and total LSK numbers however, were significantly higher in the Kit^{V558Δ;T669I/+} spleen compared to wild-type (Fig 4A-B). Based on these results, we further analyzed if long term-HSC (LT-HSC) compartment defined phenotypically as Lin^-Sca⁺Kit⁺CD34^-CD150⁺ (Czechowicz et al., 2007) was altered. While no significant differences in the LT-HSC numbers of wild-type and Kit^{V558Δ;T669I/+} BM (Fig 4C-D) were observed, as with the LSK subset interestingly LT-HSC frequency and total cell numbers were significantly higher in the Kit^{V558Δ;T669I/+} spleen compared to wild-type (Fig 4C-D). In addition higher frequency of LT-HSC was also found in the peripheral blood of Kit^{V558Δ;T669I/+} mice (Fig 4E). These results show elevated HSC numbers in Kit^{V558Δ;T669I/+} mice with dissemination of stem cells from their primary location, i.e. the BM, to extramedullary sites. Previous studies have defined a clear role for KIT and KITL in the expansion of fetal liver HSC (FL-HSC) (Bowie et al., 2007; Ikuta and Weissman, 1992). At 14.5 dpc, Kit^{V558Δ;T669I/+} fetal livers had significantly greater cellularity compared to wild-type (Fig 5A). At this fetal age, equivalent numbers of LSK progenitor cells were observed in fetal livers of Kit^{V558Δ;T669I/+} and wild-type embryos (Fig 5B-C). We next investigated LT-HSC in the FL using the markers Lin^-Sca⁺Kit⁺CD150⁺CD48^- (McKinney-Freeman et al., 2009; Ogawa et al., 2001). Interestingly, similar to the Kit^{V558Δ;T669I/+} spleen, LT-HSC frequency and total cell numbers were significantly higher in the Kit^{V558Δ;T669I/+} fetal livers (Fig 5D,E). Lineage analysis also revealed significantly higher number of R1 erythroid progenitors in Kit^{V558Δ;T669I/+} fetal livers (Fig 5F). However, total B220⁺ or Gr1⁺ cells were similar in Kit^{V558Δ;T669I/+} and wild-type fetal livers (Fig 5G,H).

Figure 4. Analysis of the stem cell compartment in wild-type and Kit^{V558Δ;T669I/+} mice.

(A-B) Total BM and spleen cells were stained against lineage cocktail and gated as Lin^-Kit⁺Sca⁺ (LSK). LSK frequency and total cell numbers were not significantly altered in the Kit^{V558Δ;T669I/+} BM compared to wild-type BM but in the Kit^{V558Δ;T669I/+} spleen total LSK numbers were 11-fold elevated compared to wild-type spleen (n=5). (C-D) LT-HSC defined as Lin^-Sca⁺Kit⁺CD150⁺CD34^- were significantly higher in frequency and total cell numbers were 6-fold higher in Kit^{V558Δ;T669I/+} spleen compared to wild-type spleen. LT-HSC numbers were not significantly different in the BM of the two mice (n=4-5). (E) The frequency of LT-HSC numbers was increased in the peripheral blood of Kit^{V558Δ;T669I/+} compared to wild-type mice (n=4). (F) Representative staining profiles for LSK and LT-HSC from BM and spleen. Lineage^- cells were electronically gated into LSK subset. LSK cells were subdivided into LT-HSC, multipotent progenitors (MPP) and lymphoid-primed multipotent progenitors (LMPP). Numbers represent frequency of cells in total BM or spleen. Data is representative of at least three independent experiments. Values are mean±SEM. n-indicates number of mice.

Figure 5. Analysis of fetal livers from wild-type and Kit^{V558Δ;T669I/+} mice.

(A) The total number of nucleated cells is higher in Kit^{V558Δ;T669I/+} fetal livers at 14.5 dpc. (B,C) LSK frequency and numbers were not significantly different in Kit^{V558Δ;T669I/+} mice compared to WT. (D,E)LT-HSC frequency and number were significantly higher in Kit^{V558Δ;T669I/+} fetal livers compared to wild-type. (F) Analysis of erythroid progenitors by staining fetal livers against Ter119 and CD71 antibodies revealed significantly higher R1 progenitors in Kit^{V558Δ;T669I/+} compared to wild-type fetal livers, while the erythroid subsets A, B and C did not vary significantly between the two groups. (G,H) Lineage analysis revealed no significant difference in B-lymphoid (B220⁺), or myeloid (Gr1⁺) cell numbers between Kit^{V558Δ;T669I/+} and wild-type mice. Graphs A-H: data is representative of at least two independent experiments. n=6-8 fetal livers per group. Values are mean±SEM.

Increased self-renewal of hematopoietic stem cells in the Kit^{V558Δ;T669I/+} mice

Whereas relatively mild Kit loss-of-function mutations had been shown to diminish hematopoietic stem cell function, it was unclear how Kit gain-of-function mutations would affect HSC function (Sharma et al., 2007). To investigate whether the Kit^V558Δ;T669I mutation affected stem cell function, we carried out competitive repopulation assays. In primary transplantation experiments 5×10⁵ BM cells from wild-type or Kit^{V558Δ;T669I/+} were competed with an equal number of congenic CD45.1 BM cells. At 16 weeks post-transplantation, flow cytometric analysis of peripheral blood revealed that Kit^{V558Δ;T669I/+} cells were able to repopulate the recipients with greater efficiency compared to wild-type cells (Fig 6A). A higher percentage of myeloid (Mac1⁺Gr1⁺) chimerism was observed in the peripheral blood of recipients that received Kit^{V558Δ;T669I/+} BM cells compared to wild-type (Fig 6B). Analysis of myeloid (Mac1⁺Gr1⁺) and B-lymphoid (B220⁺) lineages in the BM and spleen of primary recipients revealed significantly higher repopulation by Kit^{V558Δ;T669I/+} than wild-type in the myeloid but not the B220⁺ lineage (Fig 6C,D). Interestingly, HSC analysis (LSKCD34^-) in the BM did not reveal significantly higher Kit^{V558Δ;T669I/+} chimerism in the primary recipients (Fig 6E). In order to test the long term self-renewal potential of HSC, BM cells from three primary recipient mice from wild-type or Kit^{V558Δ;T669I/+} groups were pooled and 2 million or 5 million cells were injected into two cohorts of lethally irradiated CD45.1 mice. Flow cytometric analysis of peripheral blood to determine donor chimerism at 16 weeks revealed significantly higher donor and myeloid chimerism in both cohorts that received Kit^{V558Δ;T669I/+} cells (Fig 6F,G). Furthermore, analysis of repopulation efficiency in the BM and spleen of secondary recipients showed significantly greater donor contribution toward myeloid and B-lymphoid lineages in the recipients receiving Kit^{V558Δ;T669I/+} cells (Fig 6H,I). Interestingly, analysis of LSKCD34^- HSC in the BM of secondary recipients revealed significantly higher donor contribution in the mice receiving Kit^{V558Δ;T669I/+} cells (Fig. 6J) indicating increased long-term self renewal of Kit^{V558Δ;T669I/+} HSC in comparison with wild-type HSC. Since the Kit^{V558Δ;T669I/+} mice showed an increased number of LT-HSC in the spleen, we tested if the HSC in the spleen were functional. Spleen cells (2×10⁶) from wild-type or Kit^{V558Δ;T669I/+} were mixed with CD45.1 competitor BM cells (2×10⁵) and transplanted into lethally irradiated congenic recipients. At 16 weeks post-transplantation, analysis of donor chimerism in peripheral blood revealed significantly higher total, myeloid and B-lymphoid donor contribution in recipients that were transplanted with Kit^{V558Δ;T669I/+} spleen cells (Fig 6K, L). Our results reveal that the Kit^V558Δ;T669I mutation confers greater advantage to stem cells in terms of self-renewal and that the HSCs in the spleen are capable of long-term self renewal.

Figure 6. Analysis of HSC self-renewal from wild-type and Kit^{V558Δ;T669I/+} BM and spleen.

(A,B) CD45.1 mice were lethally irradiated and transplanted with CD45.1 whole BM cells (5×10⁵) in competition with wild-type or Kit^{V558Δ;T669I/+} (CD45.2) BM cells (5×10⁵). Peripheral blood analyzed for donor contribution in total and myeloid compartments reveals increased donor contribution in recipients transplanted with CD45.1 and Kit^{V558Δ;T669I/+} cell mixture. (n=8-9) (C-D) BM and spleen analyzed for CD45.2 contribution in myeloid and B-lymphoid compartments reveal higher donor contribution for the myeloid but not B-lymphoid lineage in the Kit^{V558Δ;T669I/+} group. (n=4-6) (E) BM analyzed for donor contribution in LSKCD34^- stem cell compartment did not show significant difference among the two groups. (n=5). Secondary transplantation was carried out in lethally irradiated CD45.1 recipients with 2 million and 5 million pooled whole BM cells from three donors in each primary transplanted group. (F-G) Reconstitution in peripheral blood analyzed in total and myeloid populations 16 wks after transplantation showed higher donor contribution in both the 2 million and 5 million Kit^{V558Δ;T669I/+} transplanted groups (n=4-5). (H-I) CD45.2 donor contribution in BM and spleen for myeloid and B-cell subsets was also higher in the Kit^{V558Δ;T669I/+} groups (n=4-5). (J) Significantly higher CD45.2 contribution was observed in the LSKCD34^- stem cell compartment in the BM of recipients that were injected with 5 million CD45.1: Kit^{V558Δ;T669I/+} cells. (n=4-5). (K-L) CD45.1 mice were also transplanted with 2×10⁶ whole spleen cells from WT or Kit^{V558Δ;T669I/+} and co-transplanted with 2×10⁵ CD45.1 whole BM cells. Reconstitution was measured in peripheral blood 16 weeks post-transplantation. Significantly higher donor contribution was observed in total, myeloid and B-lymphoid subsets in the Kit^{V558Δ;T669I/+} donor group n=5 mice in each group. Data is representative of at least two independent experiments. Values in graphs A-L are mean±SEM.

Cell cycle status and 5-FU stress response of wild-type and Kit^{V558Δ;T669I/+} HSC

We next ascertained whether the oncogenic Kit mutation conferred any changes toward the proliferative rate of stem cells by analyzing BrdU incorpotation in stem cells. Interestingly, BM LSKCD34^- stem cells from Kit^{V558Δ;T669I/+} mice incorporated more BrdU compared to wild-type (Fig. 7A, B). Although a significantly higher number of BrdU positive stem cells was observed in the mutant BM, the difference between the wild-type and mutant was less than two-fold. The self-renewal property of HSCs is linked to the cycling status of HSC, especially to cellular quiescence. In order to distinguish if there was any difference in the quiescent G₀ and resting G₁ phases of the cell cycle we stained LSKCD34^- stem cells with Ki67 in conjunction with DAPI which allows distinction between G₀ and G₁ phases of the cell cycle. The percentage of quiescent and resting LSKCD34^- stem cells in the mutant BM (Fig. 7C, D) or spleen (Fig. 7E, F) did not differ significantly from wild-type. Moreover, stem cells from the mutant and wild-type did not differ significantly in the S/G₂/M phase of cell cycle. To test whether the Kit^V558Δ;T669I mutation provided an advantage to HSC during stress hematopoiesis, we subjected wild-type and Kit^{V558Δ;T669I/+} mice to a single dose of 150 mg/kg 5-FU. 5-FU results in the elimination of proliferating stem and progenitor cells thereby inducing activation of the quiescent HSCs to enter the cell cycle and replenish the lost progenitors. A comparison of blood parameters showed that as expected 5-FU treatment caused a rapid decline in the WBC and neutrophil counts in both wild-type and Kit^{V558Δ;T669I/+} mice between d4 and d8 of 5-FU treatment (Fig 7G,H). At d16 post-5FU, the recovery of WBC and neutrophil counts was near baseline levels in wild-type mice in contrast to Kit^{V558Δ;T669I/+} where an increase in their numbers above baseline was observed (Fig 7G,H). Between d4 and d8 %HCT levels in wild-type mice also declined from an average 50% at baseline to 25-30%, which is in accordance with our previous results (Agosti et al., 2008). In sharp contrast, %HCT levels showed only a small decline during this timeframe in Kit^{V558Δ;T669I/+} mice (Fig 7I). In order to assess whether 5-FU induces an accelerated cycle entry in Kit^{V558Δ;T669I/+} HSC, 5-FU treated mice were administered BrdU and the cells from the BM were subjected to cell cycle analysis. Since it has been established previously that 5-FU treatment results in a transient loss of KIT expression on HSC and that repopulating activity is predominant in the Lin^-Sca⁺ population, we analyzed the Lin^-Sca⁺ population for changes in cell cycle status (Nemeth et al., 2006; Randall and Weissman, 1997). While the frequency of Lin^-Sca⁺ cells in the BM did not change in Kit^{V558Δ;T669I/+} compared to wild-type between d4-d16 post-5-FU (data not shown), the fraction of Lin^-Sca⁺ cells that were BrdU positive was significantly higher in Kit^{V558Δ;T669I/+} at d8, and the numbers remained higher than wild-type even at d16 (Fig 7J). Taken together these results then imply that 5-FU induced hematopoietic stress accelerates HSC proliferation of Kit^{V558Δ;T669I/+} cells compared to wild-type cells.

Figure 7. Stem cell proliferation in wild-type and Kit^{V558Δ;T669I/+} BM and spleen, and response to 5-FU.

(A) Stem cell turnover was measured in the BM by BrdU incorporation. Graphical representation of percentage of BrdU⁺BM LSKCD34^- stem cells in mutant mice. Values are mean±SEM; n=7 mice in each group. (B) Representative BrdU staining profile of LSKCD34^- stem cells from wild-type and mutant BM. (C) BM LSKCD34^- stem cells in G₀ (quiescent), G₁ (resting) and S/G₂/M (proliferating) phases of cell cycle as measured by Ki67 and DAPI staining. .(D) Representative staining profile of LSKCD34^- cells from wild-type and mutant BM stained with Ki67 and DAPI. (E) Spleen LSKCD34^- stem cells from wild-type and mutant in G₀, G₁ and S/G₂/M phases of cells cycle measured by Ki67 staining. (F) Representative staining profile of LSKCD34^- cells from wild-type and mutant spleen stained with Ki67 and DAPI. Values in C and E are mean±SEM; n=5 mice in each group. Data is representative of at least three independent experiments.

(G,H) A single dose of 150 mg/kg body weight 5-FU led to reduced WBC and neutrophil numbers in both wild-type and mutant at d8 of 5-FU treatment followed by a recovery in numbers by d16. (I) Hematocrit did not fall sharply in Kit^{V558Δ;T669I/+}, however in the wild-type hematocrit declined sharply at d8 after treatment (J) 5-FU treated wild-type and mutant mice were administered BrdU one day before each analysis and BM was harvested to analyze BrdU incorporation in Lin^-Sca⁺ cells by flow cytometry. At d8 the number of cycling cells is higher in the Kit^{V558Δ;T669I/+} mice compared to wild-type. Values are mean±SEM. n=at least 3 mice in each group per time-point.

DISCUSSION

A hallmark of KIT-KITL signaling is its role in the erythroid lineage evident in the form of macrocytic anemia which is observed in mice with Kit loss-of-function mutations. Recently in mice carrying the Kit^{V558Δ;T669I/+} gain of function mutation we showed that the hematocrit was elevated due to increased BFU-E progenitors which were hypersensitive to KITL concentrations (Bosbach et al., 2012). In accordance with increased numbers of BFU-E the numbers of erythroid progenitors assessed in the present study were also increased in the spleen of Kit^{V558Δ;T669I/+} mice and greater frequency of progenitors were in cell cycle in the Kit^{V558Δ;T669I/+} spleen than in the Kit^{V558Δ;T669I/+} BM. In order to test if the spleen was essential for the extramedullary erythropoiesis observed in the Kit^{V558Δ;T669I/+} mice, we performed splenectomy. Splenectomy normalized HCT in the Kit^{V558Δ;T669I/+} mice to near wild-type levels suggesting a highly significant contribution of the spleen toward maintenance of high erythroid output in Kit^{V558Δ;T669I/+} mice. Similar results have been obtained in JAK2^V617F mutant mice where splenectomy resulted in normalizing HCT levels to wild-type levels and in EPO transgenic mice where splenectomy substantially reduced hematocrit levels (Mo et al., 2009; Vogel et al., 2003). KIT has an important role in the expansion of the cells of the myeloid lineage especially mast cells. But unlike mast cells, where the role of KIT in their expansion and survival has been well documented, the importance of KIT in the expansion of cells of other myeloid lineages or their precursors is not well defined. BAC transgenic mice with a Kit^D816V mutation develop severe mastocytosis, have increased erythroid cell numbers and develop B-cell tumors yet these mice did not have increased myeloid cell numbers (Gerbaulet et al., 2011). Interestingly, in many patients the Kit^D816V mutation was detected not only in mast cells but also in earlier hematopoietic progenitor cells and other myeloid cells indicating that this mutation may be conferring proliferative capacity not only to mast cells but also to myeloid progenitors (Garcia-Montero et al., 2006). Notably, Kit^{V558Δ;T669I/+} mice have increased granulocyte and monocyte numbers in the BM and spleen. Earlier work measuring CMP numbers based on CFU-S assays had shown that in Kit loss-of-function mutants such as W^v/+, W^42/+ and W^41/+, CMP numbers were reduced suggesting that Kit signaling may be important for proliferation of myeloid progenitors (Sharma et al., 2007). While we did not see changes in CMP or GMP or CFU-S colony numbers in the BM of Kit^{V558Δ;T669I/+} mice, in accordance with the increased erythroid numbers, MEPs were significantly increased in the BM of mutant mice compared to wild-type. Our results compare with the PV mouse model where MEP numbers were significantly increased in the BM but not CMP or GMP (Akada et al., 2010; Mullally et al., 2010). While Kit has a role in B-cell maturation (Agosti et al., 2004) and a profound effect on B-cell proliferation was observed in BAC transgenic Kit^D816V mice with the development of B-cell tumors (Gerbaulet et al., 2011), B-cell and T-cell numbers in Kit^{V558Δ;T669I/+} mice were not statistically different from wild-type. The effects observed on the erythroid and myeloid cells suggest that the myeloerythroid lineage is preferably affected in Kit^{V558Δ;T669I/+} mice.

The changes in multi-potential progenitor cells (LSK) or stem cell (LT-HSC) numbers were minimal in the BM of Kit^{V558Δ;T669I/+} mice compared to wild-type. In contrast LSK and LT-HSC were 11-fold and 6-fold increased in the spleen of Kit^{V558Δ;T669I/+} mice compared to wild-type. Additionally stem cells were found in circulation. The presence of LT-HSC in the peripheral blood or spleen has been observed in other myeloproliferative mouse models such as the Rb^Δ/Δ mice, where the BM environment is altered due to myelofibrosis, thereby favoring the mobilization of hematopoietic cells including LT-HSC from the marrow to the spleen (Walkley et al., 2007). Since the marrow of Kit^{V558Δ;T669I/+} mice did not turn fibrotic as the mice aged (Bosbach et al., 2012) it is possible that a combination of constrained space and cell-intrinsic factors in the marrow forced the LT-HSC into the spleen. The requirement of HSC for KITL increases during ontogeny, such that adult HSC require more KITL than their fetal counterparts (Bowie et al., 2007; Ikuta and Weissman, 1992). Hence we assessed if increased KIT signaling in the Kit^{V558Δ;T669I/+} mice had any effect on fetal HSC numbers. While fetal livers of Kit^{V558Δ;T669I/+} mice had similar numbers of LSK progenitor cells compared to wild-type, LT-HSC numbers were increased more than two-fold compared to wild-type in agreement with a reduced threshold for KIT signaling in the fetal liver.

KIT has been shown to be important for self-renewal of LT-HSC. The BM from W⁴¹/W⁴¹ mice or W^42/+ are deficient in long-term generation of hematopoiesis in recipient BM (Miller et al., 1996; Sharma et al., 2007; Thoren et al., 2008) and interestingly long-term but not short-term competitive activity of W⁴¹/W⁴¹ fetal liver cells was also found to be defective (Miller et al., 1997). Evaluation of the reconstitution ability of Kit^{V558Δ;T669I/+} hematopoietic stem cells by competitive repopulation assays showed that donor contribution toward the myeloid lineages was higher with the Kit^{V558Δ;T669I/+} donors compared to wild-type in primary and secondary transplantations. Notably, while the donor derived stem cell numbers were similar for wild-type and Kit^{V558Δ;T669I/+} donors after primary transplantation, the contribution of donor derived stem cells increased in Kit^{V558Δ;T669I/+} donors following secondary transplantation, indicating increased self-renewal capacity of Kit^{V558Δ;T669I/+} derived donor LT-HSC . The presence of significant numbers of LT-HSC s in the spleen of Kit^{V558Δ;T669I/+} mice prompted us to test if they were functional (Morita et al., 2011). Interestingly, transplanting CD45.1 recipients with wild-type or Kit^{V558Δ;T669I/+} spleen cells in competition with congenic BM cells revealed significant spleen-derived multi-lineage activity at 16 weeks, showing that the Kit^{V558Δ;T669I/+} mice have functional HSC in the spleen. While the precise role of the HSC in the spleen in general is unclear, it has been suggested that the spleen may function as a storehouse of HSC to rapidly generate erythroid cells in conditions of stress (Morita et al., 2011). The increased self-renewal of HSC in our gain-of-function mutant supports earlier work which showed that KITL greatly enhances self-renewal ability of both fetal and adult stem cells, despite their different sensitivity to this factor (Bowie et al., 2007). While we observed higher BrdU incorporation in the Kit^{V558Δ;T669I/+} stem cells, we did not observe significantly decreased quiescence or increased S/G₂/M phase of Kit^{V558Δ;T669I/+} BM and spleen HSC in comparison to wild-type, indicating that the effect of Kit mutation on self-renewal is perhaps greater than its effect on proliferation of stem cells. 5-FU is known to deplete cycling HSCs and progenitors while sparing resting LT-HSCs due to their slower cycling characteristics (Lerner and Harrison, 1990). These cells then expand and regenerate the depleted progenitors. Since the Kit^{V558Δ;T669I/+} LT-HSC had increased self-renewal capacity based on competitive repopulation properties, we tested if upon 5-FU treatment, these cells would show an accelerated response to stress. We saw increased expansion of WBC and neutrophils in the 5-FU treated Kit^{V558Δ;T669I/+} mice indicating rapid expansion of progenitors in these mice compared to wild-type. Furthermore, the percentage of HSC in cycle was also increased.

In conclusion, our analysis of the hematopoietic characteristics of the Kit^{V558Δ;T669I/+} mice shows that the Kit gain-of-function mutation has a significant effect on the proliferation of the myelo-erythroid cell lineage both in the BM and the spleen, which ultimately results in elevated hematocrit and microcytosis. While there is an expansion of LT-HSCs especially in the spleen of Kit^{V558Δ;T669I/+} mice, overall the Kit gain-of-function mutation has an important role in regulating the self-renewal of LT-HSC.