Non-essential role of S100A9 in hematopoiesis in primary MPN and MDS mouse models

S100 proteins govern cellular proliferation, differentiation, apoptosis, calcium homeostasis, energy metabolism, and inflammation. This heterodimeric calcium-binding protein, consisting of the S100A8 and S100A9 subunits, modulates myeloid cell development. S100A9 elevation has been found in myeloproliferative neoplasms (MPN), myelodysplastic syndrome (MDS), and acute myeloid leukemia (AML)^1–3. Notably, transgenic mice with S100A9 overexpression develop myeloid dysplasia that closely mimics human MDS². Haploinsufficiency in Rps14, a human del(5q) MDS-associated gene, leads to impaired erythroid differentiation and triggers overexpression of S100A8/S100A9³. However, whether endogenous S100A9 is essential for normal or pathological hematopoiesis in vivo is largely unknown.

To evaluate the roles of S100A9 in hematopoiesis, we analyzed S100A9^−/− mice. Peripheral blood parameters and bone marrow (BM) or spleen (SP) cellularity analysis showed scarcely any differences between WT and S100A9 null mice (Figures S1 and S2). We therefore induced stress hematopoiesis by treating the mice with 5-fluorouracil (5-Fu) and phenylhydrazine (PHZ). Unexpectedly, S100A9 loss had minimal effects on the mice during hematopoietic recovery (Figures S3 and S4).

We next asked whether S100A9 might play roles in hematological malignant conditions, such as MPN and MDS. We first crossed S100A9^−/− mice with MPN mice (JAK2^V617F/+Vav-Cre)—a model closely resembling human polycythemia vera. The JAK2^V617F mutation attenuated survival regardless of the presence of S100A9 expression (Figure 1A). Moreover, the splenomegaly in JAK2^V617F mutant mice was not alleviated by S100A9 deletion (Figure 1B). BM and SP cellularity did not differ between JAK2^V617FVav-Cre and JAK2^V617FS100A9^−/−Vav-Cre mice (Figure 1C). Erythrocytosis promoted by JAK2^V617F mutation was not alleviated by S100A9 depletion, despite a decline in platelet counts observed in JAK2^V617FS100A9^−/−Vav-Cre mice (Figure 1D and Figure S5A–T). This outcome, combined with the previously described role of S100A9 in facilitating the formation of procoagulant platelets through GPIbα⁴, underscores S100A9’s role in promoting thrombopoiesis. Subsequently, we conducted lineage analysis through flow cytometry assays. The JAK2^V617F mutant mice exhibited aberrant elevations in both the proportions and cell counts of TER119⁺ erythroid cells in the BM and SP (Figure 1D and Figure S5U). However, the deletion of S100A9 in the presence of JAK2^V617F mutation only slightly ameliorated the absolute number of erythroid cells in the BM but not the SP (Figure 1D), and the percentages remained unchanged (Figure S5U). Likewise, myeloid and lymphoid cells in BM or SP showed no differences between JAK2^V617F mutant mice and JAK2/S100A9 double-mutant mice (Figure 1D and Figure S5V). These results suggest that S100A9 does not substantially contribute to the pathogenesis of JAK2^V617F/+-driven polycythemia. A prior study has indicated that the overexpression of S100A9, triggered by the Srsf2^P95H mutation, hindered polycythemia in JAK2^V617F/+ mice⁵. Together, these results indicate that activation, rather than deletion, of S100A9 might be a suitable approach for MPN therapy.

• Download figure
• Open in new tab
• Download powerpoint

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

S100A9 is not required for hematopoietic dysfunction in the JAK2^V617F mutant MPN and age-related MDS mouse model. (A) Kaplan–Meier survival curve showing the survival of the indicated mice. Control, n = 10; S100A9^−/−, n = 9; JAK2^VF/+Vav-Cre, n = 23; JAK2^VF/+S100A9^−/−Vav-Cre, n = 22. (B) Quantitative analyses of spleen weight vs. body weight. n = 3 per group. (C) Absolute cell counts in the BM and SP from the indicated mice. n = 3 per group. (D) Flow cytometry analysis of TER119⁺ erythroid, lymphoid, and myeloid cells in the BM and SP. Top, quantification of absolute counts of TER119⁺ erythroid cells and various differentiation stages of erythroid cells. I–VI represent proerythroblasts (ProE), basophilic erythroblasts (BasoE), polychromatic erythroblasts (PolyE), orthochromatic erythroblasts (OrthoE), reticulocytes (Ret), and red blood cells (RBCs). n = 3 per group. Bottom, absolute counts of lymphoid and myeloid cells. B cells, B220⁺; T cells, CD3e⁺; myeloid cells, CD11b⁺; monocytes/macrophages, Ly6G⁻CD11b⁺; granulocytes, Ly6G⁺CD11b⁺. n = 3 per group. (E) Merged UMAP plots showing the distribution and overlap of annotated cell populations from the BM of aged (12- to 14-month-old) control and DKO mice. (F) As in I, except that the S100A9 expression pattern is illustrated on merged UMAP plots. (G) Violin plots showing S100A9 expression levels across the indicated cell populations. (H) Relative serum S100A9 levels, determined by multiplex ELISA. Control, n = 12; DKO, n = 7. (I) Kaplan–Meier survival analysis assessing the effects of combined S100A9 deletion in Diaph1^−/− Mir146^−/− mice. Control, n = 17; DKO, n = 40; TKO, n = 46. (J) Quantitative analysis of spleen weight normalized to body weight. (K) Absolute cell numbers in the BM and SP. (L) Flow cytometry analysis of TER119⁺ erythroid, lymphoid, and myeloid cells in the BM and SP. Top, quantification of absolute counts of TER119⁺ erythroid cells and various differentiation stages of erythroid cells. I, ProE; II, BasoE; III, PolyE; IV, OrthoE; V, reticulocytes; and VI, RBCs. Bottom, absolute counts of lymphoid and myeloid cells. B cells, B220⁺; T cells, CD3e⁺; myeloid cells, CD11b⁺; monocytes/macrophages, Ly6G⁻CD11b⁺; granulocytes, Ly6G⁺CD11b⁺. Control, n = 8; DKO, n = 5; TKO, n = 7. (M) Percentages (left) and absolute cell counts (right) of LK (Lin⁻c-Kit⁺Sca1⁻), LSK (Lin⁻c-Kit⁺Sca1⁺), and LS (Lin⁻c-Kit⁻Sca1⁺) cells in the BM lineage negative cells (Lin⁻) of the indicated mice. (N) SLAM-LSK (CD150⁺CD48⁻LSK) and MPP (CD150⁻CD48⁻LSK) were gated from LSK cells of the indicated mice; frequency (left) and cell count (right) are shown. (O) Analysis of LT-HSC (CD135⁺CD34⁻LSK), ST-HSC (CD135⁻CD34⁻LSK), and MPP (CD135⁻CD34⁺LSK), similarly to N. (P) Colony formation assays of bone marrow mononuclear cells isolated from the indicated mice. Colonies were scored 14 days post-plating. (Q) Fibrosis in the sternum from the indicated mice, assessed through reticulin staining, with the fibrotic index quantified by fiber length measurement. DKO, Diaph1^−/− Mir146^−/−; TKO, Diaph1^−/−Mir146^−/− S100A9^−/−. Control, n = 8; DKO, n = 5; TKO, n = 7 in H–J; control, n = 5; DKO, n = 4; TKO, n = 3 in J–L; control, n = 10; DKO, n = 8; TKO, n = 6 in P; quantified colony numbers in Q were as follows: control, n = 15; DKO, n = 55; TKO, n = 49. All data are shown as mean ± SEM. All P values were calculated with 2-tailed unpaired Student’s t-tests, except in G, in which the Wilcoxon Rank-Sum test was used. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

We have previously reported that mice with dual deficiency in mDia1 and MiR-146a (DKO mice) show age-related anemia and ineffective erythropoiesis mimicking human MDS⁶. Subsequent single-cell RNA-sequencing analysis of BM nucleated cells revealed enriched S100A9 expression in neutrophils, dendritic cells, fibroblasts, and macrophages (Figure 1E–G)⁷. Specifically, S100A9 was aberrantly overexpressed in B cells, monocytes, stage II neutrophils, T cells, NK cells, neutrophils, and macrophages in DKO mice (Figure 1G). Moreover, serum S100A9 was significantly higher in DKO mice than controls (Figure 1H). These results collectively suggested that S100A9 might potentially contribute to MDS pathogenesis. Therefore, we crossed the DKO mice with S100A9^−/− mice to generate Diaph1^−/−Mir146^−/−S100A9^−/− triple knockout (TKO) mice to interrogate the roles of S100A9 in MDS pathogenesis. Firstly, the survival curves were similar between DKO and TKO mice during aging (Figure 1I). Accordingly, peripheral blood parameters showed minimal differences between DKO and TKO mice, except for neutrophil counts (Figure S6A–T). An enlarged SP and extramedullary erythropoiesis were consistently observed in DKO mice and were not alleviated by S100A9 ablation (Figure 1J–K). Moreover, the loss of BM cellularity in DKO mice was not ameliorated by S100A9 deletion (Figure 1K). The impaired BM erythropoiesis in DKO animals, as evidenced by the disappearance of varied developmental stages of erythroid cells, was not recovered or ameliorated after S100A9 loss in TKO mice (Figure 1L and Figure S6U). We observed similar patterns for lymphoid cells (Figure 1L and Figure S6V, B220⁺ and CD3e⁺). Notably, in agreement with the diminished peripheral neutrophil counts in TKO mice (Figure S6R), myelopoiesis was further impaired in DKO mice after S100A9 inactivation (Figure 1L and Figure S6V, CD11b⁺, monocytes/macrophages, and granulocytes). This observation might be attributable to S100A9-mediated modulation of CD11b expression on neutrophils under pathological conditions such as inflammation⁸. Secondly, our analysis of the hematopoietic stem progenitor cells (HSPCs) in the BM and SP revealed that the common progenitors (LK, CMP, and GMP), multipotent progenitors (MPP), and hematopoietic stem cells (HSCs), including long-term HSCs (LT-HSCs) and SLAM-LSK were depleted in the BM but markedly accumulated in the SP in aged DKO mice (Figure 1M–O and Figure S7A–D). Intriguingly, although S100A9 deletion did not restore BM HSPC levels, it substantially decreased splenic HSPC accumulation in DKO mice (Figure 1N). This finding was further validated through in vitro colony-forming assays in mononuclear cells (Figure 1P and Figure S7E–H). Our findings suggested that S100A9 is critical for either HSPC migration from the BM into the SP or the expansion of splenic HSPCs in the context of MDS. Thirdly, aged DKO mice exhibited fibrosis, as evidenced by reticulin staining in BM sections. However, the inactivation of S100A9 scarcely ameliorated the fibrotic conditions in DKO mice (Figure 1Q and Figure S7I).

• Received April 13, 2025.
• Accepted June 3, 2025.

• Copyright: © 2025, The Authors