INTRODUCTION
Myelodysplastic neoplasms (MDSs) are among the most common hematological neoplasms in the elderly population. This group of clonal hematopoietic stem and progenitor cell (HSPC) disorders are characterized by ineffective hematopoiesis, peripheral cytopenia, and a risk for transformation into acute myeloid leukemia (AML). Overall, therapeutic options are limited, and allogeneic hematopoietic cell transplantation (allo‐HCT) remains the only potential curative treatment option. However, allo‐HCT is a burdensome approach associated with numerous complications and is therefore reserved for fit, high‐risk patients. Patients with lower risk (LR) disease, who make up two‐thirds of the MDS population, suffer predominantly from anemia and are mostly treated with supportive care (i.e., red blood cell [RBC] transfusions), erythropoiesis‐stimulating agents (ESAs), or luspatercept. These treatments offer only temporary relief, and supportive transfusions are associated with impaired quality of life, increased cardiovascular risk, and toxic iron overload. 1 Indeed, disease modification cannot be achieved with these therapies, and a relevant proportion of LR‐MDS patients will ultimately progress to high‐risk MDS or even AML. Therefore, new therapeutic options are urgently needed.
Aberrant innate immune responses and a pro‐inflammatory bone marrow (BM) microenvironment play significant roles in MDS pathogenesis. Both preclinical and clinical data provide a strong rationale for modulating the disease course by targeting inflammatory pathways, especially in LR‐MDS patients. 2 , 3 , 4
The danger‐associated molecular pattern (DAMP) molecules S100A8 and S100A9 are key mediators of aberrant inflammation. These molecules bind to Toll‐like receptor 4 (TLR4), triggering downstream signaling via IRAK1/4 and TRAF6. 3 This leads to increased NF‐κB‐mediated transcription of pro‐inflammatory cytokines, such as interleukin‐1β (IL‐1β), tumor necrosis factor α (TNF‐α), transforming growth factor β (TGF‐β), interleukin‐6 (IL‐6), and others, which promote further inflammation. 5 , 6 Additionally, S100A9 can stimulate myeloid‐derived suppressor cells (MDSCs) in an autocrine manner via CD33, leading to the secretion of hematopoiesis‐suppressive cytokines, such as IL‐10 and TGF‐β. 5
Thus, targeting the S100A9 pathway might be an interesting therapeutic strategy to slow down MDS progression.
Tasquinimod (TASQ, Active Biotech, Lund, Sweden) is a small‐molecule oral inhibitor and a second‐generation quinolone‐3‐carboxamide compound that binds to S100A9 and inhibits its interaction with receptors such as RAGE and TLR4. 7 , 8 , 9 It is also a histone deacetylase (HDAC) modulator that reduces HDAC‐dependent deacetylation of targets including HIF‐1α, thereby lowering HIF‐1α–mediated gene expression. 10
Originally tested for the treatment of prostate cancer, 11 TASQ has demonstrated immunomodulatory and anti‐tumor effects in several preclinical solid cancer models as well as in myelofibrosis. 12 , 13 In a recent study, it was demonstrated that TASQ exerts its primary effects on malignant JAK2V617F‐mutant megakaryocytes and monocytes in myeloproliferative neoplasms. 14 The drug is currently being tested in a Phase 1 trial in patients with multiple myeloma. 15 Moreover, blocking the interaction of S100A9 with extracellular matrix metalloproteinase inducer (EMMPRIN) controlled chronic lymphocytic leukemia progression and shifted leukemic cells toward a less inflammatory phenotype. 16
In this study, we investigated whether and how the S100A9‐mediated inflammatory activation of the myelodysplastic BM can be rescued by TASQ. We found that mesenchymal stromal cells (MSCs), a central component of the BM microenvironment, are susceptible to adopting an inflammatory phenotype, which causes dysregulation of HSPCs and may contribute to cytopenia in MDS patients. TASQ treatment of MSC cultures in vitro mitigated these effects as demonstrated by reduced expression levels of IL‐1β, IL‐18, caspase 1, and PD‐L1 resulting in improved hematopoietic support. Moreover, in NHD13 transgenic mice, TASQ significantly improved hemoglobin (Hb) levels and RBC counts, as well as bone microarchitecture, highlighting its potential as a therapeutic agent for LR‐MDS.
RESULTS
S100A9 is mainly secreted by neutrophils and macrophages in the human BM
Pro‐inflammatory signaling within the BM microenvironment mediated by alarmins, such as S100A9 among others, has been identified as a key pathogenetic driver of MDS. 3 To elucidate the differences in cellular composition and origins of S100A9 within the BM, we undertook a comprehensive approach utilizing multiplex immunofluorescence staining on biopsy sections obtained from patients with LR‐MDS and healthy donors. Through the expression of specific markers, including CD271 (MSCs), CD31 (endothelial cells), CD3 (T cells), CD66b (neutrophils), and CD68 (macrophages), we were able to discern the presence and distribution of these cell types in the BM. Intriguingly, S100A9 exhibited high expression in the MDS samples compared to those from healthy individuals (Figure 1A,B ). These findings underscore a distinctive signature within the MDS BM microenvironment.
Co‐expression analysis of multiplex immunofluorescence images revealed that the predominant producers and secretors of S100A9 were primarily neutrophils with significantly higher intensities in the MDS BM compared to the healthy controls (Figure 1C ). This highlights the pivotal role of these immune cells in the dysregulated expression of S100A9 observed in MDS. Importantly, CD271 + MSCs did not express S100A8 or S100A9, or the levels were below the detection limit.
S100A9 induces a pro‐inflammatory signature in MSCs, which can be reverted by TASQ in vitro
MSCs play a pivotal role within the BM environment, supporting both hematopoiesis and immunomodulation, as well as serving as precursors for bone‐forming cells. 17 , 18 Thus, we aimed to explore the impact of S100A9 on MSC biology and function. Given the potential for LR‐MDS MSCs to lose their inflammatory characteristics in vitro, we conducted experiments using recombinant S100A9 to induce inflammatory licensing in both healthy and MDS MSCs, although MDS BM‐derived cells might already be “pre‐primed.” Since the observed effects were more pronounced in MDS‐derived MSCs compared to healthy controls, subsequent data presentation focuses on MDS MSCs unless otherwise specified.
Initially, we investigated S100A9/TLR4‐mediated inflammasome activation by examining the expression and activation of downstream molecules through Western blot analysis. Exposure of MSCs to S100A9 induced TLR4 downstream signaling, as evidenced by increased expression of IRAK1 and NF‐κB‐p65. Notably, the S100A9 inhibitor TASQ reversed the effects and suppressed the expression of these proteins, indicating a reduction in inflammation (Figure 2A and Supporting Information S2: Figure 1 ). The expression of TLR4 remained unchanged.
Additionally, we assessed the mRNA expression levels of potential NF‐κB target genes in MSC cultures using quantitative real‐time polymerase chain reaction (PCR). Our findings revealed a significant increase in mRNA expression of the pro‐inflammatory cytokines IL‐1β and IL‐18, along with their activator caspase 1, in MDS MSCs following treatment with S100A9. Importantly, these effects were attenuated by TASQ (Figure 2B ).
We identified PD‐L1 as a potential downstream target of the TLR4/NF‐ κB signaling pathway in MSCs, where its expression was significantly upregulated by S100A9 stimulation, leading to an approximately 2.5‐fold increase compared to baseline levels (Figure 2B ). Notably, this induction of PD‐L1 expression could be effectively suppressed by TASQ treatment, reducing it to nearly 50% of the levels observed after S100A9 stimulation. A similar trend was observed at the protein level, where PD‐L1 expression was found to be inherently higher in MDS MSCs (Figure 2B,C , and Supporting Information S2: Figure 1 ).
These results demonstrate that S100A9 changes the immune‐modulatory function of MSCs and highlight its potential as a therapeutic target as well as the efficacy of TASQ in mitigating the effects.
MSC differentiation and clonogenic capacity are not significantly modulated by S100A9 and TASQ
The differentiation balance of MSCs is crucial for their hematopoietic support and for balancing bone homeostasis. Therefore, we differentiated cells from MDS patients and healthy donors in vitro for 14 days with S100A9/TASQ into the adipogenic and osteogenic lineages. Subsequently, the extent of differentiation was quantified based on the expression of adipogenic‐ and osteogenic‐specific genes, for example, PPARγ and Runx2. Although the enhanced adipogenic differentiation potential of MDS MSCs could be demonstrated by 26‐fold versus 19‐fold upregulation of expression of PPARγ compared to undifferentiated cells, neither S100A9 nor TASQ had a significant influence on these parameters (Supporting Information S3: Figure 2A ). Similarly, the osteogenic differentiation capacity of MDS MSCs was not affected by S100A9 or TASQ. Higher mRNA expression of Runx2, a marker for early stages of osteo‐chondro‐progenitor determination, was detected in MDS compared to healthy MSCs (Supporting Information S3: Figure 2B ).
Next, we compared the clonogenic potential of MDS and healthy MSCs, with and without TASQ treatment, by using colony‐forming unit (CFU)‐F assays. While MDS MSCs formed approximately twice as many colonies as healthy MSCs, treatment with TASQ had no significant effect on MDS MSCs (Supporting Information S3: Figure 2C ).
S100A9 impairs and TASQ restores the supportive function of MSCs toward hematopoiesis in vitro
To assess whether S100A9‐mediated NF‐κB signaling in MDS MSCs contributes to impaired hematopoietic support, we co‐cultured CD34 + HSPCs from healthy donors on MDS MSC monolayers pre‐treated with S100A9 or TASQ. Cobblestone area‐forming cell (CAF‐C) assays were performed weekly for up to 4 weeks to evaluate hematopoietic support. Notably, S100A9 pre‐treatment significantly reduced colony formation, whereas TASQ treatment effectively rescued colony numbers, with significant differences observed at Week 4 (Figure 3A,B ).
To further investigate the impact of MSC treatment on lineage‐specific progenitor differentiation, healthy HSPCs were co‐cultured on pre‐treated MDS MSC layers for 1 week, followed by an additional 2‐week colony‐forming assay. The colony numbers of erythroid (CFU‐E) and granulocyte‐macrophage (CFU‐GM) progenitors were compared between directly S100A9/TASQ‐treated HSPCs and those co‐cultured with MSCs pre‐treated or continuously exposed to S100A9 or TASQ. HSPCs co‐cultured with MSCs exhibited an overall increase in colony formation compared to directly treated cells. S100A9 treatment significantly reduced CFU‐E formation in both directly treated HSPCs and those co‐cultured with treated MSCs (Figure 3C ), while CFU‐GM numbers increased (Figure 3D ). Interestingly, TASQ treatment led to a significant increase in CFU‐E numbers exclusively in HSPCs co‐cultured with pre‐treated MSCs (Figure 3C ), underscoring the supportive role of MSCs during erythroid differentiation. Moreover, TASQ treatment significantly reduced CFU‐GM numbers in HSPCs co‐cultured with pre‐treated MSCs (Figure 3D ), suggesting a potential role in mitigating aberrant myeloid expansion associated with MDS.
TASQ modulation of MSCs enhances HSC erythroid differentiation
To further investigate the effects of TASQ on MSC‐mediated hematopietic stem cell (HSC) differentiation, healthy BM‐derived CD34 + cells were co‐cultured with MDS MSCs in the presence of erythropoietin (Epo), which is essential for the commitment of erythroid progenitors and their subsequent maturation into mature erythrocytes. To evaluate erythroid differentiation, the expression of CD71 (transferrin receptor), which facilitates iron uptake, and CD235a (glycophorin A), a marker for erythroid precursor and mature erythrocytes, was assessed by flow cytometry.
In the CD34 − /CD45 − fraction, TASQ treatment in the presence of Epo did not alter the percentage of CD71 + cells when HSCs were cultured alone. However, a significant increase in CD71 + cells was observed in MSC co‐culture–derived cells treated with Epo and TASQ (Figure 4A ). In contrast, CD235a expression remained lower, indicating an incomplete maturation of erythroid precursors (not shown). Moreover, the CD34 + /CD45 + fraction was significantly reduced in HSCs derived from TASQ‐treated MSC co‐cultures compared to untreated co‐cultures, to a similar extent as observed with Epo‐treated co‐cultures. However, neither Epo nor TASQ had a significant effect on this fraction in HSC monocultures (Figure 4B ). These findings highlight the potential role of TASQ in promoting erythroid differentiation through MDS MSC modulation.
TASQ modulation influences myeloid differentiation
To further explore the effects of TASQ on MDS MSC‐mediated myeloid differentiation, we assessed specific myeloid progenitor and monocyte subsets under different cytokine conditions.
The proportion of CD45 + /CD34 + /HLA‐DR + myeloid progenitor cells was significantly increased in co‐cultures with MDS MSCs compared to healthy HSCs cultured alone (Figure 4C ). Although not significant, TASQ treatment reduced this population by approximately 25% across all three differentiation conditions, but only in the presence of MSC co‐culture (Figure 4C ).
In further analyses of monocyte subset differentiation, CD14 + /CD16 − classical monocytes showed a strong increase in co‐culture under all conditions. However, TASQ treatment reduced their numbers by approximately 50% in G‐CSF‐ and M‐CSF‐driven differentiation, but only in co‐cultures (Figure 4D ). Similarly, CD14 + /CD16 + intermediate monocytes exhibited a strong increase in co‐culture, with TASQ treatment reducing their numbers by 25%–75% across all conditions, again only in co‐culture settings (Figure 4E ). Conversely, CD14 dim /CD16 + non‐classical monocytes were lower in co‐culture compared to monocultures, but TASQ treatment increased their proportion by 20%–50%, primarily in co‐culture conditions (Figure 4F ).
These findings suggest that TASQ influences HSPC differentiation in the presence of MSCs, consistent with a modulatory effect on niche‐associated signals. In this context, TASQ treatment is associated with enhanced erythroid differentiation and reduced myeloid expansion.
TASQ improves erythropoiesis in NHD13 mice in vivo
To investigate whether such effects can be observed in vivo, the NHD13 transgenic mouse model was used. The fusion protein in these mice causes a rather slow development of MDS toward high‐risk disease and/or AML. At the age of 12 weeks, NHD13 mice already showed mild anemia with decreased RBC levels (9.5 vs. 10.2 × 10 12 /L, Figure 5A ), prompting treatment initiation at this age. From 12 to 24 weeks, untreated NHD13 mice showed a consistent decline in Hb and RBC counts. Both parameters were strongly decreased in 24‐week‐old NHD13 mice compared to WT mice (Hb: 7.6 vs. 9.3 mmol/L; RBC: 7.1 vs. 10.2 × 10 12 /L). Of note, treatment with TASQ over 12 weeks significantly improved Hb levels to 8.4 mmol/L and RBC counts to 8.2 × 10 12 /L in NHD13 mice, while no significant effect was observed in WT mice (Figure 5A,B ).
The platelet count was slightly increased in TASQ‐treated NHD13 mice at 20 weeks with 673.8 × 10 9 /L compared to 556.8 × 10 9 /L in untreated animals, but this difference disappeared at 24 weeks (Supporting Information S4: Figure 3A ). The white blood cell (WBC) count was significantly lower in NHD13 than in WT mice but stayed at comparable levels throughout the experiment and was unaffected by TASQ treatment. In contrast, WT mice showed declining WBC levels, which were further decreased by TASQ (Supporting Information S4: Figure 3B ).
Given the significant improvement observed in peripheral RBC counts, we analyzed erythroid precursors in the BM to assess potential changes induced by TASQ treatment (Supporting Information S5: Figure 4A ). Notably, the CD45⁻Ter119 + population was reduced by 33% in NHD13 mice following TASQ administration (Figure 5C ), suggesting that TASQ promotes erythroid differentiation. Further analysis of the Ter119 + compartment revealed an altered distribution of erythroid precursors (Stages I–V), characterized by a significant reduction in late‐stage erythroid precursors (Stage IV) and a concomitant increase in reticulocytes (Stage V, Figure 5C ).
To investigate whether these effects were linked to changes in progenitor potential, we performed CFU assays using BM cells. Interestingly, we observed that all colonies of cells from NHD13 were smaller than those of WT mice, with significantly lower numbers of erythroid progenitors. TASQ treatment resulted in a reduction of erythroid colony formation in WT mice, but no significant effect was observed in NHD13 mice (Figure 5D ). These findings suggest that TASQ modulates erythroid differentiation but does not restore erythroid progenitor capacity in the NHD13 model.
Next, we investigated the potential effects of TASQ on the myeloid BM compartment in both NHD13 and WT mice. Flow cytometric analysis (Supporting Information S6: Figure 4B ) revealed that the frequency of CD11b + myeloid cells was significantly higher in untreated NHD13 mice compared to WT controls; however, TASQ treatment did not affect their abundance (Figure 5E ). In contrast, Ly6G + granulocytic cells were increased in TASQ‐treated NHD13 mice relative to TASQ‐treated WT mice (Figure 5F ). Additionally, Ly6C + monocytic cells were significantly elevated in NHD13 mice, and while TASQ treatment led to a reduction in Ly6C + cells in WT mice, it had no effect in NHD13 mice (Figure 5G ). Furthermore, the Ly6C + F4/80 + monocytic/macrophage population was expanded in NHD13 mice, with a slight but further increase following TASQ treatment, whereas no such effect was observed in WT mice (Figure 5H ). These findings suggest that TASQ differentially influences myeloid cell populations in the BM of NHD13 and WT mice.
Additionally, myeloid progenitors were assessed using CFU assays. While not reaching statistical significance, we observed a decrease in granulocyte‐macrophage (GM) colony formation in TASQ‐treated NHD13 mice (Figure 5I ), suggesting a potential suppressive effect of TASQ on myeloid progenitor activity.
Reduced osteoclast number and leads to bone gain in TASQ‐treated NHD13 mice
Since NHD13 mice showed impaired bone microarchitecture, 19 we investigated the potential effect of TASQ. Untreated NHD13 mice showed normal bone volume per total volume, but the trabecular number was reduced by 11%, indicating bone loss (Figure 6B ). TASQ treatment for 12 weeks increased the bone volume in both WT and NHD13 cohorts (WT and NHD13: twofold), accompanied by an increase in trabecular number (WT: +29%; NHD13: +32%; Figure 6A,B ). To evaluate the cause of changes in the bone remodeling, bone histomorphology of the femora and vertebrae was performed, and the bone turnover markers were assessed in the serum.
Surprisingly, the higher number of bone‐forming osteoblasts in NHD13 mice (+19%) and their activity, measured by procollagen Type 1 amino‐terminal propeptide (P1NP), was not increased by TASQ, and the latter was even reduced by 37% in WT mice (Figure 6C,D ). In addition, TASQ treatment led to a reduction in the bone formation rate (WT: −42%; NHD13: −51%) and osteoid surface per bone surface (WT: −66%; NHD13: −75%; Figure 6E,F ). Furthermore, a reduced number of osteoclasts (−42%) was detected in NHD13 mice (Figure 6G ). TASQ treatment led to a reduction in the number of osteoclasts of about 53% in WT mice only. However, the osteoclast activity was significantly decreased in both WT and NHD13 mice as measured by tartrate‐resistant acid phosphatase 5b (TRAcP5b) levels in the serum (WT: −20%; NHD13: −26%; Figure 6G,H ). Bone formation and resorption are reduced by TASQ. Thus, the osteoblast number was not affected the reduced bone resorption predominantly led to an increase in bone volume. Furthermore, a higher maximal force was detected in TASQ‐treated WT mice in the compression test of vertebrae (+60%), while NHD13 mice had already an elevated value (+35%), which was not affected by TASQ (Figure 6I ). Taken together, TASQ treatment enhances bone volume and strength in both WT and NHD13 mice by reducing bone turnover.
DISCUSSION
MDS is characterized by cytopenia and an increased risk of progression to AML. These features are driven in part by dysregulation of the BM microenvironment, shifting from a pro‐inflammatory state in early disease stages to an immunosuppressive milieu in later stages. 4 , 19 , 20 , 21 Current treatments are limited, highlighting the need for new therapies. Fromowitz et al. identified TLR4 and TLR6 as highly overexpressed in MDS, with TLR4 inhibition shown to promote erythroid and myeloid maturation in vitro. 22 The TLR4 ligands S100A8/A9 correlate with anemia severity, 23 suggesting a potential therapeutic target. We therefore conducted a comprehensive preclinical study to assess the effects of the S100A9‐inhibiting small molecule TASQ on the MDS BM, both in vitro and in vivo.
We and others identified neutrophils and, to a lesser extent, macrophages as the predominant source of S100A9 in the human BM. 23 , 24 While CD271 + MSCs do not appear to be directly involved in S100A9 production, they undergo significant modulation through the influence of extracellular S100A9, such as cellular senescence. 25 This process involves the induction of the TLR4 receptor downstream signaling, formation of the NLRP3 inflammasome, and subsequent secretion of IL‐1β. These findings suggest that elevated levels of S100A9 in the BM microenvironment could contribute to stromal cell aging and dysfunction. 25 On the other hand, transcriptome analysis of CD271 + niche cells in a prospective, homogeneously treated cohort of LR‐MDS patients revealed that expression of S100A8 and S100A9 was strongly correlated with a subgroup of MDS patients demonstrating significant overexpression of this heterodimer independent of established prognostic factors and activation of p53 and TLR programs in mesenchymal cells, in line with experimental data from a mouse model pointing at the existence of a p53‐S100A8/9‐TLR axis. 26 The controversial data on the relationship between S100A9 and MSCs underline the enormous heterogeneity of MSC subpopulations and complex regulatory mechanisms. Recent single‐cell analyses reveal the emergence of distinct, inflammation‐associated MSC subsets in clonal hematopoiesis and MDS that are absent in healthy BM and contribute to niche remodeling and impaired hematopoietic support, including altered expression of key stromal regulators such as CXCL12. 27
Other authors have described that stromal cells in the myelofibrotic BM were altered by the inflammatory environment leading to fibrosis‐driving cells with decreased potential to support normal hematopoiesis. 14
Notably, elevated S100A9 levels appear to be largely confined to the BM compartment. In MDS, S100A9 is produced locally by myeloid cells and acts predominantly as a paracrine or autocrine danger signal within the BM microenvironment, with limited spillover into the peripheral circulation. Consequently, systemic plasma S100A9 levels may not reliably reflect local inflammatory activity and niche dysfunction in the BM. 28
We demonstrated that exposure of both MDS and healthy MSCs to S100A9 triggered a TLR4‐dependent downstream signaling cascade in vitro, resulting in increased expression of IRAK1, NF‐kB‐p65, Caspase 1, and the pro‐inflammatory cytokines IL‐1β and IL‐18. TASQ treatment effectively dampened the S100A9‐induced inflammatory response of MSCs.
One mechanism to suppress inflammation and dampen the immune response is governed by the immune checkpoint receptor PD‐1 and its ligand PD‐L1. Cheng et al. reported that S100A9 contributes to ineffective hematopoiesis in MDS via induction of PD‐1 and PD‐L1 expression. 28 In this context, the increased death of CD34 + HSPCs and CD71 + erythroid progenitor cells plays a key role. In line with these data, we detected a significantly higher expression of PD‐L1 mRNA and protein in MSCs after S100A9 exposure, which was effectively suppressed by TASQ treatment. Moreover, effective downregulation of PD‐L1 in MDS‐derived MSCs has been associated with enhanced hematopoietic support, 28 a finding that was also reflected in our in vitro experiments.
Our results with HSPC/MSC co‐cultures align with previous reports indicating that S100A9 plays a critical role in MDS pathophysiology by promoting chronic inflammation and disrupting the BM microenvironment. 5 , 28 The observed reduction in CAF‐C numbers following S100A9 treatment supports prior evidence that TLR4/NF‐κB pathway activation in MSCs negatively impacts their hematopoietic support function. 29 , 30 This impairment may contribute to the ineffective hematopoiesis observed in MDS patients.
TASQ has been recognized for its immune‐modulating properties, particularly in cancer and inflammatory conditions. 12 Our results indicate that TASQ rescues hematopoietic support by MSCs, potentially by counteracting NF‐κB‐mediated dysfunction with potential involvement of PD‐L1 as an identified target in these conditions. These findings are consistent with previous studies demonstrating that modulating NF‐κB signaling in MSCs can restore their supportive role in hematopoiesis. 17 The lineage‐specific effects observed in our study further emphasize the impact of S100A9 on hematopoietic differentiation. The reduction in CFU‐E formation following S100A9 treatment is in line with previous reports that inflammatory signals suppress erythropoiesis. 31 Conversely, the increased CFU‐GM numbers suggest a shift toward myeloid expansion, a hallmark of MDS. 32 The ability of TASQ to restore CFU‐E numbers while reducing CFU‐GM highlights its potential role in rebalancing hematopoiesis.
Several studies have demonstrated that MSCs provide critical signals for lineage commitment and differentiation. 17 , 32 Our findings suggest that TASQ modulates MSCs in a way that promotes erythroid differentiation while reducing primitive HSC fractions, indicating a shift toward lineage commitment. The observed increase in CD71 + cells in TASQ‐treated MSC co‐cultures suggests an enhanced early erythroid commitment. This is consistent with prior reports indicating that additional erythropoietic factors and microenvironmental cues are required for complete erythroid maturation. 33
The observed reduction in CD34 + /CD45 + HSCs in TASQ‐treated co‐cultures suggests a shift away from self‐renewal and toward lineage commitment. Similar findings have been reported in studies where MSCs influence HSC fate through niche signaling. 34 This supports the notion that MSCs, when properly modulated, can act as key regulators of hematopoietic lineage specification. Importantly, the decrease in CD34 + progenitor frequency in our study aligns with reports showing that microenvironmental modulation can promote efficient erythroid differentiation while reducing the pool of undifferentiated progenitors. 35
In the NHD13 murine model, TASQ had a significant impact on RBC counts and clonogenic capacity, confirming our previous in vitro results. However, in the murine system, the regulatory mechanisms underlying these results may vary depending on the specific mouse model used. A detailed BM analysis revealed lower erythroid progenitor colonies in treated mice, which is in contrast to the CFU assays performed with isolated human MSCs and HSPCs demonstrating increased erythroid colony number after TASQ treatment. A potential alternative regulatory mechanism is that TASQ enhances erythroid maturation and survival, thereby increasing erythrocyte output from a smaller progenitor pool. Additionally, TASQ may suppress ineffective erythropoiesis by enhancing the differentiation efficiency of erythroid precursors or reducing apoptosis at later stages of maturation. 36 Although TASQ does not increase erythropoiesis to WT levels in NHD13 mice, it is evident that more severe anemia is prevented. A shift in lineage dynamics, such as accelerated progression through the CFU‐E stage, could also explain the reduced colony numbers despite increased RBC production.
The expansion of CD11b + myeloid cells and Ly6C + monocytic populations in untreated NHD13 mice is consistent with the myeloid bias and ineffective differentiation characteristic of MDS. 37 TASQ treatment did not reverse these abnormalities but selectively modulated specific subsets, such as increasing Ly6G + granulocytic cells and slightly expanding the Ly6C + F4/80 + population. The reduction in GM colony formation following TASQ treatment in NHD13 mice suggests a potential suppressive effect on myeloid progenitor activity, which may be linked to its differentiation‐promoting effects on the erythroid lineage. Prior studies have highlighted the complex interplay between erythroid and myeloid differentiation, with skewed lineage commitment being a hallmark of MDS. 38 The differential effects of TASQ in WT versus NHD13 mice suggest it modulates progenitor activity, but its impact is context‐dependent. Further studies are warranted to elucidate the mechanistic basis of these effects.
TASQ treatment not only modulated hematopoiesis by the BM microenvironment but also improved bone microarchitecture, even though the bone formation rate is lower. An explanation could be the changes of the osteoid surface and the number of osteoclasts. The osteoid surface is reduced, although the number of osteoblasts and their early activity marker P1NP are normal. This could be due to delayed osteoid formation or faster bone mineralization. In the latter case, the bone formation rate should increase, but this is not the case with TASQ. Therefore, we assume that TASQ affects late osteoblasts. Measuring serum osteocalcin is a suitable way to assess the activity of late osteoblasts. While bone formation is indeed reduced with TASQ, bone resorption is even more strongly inhibited, leading to increased bone volume.
This result is supported by our human in vitro co‐culture of MSCs and HSCs differentiated to monocytes, the precursors of osteoclasts. Since pro‐inflammatory cytokines such as IL‐6, IL‐1β, and TNF‐α enhance bone resorption, the inhibitory effects of TASQ likely contribute to bone preservation and the mitigation of bone loss in MDS patients. 39 An increase in trabecular bone volume with TASQ has also been described in multiple myeloma, but the authors did not mention whether the osteoclasts are responsible for the effect. 40
While our findings provide novel insights into the effects of TASQ on hematopoiesis and bone microarchitecture, the study has potential limitations. First, although the inhibition of the S100A9/TLR4/NF‐κB axis appears to play a relevant role, the precise mechanistic pathways through which TASQ influences MSC‐mediated hematopoiesis remain unclear. Second, the long‐term functional consequences of TASQ treatment on immune homeostasis and erythropoiesis remain to be determined. While we observed significant changes in erythroid and myeloid differentiation, it is unclear whether these effects persist over time and may even have the potential to delay disease progression. Longitudinal studies are warranted to assess whether prolonged TASQ treatment alters immune function, clonal evolution, or overall hematopoietic stability. Finally, although NHD13 mice exhibit key features of LR‐MDS and are a widely accepted murine model for the disease, they are only partly reflective of human MDS pathophysiology. In particular, differences in inflammasome activation, disease progression, and inflammatory mediators may limit the direct translational applicability of our findings. Additionally, the timing and duration of TASQ treatment could significantly impact its effects, necessitating further investigations to optimize treatment strategies and better align with human MDS pathophysiology.
In summary, our data provide compelling evidence that TASQ can attenuate pathological S100A9/TLR4/NF‐κB‐mediated inflammasome activation in the myelodysplastic BM niche resulting in improved hematopoietic support by MSCs in vitro (Figure 7 ). These findings highlight the potential of TASQ as a therapeutic agent for MDS patients, who often suffer from concomitant anemia and bone loss. Of note, the therapeutic effects of TASQ in MDS are not solely attributed to remodeling of the BM niche, as cell‐intrinsic effects have also been described by other groups. Notably, Schneider et al. identified significant expression of S100A8/9 in mutant erythroblasts, leading to a p53‐mediated defect in erythroid differentiation, which could be reversed by genetic inactivation of S100A8. 24 However, our study is the first to elucidate the niche‐specific effects of TASQ in MDS and their broader impact on the disease phenotype. Given the limited success of various anti‐inflammatory therapies in the past (e.g., Canakinumab trial, IRAK4 inhibitors), we propose that an S100A9‐targeted anti‐inflammatory approach, which intervenes more centrally in the inflammatory cascade, presents a promising therapeutic strategy. Further investigation, particularly in clinical trials, is warranted to fully evaluate the therapeutic potential of TASQ in MDS.
MATERIALS AND METHODS
Patients
MSCs were collected from hematologically healthy donors undergoing hip replacement ( n = 4, age 63–82 years), BM stem cell donors ( n = 5, age 23–44), and untreated patients with MDS ( n = 12, age 41–80 years, patient's characteristics are depicted in Supporting Information S7: Table 1 ) under written informed consent (according to the Declaration of Helsinki) as part of the MDS registry (EK289112008) and the BoHemE Study ( NCT02867085 ) at the University Hospital in Dresden. Heparinized BM samples were obtained at diagnostic aspiration or during hip total arthroplasty surgery.
Human BM immunofluorescence staining
To analyze the BM using immunofluorescence, 3–5 μm thick paraffin sections were used. Deparaffinization, rehydration, and epitope retrieval were done manually via an established protocol. Briefly, sections were deparaffinized in xylene (2 × 15 min, VWR International, Fontenay‐sous‐Bois, France) and hydrated by washes of graded ethanol (Berkel AHK, Ludwigshafen, Germany) to water (B. Braun, Melsungen, Germany). Tissue sections were boiled in TRIS‐HCL buffer (Akoya Biosciences, Menlo Park, California, USA) at pH 9.0 for 20 min for antigen retrieval. Multiplex‐immunohistochemistry of BM sections was performed via cyclic immunofluorescence staining on MACSima imaging platform (Miltenyi Biotec, Bergisch‐Gladbach, Germany). The procedure is based on cyclic immunofluorescence staining. Every sequence includes the following: staining the tissue with fluorophore‐coupled antibodies (Supporting Information S8: Table 2 ), imaging the region of interest (ROI), and eliminating the fluorescence signal by bleaching. The staining is performed with commercially available antibodies that are directly linked to a fluorophore. Three dyes and DAPI are measured in one cycle with minimal spectral overlap. The staining cycles are conducted in the fully automated system of the MACSima Imaging Platform. Analysis software MACSiQView was used for image data evaluation, and nuclei segmentation was performed with a StarDist 2D plugin in ImageJ. 41
Cell culture
BM mononuclear cells were collected using a Ficoll gradient and cultured at 37°C and 5% CO 2 in Dulbecco's modified Eagle's medium (DMEM, Invitrogen) supplemented with 10% fetal calf serum (FCS). After 24 h, non‐adherent cells were removed, and adherent cells were expanded. Primary MSCs from Passages 2–5 were used for the experiments. Sub‐confluent cultures were treated with 1.5 µg/mL S100A9 (Abcam) and 10 µM TASQ (provided by Active Biotech) for up to 1 week. Cell growth and proliferation were calculated by cell counting with trypan blue.
HSPCs were isolated using CD34 antibody‐conjugated magnetic beads, according to the manufacturer's instructions (Miltenyi Biotec). The purity of the isolated CD34 + population was confirmed by flow cytometry, and only fractions with >95% positivity were used in the analysis.
Clonogenic assays
CAF‐C assays were performed over 4 weeks using pre‐treated healthy or MDS MSC layers. One thousand magnetically isolated CD34 + cells were added in LTC‐IC medium (Stem Cell Technologies) supplemented with 1 × 10 −6 M hydrocortisone (Sigma Aldrich).
CFU assays were carried out using cells harvested after 1 week of co‐culture, with 1 × 10 4 cells being plated in enriched methylcellulose medium with recombinant cytokines (MethoCult H4435, Stem Cell Technologies). Colonies were counted after 2 weeks and classified under a microscope or with the StemVision system (Stem Cell Technologies).
Mice
For the in vivo experiments, 3‐month‐old heterozygous NHD13 (C57BL/6‐Tg[Vav1‐NUP98/HOXD13]G2Apla/J) and their littermate wild‐type (WT) mice were used. The NUP98‐HOXD13 (NHD13) fusion gene has been detected in patients with MDS, and mice expressing the NHD13 fusion gene under the vav1 promotor develop MDS‐like symptoms and BM alterations. 42 , 43
All mouse experiments were approved by the institutional animal care committee as well as the Federal State of Saxony (TVV 64/2022). Mice were fed a standard diet as well as water ad libitum, and they were exposed to a 12‐h light/dark cycle in an air‐conditioned room at 23°C. To analyze the effects of TASQ on hematopoiesis and bone homeostasis, NHD13 and WT mice were treated with normal water (control) or TASQ water (30 mg/kg per body weight) for 12 weeks. The blood count was monitored at the baseline and every 4 weeks until the end of the experiment using the XN‐1000 (Sysmex, Norderstedt, Germany). All mice received an intraperitoneal calcein injection (20 mg/kg body weight) 2 and 5 days before sacrifice. Finally, long bones, spine, BM, and serum were collected for subsequent analysis. One femur and the spine per mouse were fixed in 4% PFA for 48 h and stored in 50% ethanol.
The NUP98 gene encodes a protein that is involved in RNA and protein transport across the nuclear membrane. 43 Fusion genes with NUP98 have been identified in various hematologic malignancies, including MDS.
Flow cytometry
Flow cytometry was applied to phenotypically characterize HSPCs after isolation and co‐culture experiments. The used antibodies are depicted in Supporting Information S10: Table 4 . Furthermore, the BM of NHD13 and WT mice was flushed with phosphate‐buffered saline (PBS) supplemented with 5% FCS (Biochrom) as well as 5 mM EDTA, and cells were stained with an antibody cocktail consisting of anti‐murine antibodies (Supporting Information S10: Table 4 ). Data were evaluated using FlowJo 10.10.0 software.
Mouse BM serum analysis
Concentrations of the bone turnover markers P1NP and TRAcP5b were measured in serum using ELISA according to the manufacturers' protocols (IDS, Frankfurt/Main, Germany).
Bone microarchitecture
For analysis of the bone microarchitecture, μCT (vivaCT40, Scanco Medical) was performed on excised femora with an isotropic voxel size of 10.5 μm (70 kVP; 114 μA; 200 ms integration time). We used established protocols from Scanco Medical to evaluate the trabecular bone parameters and analyzed 100 slices beneath the growth plate.
Bone histomorphometry and bone strength
The lumbar vertebrae (L1–5) were isolated for further analysis. L1/L2 were embedded in paraffin, and 2 µm sections were used for TRAP staining to determine the number of osteoclasts, osteoblasts, and osteocytes per bone perimeter. Von Kossa/van Gieson staining was performed on 4 µm methyl methacrylate section of L3/L4 to evaluate bone mineralization. In addition, 7 µm sections of L3/L4 were used to assess the distance between the two calcein labeling lines by fluorescence microscopy. All analyses were performed with the Osteomeasure software, and pictures were taken with the CellSens program or the AxioVision 4.8 program.
To determine the vertebral stiffness, first, L5 was transferred to PBS 1 day before compression testing (Zwick Roell, Ulm, Germany). The L5 was placed onto the lower plate, and the mechanical force was applied vertically onto the vertebra via the upper plate. The maximal load (Fmax) was measured, and the bone stiffness was calculated using testXpert II—V3.7 software (Zwick Roell, Ulm, Germany).
Statistical analysis
Statistical analysis was performed using GraphPad Prism software version 9.5.0 (GraphPad Software). Data are presented as mean ± standard deviation (SD). All experiments were repeated at least three times. For multiple group comparisons, two‐way analysis of variance (ANOVA) with post hoc Tukey's, Sidak's, or Bonferroni test was performed. A P‐value of less than 0.05 was regarded as being statistically significant.
Additional methods
Additional methods are described under Supporting Information S12 .
AUTHOR CONTRIBUTIONS
Manja Wobus : Conceptualization; methodology; writing—original draft; funding acquisition; visualization; project administration; writing—review and editing; formal analysis; validation. Heike Weidner : Conceptualization; methodology; investigation; writing—original draft; writing—review and editing; visualization; formal analysis; validation. Rebekka Wehner : Methodology; investigation; writing—review and editing; visualization; formal analysis. Anna‐Lena Baumann : Writing—review and editing; methodology; investigation; formal analysis. Kristin Möbus : Methodology; visualization; writing—review and editing; formal analysis. Ekaterina Balaian : Conceptualization; writing—review and editing; methodology; formal analysis; validation. Marie Törngren : Conceptualization; writing—review and editing; resources. Erik Vahtola : Conceptualization; writing—review and editing; resources. Helena Eriksson : Conceptualization; writing—review and editing; resources. Susann Winter : Writing—review and editing; methodology. Uwe Platzbecker : Supervision; writing—review and editing. Triantafyllos Chavakis : Writing—review and editing; supervision. Lorenz Christian Hofbauer : Supervision; writing—review and editing; resources. Martina Rauner : Writing—review and editing; methodology. Martin Bornhäuser : Writing—review and editing; supervision; conceptualization; resources. Katja Sockel : Project administration; conceptualization; methodology; funding acquisition; writing—original draft; writing—review and editing; resources.
CONFLICT OF INTEREST STATEMENT
Manja Wobus and Katja Sockel received research funding from Active Biotech AB, Sweden.
ETHICS STATEMENT
Donor and patient samples were obtained after written informed consent according to the Declaration of Helsinki as part of the MDS registry (EK289112008) and the BoHemE Study (EK393092016) at the University Hospital in Dresden.
All mouse experiments were approved by the institutional animal care committee as well as the Federal State of Saxony (TVV 64/2022).
FUNDING
This work was supported by a research grant from Active Biotech (Lund, Sweden) to M.W. and K.S. K.S. and E.B. are supported by the German Federal Ministry of Research, Technology and Space (BMFTR) (CAMINO, 01EO2101). Open Access funding enabled and organized by Projekt DEAL.
Supporting information
Supporting Information.
Supporting Information.
Supporting Information.
Supporting Information.
Supporting Information.
Supporting Information.
Supporting Information.
Supporting Information.
Supporting Information.
Supporting Information.
Supporting Information.
Supporting Information.
ACKNOWLEDGMENTS
The authors would like to thank Katrin Müller, Ivonne Habermann, Claudia Richter, Claudia Dill, and Robert Kuhnert for excellent technical support. Tasquinimod was generously provided by Active Biotech AB, Sweden.
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.
REFERENCES
1 Platzbecker U , Hofbauer LC , Ehninger G , Hölig K . The clinical, quality of life, and economic consequences of chronic anemia and transfusion support in patients with myelodysplastic syndromes . Leuk Res . 2012 ; 36 ( 5 ): 525 ‐ 536 . 10.1016/J.LEUKRES.2012.01.006 22300879
2 Cluzeau T , McGraw KL , Irvine B , et al. Pro‐inflammatory proteins S100A9 and tumor necrosis factor‐α suppress erythropoietin elaboration in myelodysplastic syndromes . Haematologica . 2017 ; 102 ( 12 ): 2015 ‐ 2020 . 10.3324/haematol.2016.158857 28983059 PMC5709100
3 Sallman DA , List A . The central role of inflammatory signaling in the pathogenesis of myelodysplastic syndromes . Blood . 2019 ; 133 ( 10 ): 1039 ‐ 1048 . 10.1182/BLOOD-2018-10-844654 30670444 PMC7022316
4 Winter S , Shoaie S , Kordasti S , Platzbecker U . Integrating the “immunome” in the stratification of myelodysplastic syndromes and future clinical trial design . J Clin Oncol . 2020 ; 38 ( 15 ): 1723 ‐ 1735 . 10.1200/JCO.19.01823 32058844
5 Chen X , Eksioglu EA , Zhou J , et al. Induction of myelodysplasia by myeloid‐derived suppressor cells . J Clin Invest . 2013 ; 123 ( 11 ): 4595 ‐ 4611 . 10.1172/JCI67580 24216507 PMC3809779
6 Ehrchen JM , Sunderkötter C , Foell D , Vogl T , Roth J . The endogenous Toll‐like receptor 4 agonist S100A8/S100A9 (calprotectin) as innate amplifier of infection, autoimmunity, and cancer . J Leukoc Biol . 2009 ; 86 ( 3 ): 557 ‐ 566 . 10.1189/jlb.1008647 19451397
7 Osanto S , van Poppel H , Burggraaf J . Tasquinimod: a novel drug in advanced prostate cancer . Future Oncol . 2013 ; 9 ( 9 ): 1271 ‐ 1281 . 10.2217/fon.13.136 23980674
8 Källberg E , Vogl T , Liberg D , et al. S100A9 interaction with TLR4 promotes tumor growth . PLoS One . 2012 ; 7 ( 3 ): e34207 . 10.1371/journal.pone.0034207 22470535 PMC3314596
9 Raymond E , Dalgleish A , Damber J‐E , Smith M , Pili R . Mechanisms of action of tasquinimod on the tumour microenvironment . Cancer Chemother Pharmacol . 2014 ; 73 ( 1 ): 1 ‐ 8 . 10.1007/s00280-013-2321-8 24162378 PMC3889691
10 Isaacs JT , Antony L , Dalrymple SL , et al. Tasquinimod is an allosteric modulator of HDAC4 survival signaling within the compromised cancer microenvironment . Cancer Res . 2013 ; 73 ( 4 ): 1386 ‐ 1399 . 10.1158/0008-5472.CAN-12-2730 23149916 PMC3578133
11 Sternberg C , Armstrong A , Pili R , et al. Randomized, double‐blind, placebo‐controlled phase III study of tasquinimod in men with metastatic castration‐resistant prostate cancer . J Clin Oncol . 2016 ; 34 ( 22 ): 2636 ‐ 2643 . 10.1200/JCO.2016.66.9697 27298414
12 Shen L , Pili R . Tasquinimod targets suppressive myeloid cells in the tumor microenvironment . Oncoimmunology . 2019 ; 8 ( 10 ): e1072672 . 10.1080/2162402X.2015.1072672 31646064 PMC6791429
13 Leimkühler NB , Gleitz HFE , Ronghui L , et al. Heterogeneous bone‐marrow stromal progenitors drive myelofibrosis via a druggable alarmin axis . Cell Stem Cell . 2021 ; 28 ( 4 ): 637 ‐ 652.e8 . 10.1016/j.stem.2020.11.004 33301706 PMC8024900
14 Gleitz HFE , Fuchs SNR , Snoeren IAM , et al. Inhibiting the alarmin‐driven hematopoiesis‐stromal cell crosstalk in primary myelofibrosis ameliorates bone marrow fibrosis . HemaSphere . 2025 ; 9 ( 8 ): e70179 . 10.1002/hem3.70179 40823315 PMC12351185
15 Vogl DT , Nefedova Y , Wileyto EP , et al. Tasquinimod, an S100A9 inhibitor, in combination with ixazomib, lenalidomide, and dexamethasone for relapsed and refractory multiple myeloma . Blood . 2024 ; 144 ( Suppl 1 ): 7060 . 10.1182/blood-2024-204155
16 Uriepero‐Palma A , Márquez ME , Payque E , et al. Targeting S100A9‐mediated inflammation: a novel therapeutic approach for CLL . Blood Adv . 2025 ; 9 ( 20 ): 5219 ‐ 5233 . 10.1182/bloodadvances.2025016061 40720748 PMC12550715
17 Kfoury Y , Scadden DT . Mesenchymal cell contributions to the stem cell niche . Cell Stem Cell . 2015 ; 16 ( 3 ): 239 ‐ 253 . 10.1016/j.stem.2015.02.019 25748931
18 Pittenger MF , Discher DE , Péault BM , Phinney DG , Hare JM , Caplan AI . Mesenchymal stem cell perspective: cell biology to clinical progress . NPJ Regen Med . 2019 ; 4 : 22 . 10.1038/s41536-019-0083-6 31815001 PMC6889290
19 Weidner H , Wobus M , Hofbauer LC , Rauner M , Platzbecker U . Luspatercept mitigates bone loss driven by myelodysplastic neoplasms and estrogen‐deficiency in mice . Leukemia . 2022 ; 36 : 2715 ‐ 2718 . 10.1038/s41375-022-01702-1 36175549 PMC9613459
20 Tentori CA , Zhao LP , Tinterri B , et al. Immune‐monitoring of myelodysplastic neoplasms: recommendations from the i4MDS consortium . HemaSphere . 2024 ; 8 ( 5 ): e64 . 10.1002/hem3.64 38756352 PMC11096644
21 Balaian E , Wobus M , Bornhäuser M , Chavakis T , Sockel K . Myelodysplastic syndromes and metabolism . Int J Mol Sci . 2021 ; 22 ( 20 ): 11250 . 10.3390/ijms222011250 34681910 PMC8541058
22 Fromowitz A , Aluri S , Gonzalez Lugo JD , et al. Inhibition of Toll‐like receptor 4 (TLR4) promotes erythroid and myeloid differentiation in myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML) . Blood . 2023 ; 142 : 5615 . 10.1182/blood-2023-182059
23 Li X , Li Q , Xiang X , Zhang X , Wu Y . The diagnostic value and clinical correlations of bone marrow supernatant S100A8 and S100A9 in myelodysplastic neoplasms . Cytokine . 2025 ; 187 : 156856 . 10.1016/j.cyto.2025.156856 39799746
24 Schneider RK , Schenone M , Ferreira MV , et al. Rps14 haploinsufficiency causes a block in erythroid differentiation mediated by S100A8 and S100A9 . Nat Med . 2016 ; 22 ( 3 ): 288 ‐ 297 . 10.1038/nm.4047 26878232 PMC4870050
25 Shi L , Zhao Y , Fei C , et al. Cellular senescence induced by S100A9 in mesenchymal stromal cells through NLRP3 inflammasome activation . Aging . 2019 ; 11 ( 21 ): 9626 ‐ 9642 . 10.18632/aging.102409 31727865 PMC6874461
26 Zambetti NA , Ping Z , Chen S , et al. Mesenchymal inflammation drives genotoxic stress in hematopoietic stem cells and predicts disease evolution in human pre‐leukemia . Cell Stem Cell . 2016 ; 19 ( 5 ): 613 ‐ 627 . 10.1016/j.stem.2016.08.021 27666011
27 Prummel KD , Woods K , Kholmatov M , et al. Inflammatory stromal and T cells mediate human bone marrow niche remodeling in clonal hematopoiesis and myelodysplasia . Nat Commun . 2025 ; 16 ( 1 ): 10042 . 10.1038/s41467-025-65803-y 41253811 PMC12627676
28 Cheng P , Eksioglu EA , Chen X , et al. S100A9‐induced overexpression of PD‐1/PD‐L1 contributes to ineffective hematopoiesis in myelodysplastic syndromes . Leukemia . 2019 ; 33 ( 8 ): 2034 ‐ 2046 . 10.1038/S41375-019-0397-9 30737486 PMC6687540
29 Yang L , Qian Y , Eksioglu E , Epling‐Burnette PK , Wei S . The inflammatory microenvironment in MDS . Cell Mol Life Sci . 2015 ; 72 ( 10 ): 1959 ‐ 1966 . 10.1007/s00018-015-1846-x 25662443 PMC11113192
30 Meisel R , Brockers S , Heseler K , et al. Human but not murine multipotent mesenchymal stromal cells exhibit broad‐spectrum antimicrobial effector function mediated by indoleamine 2,3‐dioxygenase . Leukemia . 2011 ; 25 ( 4 ): 648 ‐ 654 . 10.1038/leu.2010.310 21242993
31 Morceau F , Dicato M , Diederich M . Pro‐inflammatory cytokine‐mediated anemia: regarding molecular mechanisms of erythropoiesis . Mediators Inflamm . 2009 ; 2009 : 405016 . 10.1155/2009/405016 20204172 PMC2830572
32 Gañán‐Gómez I , Wei Y , Starczynowski DT , et al. Deregulation of innate immune and inflammatory signaling in myelodysplastic syndromes . Leukemia . 2015 ; 29 ( 7 ): 1458 ‐ 1469 . 10.1038/leu.2015.69 25761935 PMC4857136
33 Dzierzak E , Philipsen S . Erythropoiesis: development and differentiation . Cold Spring Harbor Perspect Med . 2013 ; 3 ( 4 ): a011601 . 10.1101/cshperspect.a011601 PMC3684002 23545573
34 Méndez‐Ferrer S , Michurina TV , Ferraro F , et al. Mesenchymal and haematopoietic stem cells form a unique bone marrow niche . Nature . 2010 ; 466 ( 7308 ): 829 ‐ 834 . 10.1038/nature09262 20703299 PMC3146551
35 Gori JL , Butler JM , Chan Y‐Y , et al. Vascular niche promotes hematopoietic multipotent progenitor formation from pluripotent stem cells . J Clin Invest . 2015 ; 125 ( 3 ): 1243 ‐ 1254 . 10.1172/JCI79328 25664855 PMC4362238
36 Balderman SR , Li AJ , Hoffman CM , et al. Targeting of the bone marrow microenvironment improves outcome in a murine model of myelodysplastic syndrome . Blood . 2016 ; 127 ( 5 ): 616 ‐ 625 . 10.1182/blood-2015-06-653113 26637787 PMC4742549
37 Ding L , Saunders TL , Enikolopov G , Morrison SJ . Endothelial and perivascular cells maintain haematopoietic stem cells . Nature . 2012 ; 481 ( 7382 ): 457 ‐ 462 . 10.1038/nature10783 22281595 PMC3270376
38 Pellagatti A , Cazzola M , Giagounidis A , et al. Deregulated gene expression pathways in myelodysplastic syndrome hematopoietic stem cells . Leukemia . 2010 ; 24 ( 4 ): 756 ‐ 764 . 10.1038/leu.2010.31 20220779
39 Takayanagi H . Osteoimmunology: shared mechanisms and crosstalk between the immune and bone systems . Nat Rev Immunol . 2007 ; 7 ( 4 ): 292 ‐ 304 . 10.1038/nri2062 17380158
40 Fan R , Satilmis H , Vandewalle N , et al. Tasquinimod suppresses tumor cell growth and bone resorption by targeting immunosuppressive myeloid cells and inhibiting c‐MYC expression in multiple myeloma . J Immunother Cancer . 2023 ; 11 ( 1 ): e005319 . 10.1136/jitc-2022-005319 36650020 PMC9853259
41 Schmidt U , Weigert M , Broaddus C , Myers G . Cell detection with star‐convex polygons . In: Frangi AF , Schnabel JA , Davatzikos C , Alberola‐López C , Fichtinger G , editors. BT‐medical image computing and computer assisted intervention—MICCAI 2018 . Cham : Springer International Publishing ; 2018 : 265 ‐ 273 .
42 Weidner H , Rauner M , Trautmann F , et al. Myelodysplastic syndromes and bone loss in mice and men . Leukemia . 2017 ; 31 ( 4 ): 1003 ‐ 1007 . 10.1038/LEU.2017.7 28074069
43 Lin YW , Slape C , Zhang Z , Aplan PD . NUP98‐HOXD13 transgenic mice develop a highly penetrant, severe myelodysplastic syndrome that progresses to acute leukemia . Blood . 2005 ; 106 ( 1 ): 287 ‐ 295 . 10.1182/blood-2004-12-4794 15755899 PMC1201424