Longitudinal Localization of Leukemia Stem Cells Between Metaphysis and Central Marrow Governs Leukemic Stem Cell Behavior

Longitudinal Localization of Leukemia Stem Cells Between Metaphysis and Central Marrow Governs Leukemic Stem Cell Behavior | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Longitudinal Localization of Leukemia Stem Cells Between Metaphysis and Central Marrow Governs Leukemic Stem Cell Behavior Xunlei kang, Chen Wang, Yi Pan, Ruochen Dong, WenXuan Zhou, XiaDuo Meng, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6515437/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 24 Apr, 2026 Read the published version in Nature Cell Biology → Version 1 posted You are reading this latest preprint version Abstract Leukemic stem cells (LSCs) reside in protective bone marrow (BM) niches, evading chemotherapy and driving relapse. Disrupting LSC-niche represents a promising therapeutic strategy, yet effective approach remains elusive. Our study characterized longitudinal BM niches in support of LSC survival and proliferation, i.e. metaphysis vs. central marrow, compared to well-accepted endosteum vs. sinusoid niches. Quiescent LSCs mostly localize to metaphysis and show reduced stemness and aggressiveness when mobilized to the central marrow, the composition of both endosteum and sinusoid in the central marrow. Building on this, we developed an approach to restrict LSCs within the BM and induce LSC apoptosis by targeting DPP4 in AML cells. Genetic deletion of Dpp4 in AML cells alters CXCL12 gradient across three scales: 1) System-wide : A reversed CXCL12 gradient between the BM and peripheral blood confines AML cells within the BM, limiting their circulation. 2) BM level : Perturbation of the CXCL12 gradient between the metaphysis and central marrow mobilizes LSCs out of their protective metaphysis niche, leading to exhaustion in the sinusoidal region, despite the higher CXCL12 level in the sinusoidal region. 3) Microscale within the metaphysis : Loss of the CXCL12 gradient between N-cadherin + mesenchymal stromal cells and the surrounding matrix impairs LSC recruitment to N-cad + cells, further driving their exhaustion. These alterations stem from the CXCL12-DPP4-GPC3 axis. DPP4, highly expressed by AML cells, deactivates CXCL12, while GPC3, enriched in N-cad + cells, inhibits DPP4's enzymatic activity. This axis establishes a favorable CXCL12 gradient that attracts LSCs to N-cad + cell-rich niches in the metaphysis and facilitates their dissemination via circulation. These findings highlight the therapeutic potential of targeting CXCL12-DPP4-GPC3 axis to disrupt LSC niches and enhance AML treatment. Health sciences/Diseases/Cancer/Cancer microenvironment Health sciences/Diseases/Cancer/Cancer stem cells Biological sciences/Cancer/Cancer microenvironment Biological sciences/Cancer/Cancer stem cells Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Introduction Acute myeloid leukemia (AML) is an aggressive hematopoietic malignancy driven by leukemic stem cells (LSCs), which are central to disease initiation, progression, and relapse. Thus, targeting LSCs has emerged as a promising strategy to combat AML effectively. LSC behavior is governed by two critical factors: their trafficking through circulation and their localization within specialized bone marrow (BM) niches. However, despite extensive studies on LSC-niche interactions and significant efforts to translate these insights into clinical therapies 1–3 , no effective treatment has been developed to restrict leukemia dissemination or eradicate LSCs. This highlights an incomplete understanding of the complex interplay between LSCs and their niches. Previous research has extensively characterized how niche cells regulate LSCs through single-direction communication 3,4 . However, the dynamic cross talk between LSCs and niche cells in co-creating a leukemia-supportive microenvironment remains poorly understood. Additionally, two BM niches have been well-described in regulating LSC behavior: the endosteal niche, which promotes LSC quiescence and survival, and the sinusoidal niche, which facilitates LSC proliferation and dissemination 5 . The endosteal niche comprise two anatomically distinct regions: the metaphysis (trabecular bone) and the diaphysis (cortical bone). While interactions between LSCs and the diaphyseal endosteum at a local transverse scale (diaphysis vs. sinusoid) have been well-studied, the broader longitudinal scale (metaphysis vs. sinusoid) remains underexplored. CXCL12 is a well-established chemoattractant for leukemic cells 6,7 , playing a crucial role in their migration and localization. DPP4, a key regulator of CXCL12 activity 8,9 , has been implicated in promoting AML progression 10 . While GPC3 is recognized as an inhibitor of DPP4 11,12 , its specific role in leukemia development remains unexplored. In this study, we identified the metaphysis and central marrow as two key LSC niche areas, emphasizing the crucial role of N-cadherin-expressing mesenchymal stromal/stem cells (N-cad + MSCs) 13–15 in the metaphysis for maintaining LSCs. Furthermore, we uncovered a three-way crosstalk between the BM matrix, N-cad + MSCs, and LSCs through the CXCL12-DPP4-GPC3 axis that governs LSC behavior. Targeting DPP4 mobilized LSCs from the metaphysis to the central marrow niche, disrupting their protective microenvironment, inducing exhaustion, and limiting their dissemination into peripheral blood. This therapeutic approach significantly enhances the efficacy of targeting LSCs, providing a novel avenue for AML treatment. Results Dpp4 deficiency re-distributes AML cells and induces LSC exhaustion Our group recently identified DPP4 as a promising therapeutic target for AML due to its selectively high expression in AML cells and its role in blocking AML progression upon depletion 10 . Strikingly, our experiments revealed that Dpp4 deletion alters the spatial distribution of AML cells within the BM, as observed in whole BM sections (Fig. 1 a-b). In control AML cells transplanted mice ( Dpp4 +/+ AML mice; 2 weeks post-transplantation), AML cells predominantly localize around the metaphysis area, especially the proximal metaphysis (PM). However, in Dpp4 KO AML cells transplanted mice ( Dpp4 −/− AML mice; 12 weeks post-transplantation), AML cells exhibited a more even distribution throughout the BM, with increased centralization towards the central marrow (CM) in both MLL-AF9 (MA9) and AML-ETO9a (AE9) models (Fig. 1 a-b). To eliminate potential bias due to differences in AML progression since Dpp4 KO AML mice rarely reach moribund stage and sustain minimal AML cells in peripheral blood (PB), we transplanted 1x10 6 control or Dpp4 KO primary AML cells (GFP + ) into C57/B6 recipients and compared AML distribution within the BM when PB GFP + cells reach ~ 30% (Fig. 1 c). Notably, this approach controlled for tumor burden, ruling out the possibility that redistribution was secondary to delayed AML progression in Dpp4 KO mice. Consistently, Dpp4 KO resulted in a similar redistribution of AML cells across the BM. This redistribution correlated with a significant increase in LSC populations (L-GMPs: GFP + IL-7R − Lin − Sca-1 − c-Kit + CD34 + FcγRII/III + ) in the BM of MA9 model (Fig. 1 d-f). Additionally, L-GMPs in Dpp4 KO mice exhibited increased cell division (Fig. 1 g) and higher apoptosis (Fig. 1 h), suggesting an exhaustion phenotype. The PM (proximal metaphysis) and DM (distal metaphysis) mark distinct niches versus CM for maintaining LSCs While we observed that L-GMPs exhibit reduced quiescence and increased exhaustion-like behavior after Dpp4 KO, it is not clear that whether such phenomenon results from the intracellular effects of Dpp4 KO or is associated with altered LSC niche. To understand this, we evaluated L-GMPs across the three anatomical partitions (PM, DM, and CM) of the BM in control AML mice. Our analysis revealed that the PM and DM harbored a significantly higher percentage of L-GMPs compared to the CM (Fig. 2 a-b). Among the L-GMPs, in each niche, those derived from the PM and DM displayed lower cell division activity (Fig. 2 c-d). To functionally assess LSC from PM/DM vs CM niches, we transplanted L-GMPs from the PM, DM or CM into C57/B6 recipient, and found those from PM and DM induced a more aggressive AML progression (Fig. 2 e). Furthermore, limiting dilution assays show PM/DM-derived AML cells exhibited higher stemness (Fig. 2 f). These findings strongly suggest that the PM and DM, as opposed to the CM, constitute specialized LSC-supporting niches that promote LSC maintenance and functionality. Dpp4 deficiency confines AML cells within the BM through reversed BM-to-PB CXCL12 gradient Another hallmark finding in Dpp4 KO AML mice is the dramatic decrease in peripheral AML cells and an enrichment of AML cells within the BM after Dpp4 deletion in both MA9 and AE9 models (Fig. 3 a). To investigate whether Dpp4 KO AML cells exhibit impaired homing or show reduced peripheralization despite successful homing, we compared the homing efficiency of control and Dpp4 KO AML cells 16 hours post-transplantation. Our findings indicate comparable homing of control and Dpp4 KO AML cells to all hematopoietic organs (Fig. 3 b), suggesting that Dpp4 deletion does not impair homing but instead restricts AML cells within the host BM. This restriction persists even in the late stage of AML development (Fig. 3 a) and resulted in prolonged mouse survival (Fig. 3 c). However, this finding challenges the existing knowledge that the CM, rich in sinusoids, serves as a crucial niche for LSCs and hematopoietic stem cells (HSCs) trafficking to the PB 4,5 . To explore the mechanism behind this intriguing observation, we next focused on MA9 AML model. We first hypothesized that Dpp4 KO might impair AML cells' access to blood vessels. However, immunofluorescence analysis revealed that both GFP-labeled control and Dpp4 KO AML cells were equidistant from endomucin-labeled vessels throughout the BM, indicating unimpeded vascular access (Fig. 3 d-f) 16,17 . This prompted us to hypothesize that altered cytokine / chemokine gradients might be responsible for the observed AML cell confinement. We focused on DPP4 substrates and cytokines or chemokines implicated in AML cell trafficking 18–22 . Among the factors tested, including CCL11, CCL22, CXCL12, IL-6, IL-15, etc., only CXCL12 exhibited a reversed gradient from plasma to BM in Dpp4 KO mice (Fig. 3 g-j). Despite prior reports implicating G-CSF and NPY in HSPC mobilization 23–25 , their levels were comparable between BM extracellular fluid (BMEF) and plasma in our model, excluding their role in the observed phenotype. Exogenous CXCL12 administration mobilized both control and Dpp4 KO AML cells at high doses (Fig. 3 k), but Dpp4 KO cells showed tolerance at lower doses. This was not due to altered CXCR4 expression (Fig. 3 l), or migration capacity ( Fig. 3 m-n ) , suggesting that the reversed gradient itself—not cell-intrinsic changes—drives confinement. Niche-specific CXCL12 production by N-cad ⁺ MSCs dictates LSC spatial distribution Our previous findings demonstrated that a reversed BM-to-PB CXCL12 gradient confines Dpp4 KO AML cells to the BM. To dissect the mechanism driving AML redistribution, we quantified CXCL12 levels across BM subregions. Strikingly, Dpp4 KO mice exhibited a reversed PM/DM-CM CXCL12 gradient, with low CXCL12 in the CM of controls but elevated levels in Dpp4 KO mice (Fig. 4 a–b). This suggested compartmentalized CXCL12 regulation by DPP4, prompting us to identify its cellular source. To identify the specific niche cells secreting CXCL12, we performed single-cell RNA sequencing (scRNA-seq) on non-hematopoietic PM and DM cells from AML mice three weeks post-transplantation. Through the analysis of 11,512 cells (median of 19,367.5 molecules and 4,391 genes per cell), we identified 14 distinct clusters, spanning mesenchymal stem/stromal cells (MSC), osteolineage cells (OLC), chondrocytes (Chondro) and chondrocytes of possible transitional states (c/r, cycling/resting; pro, progenitor), fibroblasts (Fibro), endothelial cells (EC), pericytes (Fig. 4 c). Consistent with prior reports, Cxcl12 was predominantly expressed by the MSC cluster (Fig. 4 d) 1 . Given the heterogeneity of MSCs, we compared Cxcl12 expression across known subpopulations marked by LepR, N-cadherin ( Cdh2 ), Prx-1, Osx ( Sp7 ), and Nestin 13,26–31 . Transcriptomic association analysis revealed that Cxcl12 was most strongly correlated with Lepr⁺ and Cdh2⁺ MSCs (Fig. 4 e). LepR + MSCs are known as a heterogenous population of MSCs spanning across the BM, including metaphysis (but mainly perivascular localization in metaphysis 28 ) and CM region. The N-cad protein is predominantly detected in the endosteum of metaphysis 14 . In contrast, other MSC subpopulations ( Prx-1 + , Osx + , and Nes + MSCs), were minimally represented in our dataset (Fig. 4 f), aligning with prior studies showing Prx-1 + MSCs enriched in periosteal region 32,33 , Osx + MSCS enrich in the bone tissue 34 , and Nes + MSCs enriched in perivascular region 31 . Critically, despite prior reports implicating Nestin + MSCs in AML chemoresistance 35 their Cxcl12 expression was negligible ( Fig. 4 f ) To rule out functional redundancy, we deleted Cxcl12 in Nestin + cells ( Nestin-CreER; Cxcl12 fl/fl ), and observed no impact on AML progression or BM retention (Supplementary Fig. 1a-b) , confirming Nestin + MSCs are dispensable for CXCL12-mediated LSC maintenance in our model. N-cad⁺ MSCs co-expressed Cxcl12 with other LSC-supportive factors (e.g., Gas6 , Angpt1 , Kitl ; Fig. 4 g) 4,36–38 . Given reports that SCF ( Kitl) promotes AML adhesion and survival 39 , we rigorously tested its role. Conditional deletion of Scf in N-cad + cells ( N-cad-CreER; Scf fl/fl ) yielded no difference in mice survival, AML cell engraftment and distribution in the BM ( Supplementary Fig. 2a–e ). This negative result underscores that the observed niche effects are CXCL12-specific and not confounded by SCF. To spatially assess the relationship between LSCs and N-cad + cells, we transplanted control L-GMPs into N-cad-tdTomato (N-cad-TdT) reporter mice, where Tomato + cells mark N-cadherin expression. AML cells (GFP⁺) and N-cad⁺ stromal cells were co-enriched in the proximal metaphysis (Fig. 4 h). Strikingly, LSCs (GFP⁺ Kit⁺) were significantly more likely to reside within 5 µm of N-cad⁺ cells (45.3% ± 5.4%) compared to differentiated AML cells (GFP⁺ Kit⁻; 10.6% ± 2.2%) (Fig. 4 h, bottom right ), suggesting active niche-LSC crosstalk. To test whether N-cad + -derived CXCL12 drives LSC retention, we conditionally deleted Cxcl12 in N-cad + cells ( N-cad-CreER; Cxcl12 fl/fl ) (Fig. 4 i) and transplanted control L-GMPs (Fig. 4 j). N-cad; Cxcl12 −/ − mice recapitulated the Dpp4 KO phenotype, with AML cells redistributed from PM/DM to CM (Fig. 4 k-l). Shared LSC behavior and transcriptomic Signatures in N-cad; Cxcl12 −/− and Dpp4 KO AML mice The redistribution of LSCs from PM/DM to CM in N-cad; Cxcl12 −/− mice mirrored the phenotype observed in Dpp4 KO AML, prompting us to investigate whether these models shared LSC properties and underlying molecular mechanisms. CM-localized AML cells from N-cad; Cxcl12 −/− mice exhibited hallmark features of LSC exhaustion, including increased cell cycle activity and apoptosis, alongside impaired self-renewal capacity (Fig. 5 a–d). These phenotypic changes were accompanied by delayed AML progression, as evidenced by prolonged survival and reduced organ infiltration (Fig. 5 e-f). To determine whether the shared spatial redistribution phenotype reflected convergent molecular reprogramming, we performed transcriptomic analyses. Principal component analysis (PCA) revealed striking similarity between Dpp4 KO AML cells and control AML cells from N-cad; Cxcl12 −/− recipients, with both populations clustering distinctly from their respective controls (Fig. 5 g). RNA sequencing further demonstrated consistent transcriptional changes, with upregulation of cell cycle and metabolic genes and downregulation of stemness-associated pathways and migration (Fig. 5 h–i). Together, these findings demonstrate that disruption of the CXCL12-DPP4 axis induces convergent transcriptomic rewiring that drives LSC exhaustion and impairs disease progression. N-cad + cells support LSCs through GPC3 mediated attraction While we have shown that DPP4, by deactivating CXCL12, specifically create an intra-BM CXCL12 gradient that favors LSC localization to the PM/DM for maintenance, two key questions remain unanswered: 1) Why does DPP4 selectively deactivate CXCL12 in the CM while preserving CXCL12 in the PM and DM in control AML mice (Fig. 4 b)? 2) How do control LSCs achieve close proximity to N-cad + niche cells (Fig. 4 h)? Prior studies have identified CXCL12 hotspots as critical HSC niches that attract HSCs to close proximity for their maintenance 40 . Consistent with this, our scRNA-seq analysis revealed that N-cad + cells express high levels of Cxcl12 (Fig. 4 e-f). To further investigate, we compared the spatial relationship between LSC-enriched populations (GFP + Kit + ) and N-cad + cells in control and Dpp4 KO AML mice. Interestingly, Dpp4 KO LSCs resided significantly farther from N-cad + cells than control LSCs, with only 10.5% ± 2.1% of Dpp4 KO LSCs located within 5 µm of N-cad + cells, compared to 58.6% ± 7.3% of control LSCs (Fig. 4 h, 6 a-b). Additionally, in the PM/DM, Dpp4 KO L-GMPs exhibited exhaustion-like behavior (Fig. 6 c), underscoring the importance of close proximity to N-cad + cells for LSC maintenance. As such, we hypothesized that an N-cad + cell-derived factor interacts with DPP4 to establish the CXCL12 gradient at both the macroscopic (PM/DM-CM) and the microscale level. To test this, we conducted transcriptional profiling of 10 known CXCL12 or DPP4 regulators (e.g., DPP8, elastase, MMPs, cathepsin G, TFPI, and GPC3) 11,41,42 in both AML cells and N-cad + cells from the PM/DM. Notably, Glypican-3 (GPC3) was highly expressed in N-cad + cells compared to other factors (Fig. 6 d-e). GPC3 is known as an inhibitor of DPP4, suggesting its potential role in suppressing DPP4 activity and preserving CXCL12 near N-cad + cells. Indeed, in vitro experiments demonstrated that GPC3 binds to DPP4 on AML cells and inhibits DPP4 enzymatic activity (Fig. 6 f-g). Immunostaining further revealed significantly higher GPC3 expression in N-cad + cells (79.8% ± 3.3%) compared to N-cad − cells (10.5% ± 2.1%) in the PM region (Fig. 6 h-i), with 73.5% of GPC3 and DPP4 co-localized between N-cad + cells and LSCs (Fig. 6 j). These data indicate that GPC3, expressed by N-cad + cells, inhibits DPP4 and sustains CXCL12 at the microscale, thereby attracting LSCs to CXCL12 hotspots for their maintenance. Additionally, at the macroscopic level, GPC3-mediated DPP4 inhibition contributes to the intra-BM CXCL12 gradient between the PM/DM and CM, further facilitating LSC localization to supportive niche environments. Gpc3 knockout in N-cad + redistributes AML cells and impairs LSCs To functionally assess the role of the GPC3-DPP4 interaction in LSC maintenance, we conditionally deleted Gpc3 from N-cad + cells (N-cad; Gpc3 −/− ) and transplanted LSCs into both N-cad; Gpc3 −/− and control mice (N-cad; Gpc3 +/+ ) (Fig. 7 a). Loss of GPC3 in N-cad + cells led to a significant reduction in CXCL12 levels in the PM and DM (Fig. 7 b), resulting in an altered AML cell distribution pattern similar to that observed in Dpp4 KO and N-cad; Cxcl12 −/− mice (Fig. 6 c). This redistribution was accompanied by LSC exhaustion, as evidenced by increased L-GMPs division and apoptosis (Fig. 7 d-e). The disruption of the CXCL12 gradient and loss of proximity to supportive N-cad + cells impaired LSC maintenance, ultimately prolonging mouse survival (Fig. 7 f), confirming the role of GPC3 in regulating LSC localization. Discussion Our study identified two pivotal findings: the redistribution of LSCs, leading to their exhaustion, and the confinement of AML cells within the BM. Both phenomena are fundamentally mediated by CXCL12, a key chemoattractant for AML cells (Fig. 8 ) 6,7 . While CXCL12's role in LSC biology has been controversial 43–46 , our data reconcile these discrepancies by demonstrating that spatial compartmentalization—not absolute CXCL12 levels—governs LSC fate. Previous in vitro studies suggest that CXCL12 supports LSC survival, either independently or via MSCs 45,47,48 , while in vivo evidence remained conflicting. For instance, studies utilizing mouse models have shown that knocking out Cxcl12 in Prx-expressing MSCs or Tek-expressing endothelial cells had no significant effect on LSC survival in MLL-AF9-driven AML models 43 . Conversely, other work has demonstrated that global or conditional deletion of Cxcl12 in chronic myeloid leukemia (CML) models enhanced survival and chemotherapy efficacy 2,49 . Our work resolves this paradox by demonstrating that the functional impact of CXCL12 depends critically on its anatomical source and spatial gradient within the BM niche. A key advance of our study is the identification of N-cadherin + MSCs as the functionally relevant CXCL12 source for LSC maintenance. This contrasts with previous work focusing on Prx + periosteal 32,33 or Tek + endothelial niches 43 , which our scRNA-seq revealed are minimally involved in metaphyseal LSC maintenance (Fig. 4 f). Other well-known MSC markers, such as Lepr, Nestin, and Osx, are either widespread expression across the BM 28 , enriches in perivascular region 31 , or localized in the bone tissue 34 . Importantly, we confirmed through genetic deletion that neither Nestin + MSC-derived CXCL12 (despite their reported role in chemoresistance) nor N-cad + MSC-derived SCF contribute meaningfully to the phenotypes we observed ( Supplementary Figs. 1–2 ), underscoring the specificity of the N-cadherin + MSC-CXCL12 axis in our model. The discovery of a reverse PM/DM-CM CXCL12 gradient in control AML mice provides a mechanistic basis for LSC niche specificity. This gradient depends on regional regulation of DPP4 activity: while DPP4 degrades CXCL12 in the CM, GPC3 from N-cad + MSCs inhibits DPP4 in the PM/DM, creating protected CXCL12 microdomains. This explains why prior studies using high-resolution mapping in normal hematopoiesis failed to detect long-range CXCL12 gradients 40 - the AML-specific expression of DPP4 creates a pathological gradient not present in steady state 50 . Collectively, our work identifies the CXCL12-DPP4-GPC3 axis as a master regulator of LSC niche interactions, where: GPC3 from N-cad + MSCs locally preserves CXCL12 to anchor LSCs in protective metaphyseal niches, DPP4 from AML cells degrades CXCL12 systemically to enable dissemination. This axis offers immediate clinical promise, as DPP4 inhibitors could exploit this mechanism: (1) by reversing the BM-PB CXCL12 gradient to confine AML cells, reducing life-threatening extramedullary complications—such as leukocytosis, clotting abnormalities, respiratory distress, and stroke—thereby reducing leukemia-associated morbidity 51,52 ; (2) by disrupting protective niche interactions to sensitize LSCs to chemotherapy; and (3) through direct induction of LSC exhaustion. The established safety profile of DPP4 inhibitors (e.g., sitagliptin) underscores the translational potential of this approach 10,53 . Important remaining questions include whether other DPP4 substrates contribute to these effects, and how metabolic changes in redistributed LSCs influence their exhaustion. Nevertheless, our work establishes the CXCL12-DPP4-GPC3 axis as a critical regulator of LSC spatial organization and viability, providing a strong rationale for niche-targeted therapies in AML (Fig. 8 ). Methods Mice C57BL/6 mice were obtained from Charles River, Inc. DPP4 flox/flox mice were generated by breeding targeted C57Bl/6NTac-DPP4tm1a Wtsi/Ics mice (European Mouse Mutant Cell Repository, EUCOMM) with 129S4/Bl6-Gt (ROSA) 26Sortm2(FLP*) Sor/J (stock #012930, The Jackson Laboratory, Bar Harbor, ME). The offspring were further crossed with Vav-iCre mice (stock #018968, The Jackson Laboratory, Bar Harbor, ME) 54 , to generate DPP4 fl/fl ；Vav-Cre mice 10 . N-cad-tdTomato (N-cad-TdT) and N-cad-CreER strains were generated by Dr. Linheng Li’s lab 13,15 . Cxcl12 fl/fl were purchased from Jackson Lab. To induce expression of Cre-ER recombinase, mice received tamoxifen via intraperitoneal injection (Sigma, 75 mg tamoxifen/kg body weight) as described 55 . All mouse strains used in this study had a C57BL/6J genetic background. Animals were randomly assigned to experimental groups based on genotyping results. Investigators were blinded to group allocation during data analysis but not during experimental procedures. Sample sizes for each experiment are detailed in figure legends. All animal procedures were conducted in accordance with protocols approved by the Institutional Animal Care and Use Committee (IACUC) of the University of Missouri. AML transplantation For Fig. 1a, we transplanted infected Dpp4 KO or control Lin − cells (2 x 10 5 ; infection efficiency consistently 40-50%) into lethally irradiated (1,000 cGy) C57/B6 mice (6–8 weeks old) and evaluated 2-50 weeks after transplantation. For other Dpp4 KO vs control AML models, 1 × 10 6 primary AML cells were transplanted C57/B6 recipients. For homing assay, GFP + AML cells were evaluated at 16 hours. Engraftment was assessed 2- and 4-weeks post-transplantation for MLL-AF9 and AML-ETO9a models, respectively. For control AML cells transplantation into N-cad + mice with or without Cxcl12 or Gpc3 , 500 L-GMPs were transplanted into sublethally irradiated (500cGy) mice and engraftment assessed at 2-8 weeks post-transplantation. To preserve niche cell integrity, all mice were sub-lethally irradiated (500 cGy) before transplantation. Flow cytometry Peripheral blood (PB), bone marrow (BM), spleen, and liver hematopoietic cells were labeled with the following antibodies (all from BioLegend unless otherwise noted): Anti-CD3e-PE/Cyanine5 (#100310, 1:200), Anti-Ly-6G/Ly-6C (Gr-1)-PE/Cyanine5 (#108410, 1:200), Anti-CD11b-PE/Cyanine5 (#101210, 1:200), Anti-CD45R-PE/Cyanine5 (#103210, 1:200), Anti-Ter-119-PE/Cyanine5 (#116210, 1:200), Anti-CD117 (c-Kit)-APC (#105812, 1:200), Anti-Sca-1-PE-Cy7 (#108114, 1:200), Anti-CD150-PE (#115904, 1:200), Anti-CD48-APC/Cyanine7 (#103432, 1:200), Anti-Ki67-FITC (#652410, 1:200), Hoechst 34580 (BD Pharmingen, #565877), Anti-CD16/32-PE (#101308, 1:200), Anti-CD34-FITC (eBioscience, #11-0341-82, 1:200), Anti-CD127-APC/Cyanine7 (#135040, 1:200), Anti-CD135-Brilliant Violet421 (#135314, 1:200), Annexin V (#640941, 1:20), and PI (#421301, 1:50). Intracellular staining was performed using the Foxp3/Transcription Factor Staining Kit (eBioscience) according to the manufacturer’s protocol. Flow cytometry analyses were performed independently in triplicate, with biological replicates from at least five mice per condition. Technical replicates were included for measurement accuracy. Immunofluorescence staining and quantification Femurs were perfused with PBS, fixed with 4% paraformaldehyde, and subjected to frozen sectioning. Antigen retrieval was performed with 1 μg/ml proteinase K in TE buffer (100 mM Tris-HCl, pH 8.0, 50 mM EDTA) at 37°C for 30 minutes. Sections were blocked with Universal Blocking Reagent (BioGenex) and incubated overnight at 4°C with primary antibodies, including anti-Endomucin, biotin-lineage antibody mixture (BioLegend, 1:200), IL-7R, Sca-1 (BioLegend, 1:200), PE-CD34 (BioLegend, 1:200), Alexa Fluor 647-Kit (Invitrogen, 1:200), and anti-GPC3. Secondary staining was performed with donkey anti-goat Alexa Flour® 555 (Invitrogen; 1:500), goat anti-rabbit Alexa Fluor® 750 (Invitrogen; 1:500) and Brilliant violet 421® Streptavidin (Biolegend; 1:500) at room temperature for 1 hour. DAPI stock solution was diluted to 300 nM in PBS and 300 mL was added to the coverslip preparation for 1 minute. Sections were rinsed 3 times in PBS, excess buffer drained from the coverslip and mounted with Shandon™ Immu-Mount™ (Fisher Scientific). Image stitching was done to capture the entire specimen at high magnification and seamlessly create a single high-resolution image. Sections were imaged using a Keyence BZ-X800 fluorescence microscope at 20× magnification (resulting in 200× magnification) and 60× magnification (resulting in 600× magnification). Quantification was performed using Keyence BZ-X800 analyzer software, assessing GFP + AML cell distribution across BM areas (PM, CM, DM), L-GMP localization, and apoptosis. Distances between GFP + AML cells and N-cad+ cells were measured using a minimum of 100 GFP + Kit + and 80 GFP + Kit − AML cells per dataset. A minimum of five mice per condition were used for each quantification dataset. Cytokine analyses Cytokine quantification in plasma and bone marrow extracellular fluid (BMEF) were determined using the LEGENDplex Multi-Analyte Flow Assay Kit (BioLegend, San Diego CA), a bead-based immunoassay that quantifies multiple cytokines simultaneously via flow cytometry. Briefly, a custom mouse cytokine and chemokine panel was employed to measure the concentrations of the designed cytokines/chemokines. The LSRFortessa X-20 Cell Analyzer (BD Biosciences) was used for data acquisition, and results were analyzed using the LEGENDplex Data Analysis software. Assays were performed in 96-well plates following manufacturer protocols, with data recorded using a Fisherbrand microplate photometer. BMEF and plasma were used for ELISA assay collected by using the mouse SDF-1 alpha ELISA Kit (Invitrogen) following the manufacturer’s protocol. Migration assay In vitro: The Transwell migration assay was utilized to assess cell migration. DPP4 +/+ and DPP4 −/− AML cells were cultured and seeded in serum-free medium into the upper chamber of Transwell inserts (Corning, Inc.) with an 8 μm pore-size Matrigel-coated membrane. The upper wells contained CXCL12 at concentrations of 0 ng/ml. The lower chambers were filled with culture medium containing 100 ng/ml CXCL12. After a 4-hour incubation at 37°C and 5% CO2, non-migratory cells on the upper membrane surface were removed using a cotton swab. Migratory cells on the lower membrane surface were visualized and quantified in multiple random fields under a microscope. Migration rates and statistical significance were analyzed accordingly. In vivo: Mice (n=5) were intravenously injected with PBS or CXCL12 (0-500 ng/g per mouse as indicated in figure). The percentage of GFP + AML cells in peripheral blood was measured before and 17 hours after injection. Colony assays Mouse AML cells were diluted to the indicated concentration in IMDM with 2% FBS and were then seeded into methylcellulose medium M3534 (STEMCELL Technologies, Cambridge MA) for myeloid colony formation analysis, as previously described 56 . Each colony-forming unit (CFU) assay was performed in triplicate using independent biological replicates for AML cells obtained from distinct mice. The consistency of clone formation was validated across all replicates. DPP4 activity assay DPP4 activity was measured in plasma-EDTA, BMEF, and cell lysates. For each assay, 20 μl of serum or BMEF, or 100 nM of GPC3 protein, was diluted in DPP4 assay buffer (Tris-HCl [pH 8.0], 150 mM NaCl, and protease inhibitor cocktail) in a black 96-well plate to a final volume of 50 μl. An equal volume (50 μl) of 200 mM H-Ala-Pro-AFC substrate (I-1680; Bachem Americas, Torrance, CA) was added, and the plate was incubated for 10 minutes at room temperature in the dark. Fluorescence was measured using a Synergy Microplate Reader at excitation/emission wavelengths of 405/535 nm, and results were reported as relative light units (RLUs). Ligand binding assay Recombinant His-tagged GPC3 binding to DPP4+/+ AML cells was assessed as similarly described 57 . Briefly, 1 × 10 6 Dpp4 +/+ AML cells were incubated with or without His-GPC3 (100 nM) in 200 ml PBS/1% BSA for 3 h at 25°C. Nonspecific binding was subtracted. After incubation, cells were washed twice by centrifugation, resuspended in ice-cold PBS/1% BSA, and stained with Alexa Fluor® 488 anti-His Tag Antibody for flow cytometry analysis. Quantitative RT-PCR RT-PCR was performed using 5 ng total RNA, gene-specific primers, and a QIAGEN One Step RT-PCR kit (210210; Qiagen, Germantown, MD) following the manufacturer’s instructions. 18s rRNA was used as an internal control for normalization. The primer sequences used are listed below: TFPI: forward(5′-GGG CTC CGT TCT TGG TCT C-3′) and reverse(5′- TTG AAT CTG CGG CAC TTT TGC-3′), MMP9: forward(5′- CTG GAC AGC CAG ACA CTA AAG-3′) and reverse (5′- CTC GCG GCA AGT CTT CAG AG-3′), MMP2: forward(5′- CAA GTT CCC CGG CGA TGT C -3′) and reverse (5′- TTC TGG TCA AGG TCA CCT GTC-3′), MMP3: forward(5′- ACA TGG AGA CTT TGT CCC TTT TG-3′) and reverse (5′- TTG GCT GAG TGG TAG AGT CCC -3′), MMP13: forward(5′- CTT CTT CTT GTT GAG CTG GAC TC -3′) and reverse (5′- CTG TGG AGG TCA CTG TAG ACT-3′), MMP14: forward(5′- CAG TAT GGC TAC CTA CCT CCA G-3′) and reverse (5′- GCC TTG CCT GTC ACT TGT AAA-3′), cathepsin G: forward(5′- AGG GTT TCT GGT GCG AGA AG-3′) and reverse (5′- GTT CTG CGG ATT GTA ATC AGG AT-3′), Elastase: forward(5′- AGC AGT CCA TTG TGT GAA CGG-3′) and reverse (5′- CAC AGC CTC CTC GGA TGA AG-3′), DPP8: forward (5′- GGG AAA TGG TGA ATC ACA GGA C-3′) and reverse (5′- ATG TAG CCG TGG TAT TTT CTG G-3′), GPC3: forward (5′- CAG CCC GGA CTC AAA TGG G -3′) and reverse (5′- CAG CCG TGC TGT TAG TTG GTA -3′). Cell preparation for single-cell RNA-sequencing Tissue harvesting: Femurs were collected post-euthanasia and immediately placed in ice-cold PBS. The PM and DM regions of the bone were isolated. Bone marrow was extracted by crushing the bones with a mortar and pestle, followed by enzymatic digestion with collagenase/dispase at 37°C for 45 minutes. The resulting cell suspension was washed, lysed, and filtered through a 100 μm strainer. FACS isolation of non-hematopoietic cells: Bone marrow cells were stained for CD45, and viable triple-negative cells were sorted using a BD FACSAria™. Single cell sequencing Single cells were encapsulated into emulsion droplets using Chromium Controller (10x Genomics). scRNA-seq libraries were constructed using Chromium Single Cell 3’ v2 Reagent Kit according to the manufacturer’s protocol. Briefly, post sorting sample volume was decreased, and cells were examined under a microscope and counted with a hemocytometer. Cells were then loaded in each channel with a target output of ~4,000 cells. Reverse transcription and library preparation were performed on C1000 Touch Thermal cycler with 96-Deep Well Reaction Module (Bio-Rad). Amplified cDNA and final libraries were evaluated on an Agilent Bioanalyzer using a High Sensitivity DNA Kit (Agilent Technologies). Individual libraries were diluted to 4nM and pooled for sequencing. Pools were sequenced with 75 cycle run kits (26bp Read1, 8bp Index1 and 55bp Read2) on the Novaseq 5000 Sequencing System (Illumina). Single cell RNA-seq analysis Single-cell expression was analyzed using the Cell Ranger Single Cell Software Suite (v3.0.2) to perform quality control, sample demultiplexing, barcode processing, and single-cell 3′ gene counting. Sequencing reads were aligned to the refdata-cellranger-mm39-3.0.0 transcriptome using the Cell Ranger suite with default parameters. Single cell data were imported into Seurat 3 package in R statistical software. Only genes detected in at least two cells were considered for further analysis. Apoptotic cells were filtered out by any cell containing > 5% mitochondrial UMI counts. To detect changes in gene signatures more robustly, any cell with <1,500 genes was filtered out. Cell expression levels were normalized and scaled using default settings in Seurat. Principal component analysis for dimensionality reduction was then run on the normalized gene–barcode matrix. TSNE embedding was applied to visualize the cells in the two-dimensional space after selecting the first 20 principal components for the clustering analysis with a resolution parameter of 0.5. Cluster-specific genes were identified by running the Seurat “FindMarkers/FindAllMarkers” function with the Wilcoxon rank-sum test. Plots were made using the Seurat package and R. Quantification and statistical analysis Data are expressed as mean±SE mean (SEM). Statistical analyses were performed using GraphPad Prism Version 9.0 (Graph Pad Prism Software Inc, San Diego, CA). For continuous variables, normality and homogeneity of variance were assessed by the Shapiro-Wilk and Brown-Forsythe tests, respectively. After confirming homogeneous variances and normality, 2-group comparisons for means were performed using the 2-sided Student t test, and multi-group comparisons for means were performed using 2-way ANOVA with Holm-Sidak multiple comparison test. For data that did not pass either normality or equal variance test, 2-group comparisons were performed using Mann-Whitney Rank-sum test, and multi-group comparisons were performed using the Kruskal-Wallis 1-way ANOVA on ranks test with Dunn post hoc test. P<0.05 was considered statistically significant. Randomization and blinded analyses were performed whenever possible. Declarations Acknowledgments We thank the University of Missouri Genomics Core for their support in generating data on the scRNA-seq. Funding Research reported in this publication was supported by the National Cancer Institute of the National Institutes of Health under Award Number R37CA241603 and the American Cancer Society under Award Number RSG-23-1152630. Author Contributions X.K. conceived and supervised the study; C.W. performed experiments and analyzed data; Y.P., C.W., W.Z., and X.M. conceived the experiments; C.W., W.Z. performed scRNA-seq and bulk RNA-seq. Y.P. performed bioinformatics analysis. C.W. performed immunofluorescence staining and imaging data analysis. C.W., Y.P., W.Z. and X.M. verified the reproducibility of results; R.D.H. contributes to BM imaging data analysis; L.L., R.D., R.N. provided technical assistance and contribute to data analysis. Y.P., C.W. and X.K. wrote the original draft. Declaration of interest : The authors declare no competing interests. References Baryawno, N. et al. A Cellular Taxonomy of the Bone Marrow Stroma in Homeostasis and Leukemia. Cell 177 , 1915-1932 e1916 (2019). https://doi.org/10.1016/j.cell.2019.04.040 Agarwal, P. et al. Mesenchymal Niche-Specific Expression of Cxcl12 Controls Quiescence of Treatment-Resistant Leukemia Stem Cells. Cell Stem Cell 24 , 769-784 e766 (2019). https://doi.org/10.1016/j.stem.2019.02.018 Grockowiak, E. et al. Different niches for stem cells carrying the same oncogenic driver affect pathogenesis and therapy response in myeloproliferative neoplasms. Nat Cancer 4 , 1193-1209 (2023). https://doi.org/10.1038/s43018-023-00607-x Mendez-Ferrer, S. et al. Bone marrow niches in haematological malignancies. Nat Rev Cancer 20 , 285-298 (2020). https://doi.org/10.1038/s41568-020-0245-2 Morrison, S. J. & Scadden, D. T. The bone marrow niche for haematopoietic stem cells. Nature 505 , 327-334 (2014). https://doi.org/10.1038/nature12984 Aiuti, A., Webb, I. J., Bleul, C., Springer, T. & Gutierrez-Ramos, J. C. The chemokine SDF-1 is a chemoattractant for human CD34+ hematopoietic progenitor cells and provides a new mechanism to explain the mobilization of CD34+ progenitors to peripheral blood. J Exp Med 185 , 111-120 (1997). https://doi.org/10.1084/jem.185.1.111 Wright, D. E., Bowman, E. P., Wagers, A. J., Butcher, E. C. & Weissman, I. L. Hematopoietic stem cells are uniquely selective in their migratory response to chemokines. The Journal of experimental medicine 195 , 1145-1154 (2002). https://doi.org/10.1084/jem.20011284 Hoffmann, T., Faust, J., Neubert, K. & Ansorge, S. Dipeptidyl peptidase IV (CD 26) and aminopeptidase N (CD 13) catalyzed hydrolysis of cytokines and peptides with N-terminal cytokine sequences. FEBS letters 336 , 61-64 (1993). Christopherson, K. W., 2nd, Hangoc, G., Mantel, C. R. & Broxmeyer, H. E. Modulation of hematopoietic stem cell homing and engraftment by CD26. Science 305 , 1000-1003 (2004). https://doi.org/10.1126/science.1097071 Wang, C. et al. Dipeptidylpeptidase 4 promotes survival and stemness of acute myeloid leukemia stem cells. Cell Rep 42 , 112105 (2023). https://doi.org/10.1016/j.celrep.2023.112105 Davoodi, J., Kelly, J., Gendron, N. H. & MacKenzie, A. E. The Simpson-Golabi-Behmel syndrome causative glypican-3, binds to and inhibits the dipeptidyl peptidase activity of CD26. Proteomics 7 , 2300-2310 (2007). https://doi.org/10.1002/pmic.200600654 De Cat, B. et al. Processing by proprotein convertases is required for glypican-3 modulation of cell survival, Wnt signaling, and gastrulation movements. J Cell Biol 163 , 625-635 (2003). https://doi.org/10.1083/jcb.200302152 Zhang, J. et al. Identification of the haematopoietic stem cell niche and control of the niche size. Nature 425 , 836-841 (2003). https://doi.org/10.1038/nature02041 Xie, Y. et al. Detection of functional haematopoietic stem cell niche using real-time imaging. Nature 457 , 97-101 (2009). https://doi.org/10.1038/nature07639 Zhao, M. et al. N-Cadherin-Expressing Bone and Marrow Stromal Progenitor Cells Maintain Reserve Hematopoietic Stem Cells. Cell reports 26 , 652-669.e656 (2019). https://doi.org/10.1016/j.celrep.2018.12.093 Weidenaar, A. C. et al. High acute myeloid leukemia derived VEGFA levels are associated with a specific vascular morphology in the leukemic bone marrow. Cell Oncol (Dordr) 34 , 289-296 (2011). https://doi.org/10.1007/s13402-011-0017-9 Padró, T. et al. Increased angiogenesis in the bone marrow of patients with acute myeloid leukemia. Blood 95 , 2637-2644 (2000). Kupsa, T., Vanek, J., Pavel, Z., Jebavy, L. & Horacek, J. M. Serum levels of soluble adhesion molecules in newly diagnosed acute myeloid leukemia and in complete remission suggest endothelial cell activation by myeloblasts. Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub 161 , 92-99 (2017). https://doi.org/10.5507/bp.2016.054 Tinoco, A. D., Tagore, D. M. & Saghatelian, A. Expanding the dipeptidyl peptidase 4-regulated peptidome via an optimized peptidomics platform. J Am Chem Soc 132 , 3819-3830 (2010). https://doi.org/10.1021/ja909524e Campbell, T. B. & Broxmeyer, H. E. CD26 inhibition and hematopoiesis: a novel approach to enhance transplantation. Front Biosci 13 , 1795-1805 (2008). Kei Ohnuma, C. M. DPP4. Atlas of Genetics and Cytogenetics in Oncology and Haematology (2012). Broxmeyer, H. E. et al. Dipeptidylpeptidase 4 negatively regulates colony-stimulating factor activity and stress hematopoiesis. Nat Med 18 , 1786-1796 (2012). https://doi.org/10.1038/nm.2991 Peled, A. et al. Dependence of human stem cell engraftment and repopulation of NOD/SCID mice on CXCR4. Science 283 , 845-848 (1999). Petit, I. et al. G-CSF induces stem cell mobilization by decreasing bone marrow SDF-1 and up-regulating CXCR4. Nature immunology 3 , 687-694 (2002). https://doi.org/10.1038/ni813 Singh, P. et al. Neuropeptide Y regulates a vascular gateway for hematopoietic stem and progenitor cells. J Clin Invest 127 , 4527-4540 (2017). https://doi.org/10.1172/jci94687 Wei, Q. & Frenette, P. S. Niches for Hematopoietic Stem Cells and Their Progeny. Immunity 48 , 632-648 (2018). https://doi.org/10.1016/j.immuni.2018.03.024 Calvi, L. M. et al. Osteoblastic cells regulate the haematopoietic stem cell niche. Nature 425 , 841-846 (2003). https://doi.org/10.1038/nature02040 Zhou, B. O., Yue, R., Murphy, M. M., Peyer, J. G. & Morrison, S. J. Leptin-receptor-expressing mesenchymal stromal cells represent the main source of bone formed by adult bone marrow. Cell Stem Cell 15 , 154-168 (2014). https://doi.org/10.1016/j.stem.2014.06.008 Ding, L. & Morrison, S. J. Haematopoietic stem cells and early lymphoid progenitors occupy distinct bone marrow niches. Nature 495 , 231-235 (2013). https://doi.org/10.1038/nature11885 Greenbaum, A. et al. CXCL12 in early mesenchymal progenitors is required for haematopoietic stem-cell maintenance. Nature 495 , 227-230 (2013). https://doi.org/10.1038/nature11926 Mendez-Ferrer, S. et al. Mesenchymal and haematopoietic stem cells form a unique bone marrow niche. Nature 466 , 829-834 (2010). https://doi.org/10.1038/nature09262 Lu, X. et al. Identification of the homeobox protein Prx1 (MHox, Prrx-1) as a regulator of osterix expression and mediator of tumor necrosis factor alpha action in osteoblast differentiation. J Bone Miner Res 26 , 209-219 (2011). https://doi.org/10.1002/jbmr.203 Esposito, A., Wang, L., Li, T., Miranda, M. & Spagnoli, A. Role of Prx1-expressing skeletal cells and Prx1-expression in fracture repair. Bone 139 , 115521 (2020). https://doi.org/10.1016/j.bone.2020.115521 Mizoguchi, T. et al. Osterix marks distinct waves of primitive and definitive stromal progenitors during bone marrow development. Dev Cell 29 , 340-349 (2014). https://doi.org/10.1016/j.devcel.2014.03.013 Forte, D. et al. Bone Marrow Mesenchymal Stem Cells Support Acute Myeloid Leukemia Bioenergetics and Enhance Antioxidant Defense and Escape from Chemotherapy. Cell Metab 32 , 829-843.e829 (2020). https://doi.org/10.1016/j.cmet.2020.09.001 Pinho, S. et al. VCAM1 confers innate immune tolerance on haematopoietic and leukaemic stem cells. Nat Cell Biol 24 , 290-298 (2022). https://doi.org/10.1038/s41556-022-00849-4 Ma, D. et al. Intracellular angiopoietin-1 promotes TKI-resistance via activation of JAK/STAT5 pathway in chronic myeloid leukemia. Oncogene 42 , 124-137 (2023). https://doi.org/10.1038/s41388-022-02536-y Carlesso, N. et al. Human recombinant stem cell factor stimulates in vitro proliferation of acute myeloid leukemia cells and expands the clonogenic cell pool. Leukemia 6 , 642-648 (1992). Bendall, L. J. et al. Stem cell factor enhances the adhesion of AML cells to fibronectin and augments fibronectin-mediated anti-apoptotic and proliferative signals. Leukemia 12 , 1375-1382 (1998). https://doi.org/10.1038/sj.leu.2401136 Kunz, L. & Schroeder, T. A 3D Tissue-wide Digital Imaging Pipeline for Quantitation of Secreted Molecules Shows Absence of CXCL12 Gradients in Bone Marrow. Cell Stem Cell 25 , 846-854 e844 (2019). https://doi.org/10.1016/j.stem.2019.10.003 Janssens, R., Struyf, S. & Proost, P. The unique structural and functional features of CXCL12. Cell Mol Immunol 15 , 299-311 (2018). https://doi.org/10.1038/cmi.2017.107 Cui, X. Y. et al. Tissue factor pathway inhibitor upregulates CXCR7 expression and enhances CXCL12-mediated migration in chronic lymphocytic leukemia. Sci Rep 11 , 5127 (2021). https://doi.org/10.1038/s41598-021-84695-8 Ramakrishnan, R. et al. CXCR4 Signaling Has a CXCL12-Independent Essential Role in Murine MLL-AF9-Driven Acute Myeloid Leukemia. Cell Rep 31 , 107684 (2020). https://doi.org/10.1016/j.celrep.2020.107684 Anderson, N. R. et al. CXCL12 Knock-out Enhances Leukemia Stem Cell Response to Combination Chemotherapy Plus Tyrosine Kinase Inhibition in Flt3-ITD Acute Myeloid Leukemia. Blood 136 , 7-8 (2020). https://doi.org/https://doi.org/10.1182/blood-2020-142579 Zeng, Z. et al. Targeting the leukemia microenvironment by CXCR4 inhibition overcomes resistance to kinase inhibitors and chemotherapy in AML. Blood 113 , 6215-6224 (2009). https://doi.org/10.1182/blood-2008-05-158311 Agarwal, P. et al. Mesenchymal Niche-Specific Expression of Cxcl12 Controls Quiescence of Treatment-Resistant Leukemia Stem Cells. Cell Stem Cell 24 , 769-784.e766 (2019). https://doi.org/10.1016/j.stem.2019.02.018 Schelker, R. C. et al. TGF-beta1 and CXCL12 modulate proliferation and chemotherapy sensitivity of acute myeloid leukemia cells co-cultured with multipotent mesenchymal stromal cells. Hematology 23 , 337-345 (2018). https://doi.org/10.1080/10245332.2017.1402455 Tavor, S. et al. CXCR4 regulates migration and development of human acute myelogenous leukemia stem cells in transplanted NOD/SCID mice. Cancer Res 64 , 2817-2824 (2004). https://doi.org/10.1158/0008-5472.can-03-3693 Anderson, N. R. et al. Microenvironmental CXCL12 deletion enhances Flt3-ITD acute myeloid leukemia stem cell response to therapy by reducing p38 MAPK signaling. Leukemia 37 , 560-570 (2023). https://doi.org/10.1038/s41375-022-01798-5 Namburi, S., Broxmeyer, H. E., Hong, C. S., Whiteside, T. L. & Boyiadzis, M. DPP4(+) exosomes in AML patients' plasma suppress proliferation of hematopoietic progenitor cells. Leukemia 35 , 1925-1932 (2021). https://doi.org/10.1038/s41375-020-01047-7 Cline, A., Jajosky, R., Shikle, J. & Bollag, R. Comparing leukapheresis protocols for an AML patient with symptomatic leukostasis. J Clin Apher 33 , 396-400 (2018). https://doi.org/10.1002/jca.21588 Del Prete, C., Kim, T., Lansigan, F., Shatzel, J. & Friedman, H. The Epidemiology and Clinical Associations of Stroke in Patients With Acute Myeloid Leukemia: A Review of 10,972 Admissions From the 2012 National Inpatient Sample. Clin Lymphoma Myeloma Leuk 18 , 74-77 e71 (2018). https://doi.org/10.1016/j.clml.2017.09.008 Havale, S. H. & Pal, M. Medicinal chemistry approaches to the inhibition of dipeptidyl peptidase-4 for the treatment of type 2 diabetes. Bioorg Med Chem 17 , 1783-1802 (2009). https://doi.org/10.1016/j.bmc.2009.01.061 de Boer, J. et al. Transgenic mice with hematopoietic and lymphoid specific expression of Cre. Eur J Immunol 33 , 314-325 (2003). https://doi.org/10.1002/immu.200310005 Gothert, J. R. et al. In vivo fate-tracing studies using the Scl stem cell enhancer: embryonic hematopoietic stem cells significantly contribute to adult hematopoiesis. Blood 105 , 2724-2732 (2005). https://doi.org/10.1182/blood-2004-08-3037 Kang, X. & Wang, C. CAMKs support development of acute myeloid leukemia. The Journal of experimental medicine 11 , 30 (2018). https://doi.org/10.1084/jem.20171477 10.1186/s13045-018-0574-8 Zheng, J. et al. Inhibitory receptors bind ANGPTLs and support blood stem cells and leukaemia development. Nature (2012). Additional Declarations There is NO Competing Interest. Supplementary Files SupplementaryFigures.docx Cite Share Download PDF Status: Published Journal Publication published 24 Apr, 2026 Read the published version in Nature Cell Biology → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6515437","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":447640967,"identity":"ea1a1d8e-a031-4aa0-8060-e8bca1f4e1a7","order_by":0,"name":"Xunlei kang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABFElEQVRIiWNgGAWjYBACCQhlk8DAwNzADGTxQAQMgPgAXi1pQC2Mjc2kaDkM14IEcGiRbD97+DVv2/k8fvaD7Y8LGLbJyPcfPiZdUcAgx3cjAasWaZ68NMuZbbeLJXsSG5tnMNzmMbiRliZ5xoDBWBKHFjmGHDODj223EzccAGrhAWmR4DGTbDBgSNyASwv/GzODxLZzifvPP4Roke8/A9ZSj0uLtESO8YOPbQcSN0hAbWE4kAPWkmCAQ4vkjDdmjDPOJSfOuPGwcTaPAdgvyZYNBhKGM888wKpF4nyO8WeeMrvE/v7kA595Km7bA0Ps4M2GPzbyfMex2wIEbBIItgHCLFzKQYD5Az7ZUTAKRsEoGAUMAOjsYao5ES08AAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0003-1853-1881","institution":"University of Missouri","correspondingAuthor":true,"prefix":"","firstName":"Xunlei","middleName":"","lastName":"kang","suffix":""},{"id":447640968,"identity":"25812dee-aed6-4f11-b689-bb0ab2784532","order_by":1,"name":"Chen Wang","email":"","orcid":"https://orcid.org/0000-0001-7307-9256","institution":"University of Missouri","correspondingAuthor":false,"prefix":"","firstName":"Chen","middleName":"","lastName":"Wang","suffix":""},{"id":447640969,"identity":"62a9731e-e8ee-4e1a-8729-03c0e8bdaa81","order_by":2,"name":"Yi Pan","email":"","orcid":"","institution":"University of Missouri School of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Yi","middleName":"","lastName":"Pan","suffix":""},{"id":447640970,"identity":"8c076666-8995-4158-ab35-51b1298df063","order_by":3,"name":"Ruochen Dong","email":"","orcid":"","institution":"Stowers Institute for Medical Research","correspondingAuthor":false,"prefix":"","firstName":"Ruochen","middleName":"","lastName":"Dong","suffix":""},{"id":447640971,"identity":"4b9db181-c7b2-4c21-875e-cb9bf150fd11","order_by":4,"name":"WenXuan Zhou","email":"","orcid":"","institution":"University of Missouri School of Medicine","correspondingAuthor":false,"prefix":"","firstName":"WenXuan","middleName":"","lastName":"Zhou","suffix":""},{"id":447640972,"identity":"30427e9b-95c9-492c-9425-ecb55718cc4c","order_by":5,"name":"XiaDuo Meng","email":"","orcid":"","institution":"University of Missouri School of Medicine","correspondingAuthor":false,"prefix":"","firstName":"XiaDuo","middleName":"","lastName":"Meng","suffix":""},{"id":447640973,"identity":"65e45f33-bf9f-4eb3-957a-64250ba7098b","order_by":6,"name":"Ravi Nistala","email":"","orcid":"","institution":"University of Missouri","correspondingAuthor":false,"prefix":"","firstName":"Ravi","middleName":"","lastName":"Nistala","suffix":""},{"id":447640974,"identity":"c4056843-68db-45a1-8353-6342c21e95af","order_by":7,"name":"Richard Hammer","email":"","orcid":"https://orcid.org/0000-0002-7173-9414","institution":"University of Missouri","correspondingAuthor":false,"prefix":"","firstName":"Richard","middleName":"","lastName":"Hammer","suffix":""},{"id":447640975,"identity":"7ff11ebb-3dc4-4996-b18f-7b11c09c5c60","order_by":8,"name":"Linheng Li","email":"","orcid":"https://orcid.org/0000-0001-9963-430X","institution":"Stowers Institute for Medical Research","correspondingAuthor":false,"prefix":"","firstName":"Linheng","middleName":"","lastName":"Li","suffix":""}],"badges":[],"createdAt":"2025-04-23 20:45:13","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6515437/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6515437/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41556-026-01939-3","type":"published","date":"2026-04-24T04:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":81939394,"identity":"f4e4c6ec-950b-4aa9-8797-0cfcba946c37","added_by":"auto","created_at":"2025-05-05 06:43:07","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":2149237,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eDpp4\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003e deficiency re-distributes AML cells and induces LSC exhaustion\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(a-b) Representative images of BM sections from \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e (left) or \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e−/−\u003c/sup\u003e (right) primary AML mouse models, including MLL-AF9 (MA9, a) and AML-ETO9a (AE9, b). These images were captured at weeks 2 and 12 post-transplantation and depict similar AML cell compositions in the BM (~30%). Green, GFP\u003csup\u003e+\u003c/sup\u003e AML cells; red, endomucin staining blood vessels; blue, DAPI for nuclei. The calculated proportion of GFP\u003csup\u003e+\u003c/sup\u003e signals in each of the three anatomical BM areas is listed to the left of the images. The dashed box indicates the area of focus. Scale bars, 100mm. (c) Schematic representation of the secondary AML transplantation models. 1x10\u003csup\u003e6\u003c/sup\u003e primary AML cells were transplanted into sub-lethally irradiated recipients and compared AML distribution within BM when AML mice reach ~30% GFP\u003csup\u003e+\u003c/sup\u003e in the PB, respectively. (d) Summary of secondary MA9 and AE9 GFP\u003csup\u003e+\u003c/sup\u003e signals in each BM area (PM, CM, DM), showing that \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e AML cells are more heavily concentrated in the PM, while \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e−/−\u003c/sup\u003e AML cells are more evenly distributed throughout all BM areas. For MA9 model, n = 5, BM sections from 5 mice, \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e BM sections were collected at day 14, 15, 15, 16,17 while \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e−/−\u003c/sup\u003e BM sections were collected at day 28, 29, 30 ,31,31,33. For AE9 transplanted mice. n = 5, BM sections from 5 mice. \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e BM sections were collected at day 34, 35, 35, 36, 40 while \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e−/−\u003c/sup\u003e BM sections were collected at day 65, 66, 67, 70, 73. Scale bars, 100mm. (e-f) LSC population (GFP\u003csup\u003e+ \u003c/sup\u003eIL-7R\u003csup\u003e− \u003c/sup\u003eLin\u003csup\u003e− \u003c/sup\u003eSca-1\u003csup\u003e− \u003c/sup\u003ec-Kit\u003csup\u003e+ \u003c/sup\u003eCD34\u003csup\u003e+ \u003c/sup\u003eFcgRII/III\u003csup\u003e+\u003c/sup\u003e, GMP-like leukemic cells, L-GMP) measured by flow cytometry in the BM of \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e−/−\u003c/sup\u003e and \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e AML mice at 2 weeks after transplantation (n = 5 mice). (g) Cell cycle analysis of MLL-AF9 L-GMP in \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e and \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e−/−\u003c/sup\u003e\u003csup\u003e\u003cem\u003e \u003c/em\u003e\u003c/sup\u003emice (n = 5 mice per group) by flow cytometry. (h) Flow cytometry analysis of early (detected as Annexin V\u003csup\u003e+\u003c/sup\u003e/7-AAD\u003csup\u003e− \u003c/sup\u003estaining) and late (detected as Annexin V\u003csup\u003e+\u003c/sup\u003e/7-AAD\u003csup\u003e+\u003c/sup\u003e staining) apoptotic L-GMP from the BM of MLL-AF9 mice (n = 5 mice).\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-6515437/v1/9af3364d6a562fb50cf8449b.png"},{"id":81939865,"identity":"9d63df7e-6dce-428e-b04d-b14ef1a4f068","added_by":"auto","created_at":"2025-05-05 06:51:08","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":177177,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe PM and DM mark distinct niches versus CM for maintaining LSCs\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(a-b) L-GMP measured by flow cytometry in the PM, CM and DM areas of \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e AML mice at 2 weeks after 1x10\u003csup\u003e6 \u003c/sup\u003eMA9 cells transplantation as shown in Figure 1c (n = 5 mice). (c-d) Cell cycle analysis of MA9 L-GMPs in PM, CM and DM areas from \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e\u003cem\u003e+/+\u003c/em\u003e\u003c/sup\u003e mice (n = 5 mice per group) by flow cytometry. (e) Survival curve of secondary transplanted mice receiving 500 L-GMPs from PM, CM and DM areas of \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e AML mice (n = 10 mice). (f) Table summarizing the survival ratio (number of leukemic mice / total recipients) at varying transplanted cell doses (10–5000 cells) in PM, CM, and DM groups. The frequency of tumor-initiating cells (CRUs) was calculated using LDA (Limiting Dilution Analysis), with 95% confidence intervals provided. Log-log plot showing the fraction of non-responding mice versus the number of transplanted cells (Log₁₀ scale). Each point represents the fraction of mice that did not develop leukemia at a given cell dose. The TIC frequency for each group is shown (PM: 1/76.6, CM: 1/394, DM: 1/102), with statistical comparison between groups indicating a significant difference (CM vs PM p= 0.0001811; CM vs DM p= 0.008248).\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-6515437/v1/e1df31e30e6b9e5428280422.png"},{"id":81939396,"identity":"d04b4278-404f-4f8b-a3f9-03b056fdf929","added_by":"auto","created_at":"2025-05-05 06:43:08","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":530347,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eDpp4\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003e deficiency confines AML cells within the BM through reversed BM-to-PB CXCL12 gradient\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(a) Engraftment efficiency measured by comparison of the proportions of MA9 and AE9 GFP\u003csup\u003e+\u003c/sup\u003e \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e or \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e−/−\u003c/sup\u003eAML cells in BM, PB, spleen, and liver 2 weeks (MA9) and 4 weeks (AE9) post transplantation. (n = 5 mice). (b) Homing efficiency measured by comparison of the proportions of MA9 and AE9 GFP\u003csup\u003e+\u003c/sup\u003e \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e or \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e−/−\u003c/sup\u003eAML cells in BM, PB, spleen, and liver after 16 hours post transplantation (n = 5 mice). (c) Survival curve of secondary transplanted mice receiving 1 million MA9 or AE9 GFP\u003csup\u003e+ \u003c/sup\u003e\u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e or \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e−/− \u003c/sup\u003eAML cells (n = 10 mice; P\u0026lt; 0.001, log-rank test). (d-e) Distance between AML cells and blood vessels. The numbers on the x axis indicate intervals of 5 μm (5 indicates the interval 0–5; 10 indicates 5–10, and so on). p value (\u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e) = 3.1x10\u003csup\u003e-16\u003c/sup\u003e, p value (\u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e−/−\u003c/sup\u003e) = 3.7x10\u003csup\u003e-16 \u003c/sup\u003eby the two-sample Kolmogorov-Smirnov (KS) test. (f) The observed mean distances of \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e or \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e−/−\u003c/sup\u003e MA9 AML cells to the nearest BM vessel are similar to each other. Probability distribution of the mean distances between GFP\u003csup\u003e+\u003c/sup\u003e AML cells and vessels derived from simulations of randomly positioned GFP\u003csup\u003e+\u003c/sup\u003e AML cells and actual vessels on maps of BM. Mean distances observed \u003cem\u003ein situ\u003c/em\u003e for both \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e AML cells (black line) and \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e−/−\u003c/sup\u003e AML cells (red line) are shown in relation to the grand mean (mean of the means) ± 3 SD (3σ), (dotted black and dotted red lines, respectively). The observed mean distances of both \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e AML cells (12.70 mM) or \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e−/−\u003c/sup\u003e AML cells (12.55 mM) were statistically different from the mean distance of randomly placed \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e AML cells (76.12 mM) or randomly placed \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e−/−\u003c/sup\u003e AML cells (76.11 mM) to vessels. n = 500 AML cells (\u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e and \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e−/−\u003c/sup\u003e), respectively, from three mice in three independent experiments. (g-h) Cytokine protein levels were measured in BMEF (g) and plasma (h) of \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e and \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e−/−\u003c/sup\u003e AML cells transplanted mice using the LEGEND plex Muti-Analyte Flow Assay Kit. (n = 3 mouse samples). (i-j) Quantitation of CXCL12 in BMEF (i) or plasma (j) of \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e and \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e−/−\u003c/sup\u003e AML cells transplanted mice by ELISA (n = 5 mice). (k) Mobilization of circulating GFP⁺ AML cells in peripheral blood was measured 17-hour post-injection by flow cytometry. Dpp4 KO (black) and WT, red) AML mice were intravenously injected with increasing doses of recombinant CXCL12 (0–500 ng) (n = 5 mice). (l) CXCR4 expression on GFP\u003csup\u003e+ \u003c/sup\u003eMac-1\u003csup\u003e+ \u003c/sup\u003eKit\u003csup\u003e+ \u003c/sup\u003e\u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e and \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e−/−\u003c/sup\u003e AML-SCs as determined by flow cytometry (n = 5 mice). (m) Representative images of the lower wells of the migration assays. For the \u003cem\u003ein vitro\u003c/em\u003e migration assay; the upper wells contained \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e or \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e−/−\u003c/sup\u003e AML cells; all lower wells contained 100 ng/ml CXCL12; and the concentration of CXCL12 in the upper wells is 0 ng/ml. (n) Quantification of AML cell migration after 4 hours of incubation (performed in m) at 37°C and 5% CO2 (n = 3 wells).\u0026nbsp;\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-6515437/v1/256da17e3d1862f86e4a117e.png"},{"id":81940776,"identity":"77323337-3604-4b09-8bd9-4e004c3aded3","added_by":"auto","created_at":"2025-05-05 06:59:08","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":1365380,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eNiche-specific CXCL12 production by N-cad ⁺ MSCs dictates LSC spatial distribution \u003c/strong\u003e(a-b) Cartoon of BM structure anatomical partition (a) and summary (b) of CXCL12 levels were measured in each BM area (PM, CM, and DM) from \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e−/−\u003c/sup\u003e and \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e AML mice. (c) Single-cell RNA sequencing of 11,512 non-hematopoietic cells from PM and DM regions 3 weeks post-AML transplantation revealed 14 distinct clusters, including mesenchymal stromal/stem cells (MSCs), OLCs, chondrocytes (and transitional states), fibroblasts, endothelial cells, and pericytes. (d) Violine plot shows \u003cem\u003eCxcl12\u003c/em\u003e expression among identified cell clusters. (e) Correlation analysis of \u003cem\u003eCxcl12\u003c/em\u003e expression level and MSC subtypes marked by \u003cem\u003eLepr\u003c/em\u003e, \u003cem\u003eCdh2 \u003c/em\u003e(N-cadherin), \u003cem\u003eSp7\u003c/em\u003e (Orx), \u003cem\u003ePrx\u003c/em\u003e, and \u003cem\u003eNes\u003c/em\u003e (Nestin). (f) UMAP feature plots show expression of \u003cem\u003eCxcl12\u003c/em\u003e across MSC subtypes marked by \u003cem\u003eLepr\u003c/em\u003e, \u003cem\u003eCdh2\u003c/em\u003e, \u003cem\u003eSp7\u003c/em\u003e (Orx), \u003cem\u003ePrx\u003c/em\u003e, and \u003cem\u003eNes\u003c/em\u003e (Nestin). (g) UMAP plots show the co-expression of \u003cem\u003eCdh2\u003c/em\u003e (green) and selected niche factors (\u003cem\u003eCxcl12, Angpt1\u003c/em\u003e, \u003cem\u003eKitl\u003c/em\u003e, \u003cem\u003eGas6\u003c/em\u003e, \u003cem\u003eTgfb1\u003c/em\u003e, \u003cem\u003eVcam1\u003c/em\u003e; red), with merged signals in yellow. (e) Correlation analysis between Cxcl12 expression and MSC subtype markers revealed strong associations with Lepr and Cdh2, indicating these MSC populations as primary sources of CXCL12 in the AML BM niche. (f) UMAP feature plots show broad expression of Cdh2 across the main MSC cluster, consistent with the known distribution of LepR⁺ MSCs throughout the BM, including the metaphysis and central marrow regions. Other MSC markers, including Prx, Osx, and Nes, were minimally expressed within the analyzed population. (g) UMAP plots show the spatial expression of Cdh2 (green) and selected niche factors (Cxcl12, Angpt1, Kitl, Gas6, Tgfb1, Vcam1; red), with merged signals in yellow. Overlapping expression patterns are observed within the MSC compartment. (h) Representative images of BM sections of \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e\u003cem\u003e+/+\u003c/em\u003e\u003c/sup\u003e AML cell localization in the BM of N-cad-TdT mice at day 15 after transplantation. Green, GFP\u003csup\u003e+\u003c/sup\u003e AML cells; red, N-cad+ MSCs; White, Kit; blue, DAPI for nuclei. The dashed box indicates the area of focus. Scale bars：100mm. (h\u003cstrong\u003e-\u003c/strong\u003ebottom right) Relative distance between GFP\u003csup\u003e+ \u003c/sup\u003eKit\u003csup\u003e+\u003c/sup\u003e AML cells and GFP\u003csup\u003e+ \u003c/sup\u003eKit\u003csup\u003e−\u003c/sup\u003e AML cells to N-cad\u003csup\u003e+\u003c/sup\u003e cells (n = 100 GFP\u003csup\u003e+ \u003c/sup\u003eKit\u003csup\u003e+\u003c/sup\u003e AML cells, n = 80 GFP\u003csup\u003e+ \u003c/sup\u003eKit\u003csup\u003e−\u003c/sup\u003e AML cells). n=3 mice for each distance quantification dataset. (i) CXCL12 levels were measured in each BM area (PM, CM, and DM) from N-cad; \u003cem\u003eCxcl12\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e or N-cad; \u003cem\u003eCxcl12\u003c/em\u003e\u003csup\u003e−/− \u003c/sup\u003emice. (j) Schematic representation of \u003cem\u003eCxcl12\u003c/em\u003e conditional knockout in the N-cad\u003csup\u003e+ \u003c/sup\u003ecell mouse model. 500 primary L-GMPs were transplanted into sub-lethally irradiated recipients 14 days post tamoxifen injection. (k-l) Immunostaining and statistical analysis of L-GMP markers in the BM section of N-cad; \u003cem\u003eCxcl12\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e (top) and N-cad; \u003cem\u003eCxcl12\u003c/em\u003e\u003csup\u003e−/− \u003c/sup\u003e(bottom) AML mice at day 14 and day 31 respectively after transplantation. The calculated proportion of L-GMPs / GFP\u003csup\u003e+\u003c/sup\u003e cells in each of the three anatomical BM areas is listed to the left of the images. The dashed box indicates the area of focus. Arrow heads point to the L-GMPs, scale Bars, 100mm.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-6515437/v1/b0c4d371314c2057aca06c40.png"},{"id":81939400,"identity":"220cd461-1f7a-4891-9f84-4172082fda72","added_by":"auto","created_at":"2025-05-05 06:43:08","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":709036,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eShared LSC behavior and transcriptomic signatures in N-cad; \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003eCxcl12\u003c/strong\u003e\u003c/em\u003e\u003csup\u003e\u003cstrong\u003e−/−\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e and \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003eDpp4\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003e KO AML mice \u003c/strong\u003e(a) Immunofluorescent image shows the levels of cleaved caspase-3 in the BM of N-cad; \u003cem\u003eCxcl12\u003c/em\u003e\u003csup\u003e+/+ \u003c/sup\u003eand N-cad; \u003cem\u003eCxcl12\u003c/em\u003e\u003csup\u003e−/−\u003c/sup\u003e AML mice at day 14 and day 31 respectively after transplantation. Green, GFP\u003csup\u003e+\u003c/sup\u003e AML cells; red, cleaved-caspase3, bule, DAPI; white, Kit\u003csup\u003e+\u003c/sup\u003e cells; scale bars, 100mm. (b-c) Apoptosis and cell cycle analysis of MA9 AML L-GMPs in N-cad; \u003cem\u003eCxcl12\u003c/em\u003e\u003csup\u003e+/+ \u003c/sup\u003eand N-cad; \u003cem\u003eCxcl12\u003c/em\u003e\u003csup\u003e−/−\u003c/sup\u003e AML mice (n = 5 mice per group) by flow cytometry. (d) Comparison of CFU capability of N-cad; \u003cem\u003eCxcl12\u003c/em\u003e\u003csup\u003e+/+ \u003c/sup\u003eand N-cad; \u003cem\u003eCxcl12\u003c/em\u003e\u003csup\u003e−/−\u003c/sup\u003e AML cells during serial re-plating (2,000 cells/well, n = 3 wells). (e) Survival curve of 500 control L-GMPs transplanted recipient N-cad; \u003cem\u003eCxcl12\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e or N-cad; \u003cem\u003eCxcl12\u003c/em\u003e\u003csup\u003e−/− \u003c/sup\u003emice (n= 15 mice; P\u0026lt; 0.0001, log-rank test). (f) Comparison of the proportions of GFP\u003csup\u003e+\u003c/sup\u003e N-cad; \u003cem\u003eCxcl12\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e or N-cad; \u003cem\u003eCxcl12\u003c/em\u003e\u003csup\u003e−/− \u003c/sup\u003eAML cells in BM, PB, spleen, and liver 6 weeks post transplantation (n=5 mice). (g) PCA analysis for AML cells from N-cad; \u003cem\u003eCxcl12\u003c/em\u003e\u003csup\u003e−/−\u003c/sup\u003e and \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e−/− \u003c/sup\u003eAML mice and their pairing N-cad; \u003cem\u003eCxcl12\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e and \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e AML mice. (h) Heatmap showing differentially expressed genes in AML cells from N-cad; \u003cem\u003eCxcl12\u003c/em\u003e⁺/⁺ and N-cad; \u003cem\u003eCxcl12\u003c/em\u003e⁻/⁻ mice, as well as from \u003cem\u003eDpp4\u003c/em\u003e⁺/⁺ and \u003cem\u003eDpp4\u003c/em\u003e⁻/⁻ mice. (i) Bar plot of GSEA results illustrating biological processes associated with AML cells of N-cad; Cxcl12⁻/⁻ mice and \u003cem\u003eDpp4\u003c/em\u003e⁻/⁻ mice shows the significant alteration of migration, stemness, proliferation, and the cell cycle.\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-6515437/v1/924bb6b10b51816fc9c4e891.png"},{"id":81939868,"identity":"60c88f03-9e1c-4a0f-9a2f-2c0f0b45378f","added_by":"auto","created_at":"2025-05-05 06:51:08","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":716396,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eN-cad\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e+\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e cells support LSCs through GPC3 mediated attraction\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(a) Representative images of BM sections of \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e−/−\u003c/sup\u003e AML cell localization in the BM of N-cad-TdT mice at day 15 after transplantation. Green, GFP\u003csup\u003e+\u003c/sup\u003e AML cells; red, N-cad\u003csup\u003e+\u003c/sup\u003e MSCs; White, Kit; blue, DAPI for nuclei. The dashed box indicates the area of focus. Scale bars：100mm. (b) Distance between \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e (Figure 4h) or \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e−/−\u003c/sup\u003e (Figure 5a) LSC-enriched populations (GFP\u003csup\u003e+ \u003c/sup\u003eKit\u003csup\u003e+\u003c/sup\u003e) and N-cad\u003csup\u003e+\u003c/sup\u003e cells. The numbers on the x axis indicate intervals of 5 mm (5 indicates the interval 0–5; 10 indicates 5–10, and so on, totally 200 AML cells for each group calculated). (c) Cell cycle analysis of MA9 L-GMPs in PM, CM and DM areas from \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e−/−\u003c/sup\u003e mice (n = 5 mice per group) by flow cytometry. (d) Bar graphs compare the expression levels of the indicated transcripts in N-cad\u003csup\u003e+\u003c/sup\u003e MSCs and MA9 AML cells. All transcripts’ levels were normalized to levels of actin expression (n = 3 wells). (e) Contour plots show that N-cad\u003csup\u003e+\u003c/sup\u003e BM cells have significantly higher cell surface expression of GPC3 than N-cad\u003csup\u003e−\u003c/sup\u003e BM cells. GPC3 expression level (median fluorescent intensity, MFI) in N-cad\u003csup\u003e+\u003c/sup\u003e and N-cad\u003csup\u003e−\u003c/sup\u003e populations have been indicated. (f) Flow cytometry analysis of recombinant GPC3 (100 nM) binding to DPP4\u003csup\u003e+\u003c/sup\u003e BM AML cells, MFIs are indicated. (g) Comparison of the DPP4 enzyme activity of 1 x10\u003csup\u003e6\u003c/sup\u003e mouse AML cells measured by fluorescence assay with the indicated treatment and time points. Results are reported as relative light units (RLUs) (n = 3 wells). (h) Representative BM section images of GFP-labeled \u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e AML cells in the N-cad-TdT AML mouse trabecular bone region at day 15 after transplantation. The dashed box indicates the area of focus. Dash line indicates the bone structure; scale bar, 20mM. Selected 3D image of DPP4 and GPC3 interaction. Green, GFP\u003csup\u003e+\u003c/sup\u003e AML cells; red, N-cad\u003csup\u003e+\u003c/sup\u003e s; blue, DPP4; yellow, GPC3. (i) Statistic analysis of expression and localization of GPC3. (j) Percentage of DPP4\u003csup\u003e+\u003c/sup\u003e AML cells overlap with GPC3\u003csup\u003e+\u003c/sup\u003e N-cad\u003csup\u003e+\u003c/sup\u003e MSCs.\u0026nbsp;\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-6515437/v1/2fae56f327fcae0e9883e83d.png"},{"id":81939873,"identity":"f4c3dc8f-f875-44a9-98c0-a124bc9b7efb","added_by":"auto","created_at":"2025-05-05 06:51:08","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":959225,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eGpc3\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003e knockout in N-cad\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e+\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e redistributes AML cells and impairs LSCs\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(a) Schematic representation of \u003cem\u003eGpc3\u003c/em\u003e conditional knockout in the N-cad\u003csup\u003e+ \u003c/sup\u003ecell mouse model. 500 primary L-GMPs were transplanted into sub-lethally irradiated recipients 14 days post tamoxifen injection. (b) Cxcl12 levels were measured in each BM area (PM, CM, and DM) from N-cad; \u003cem\u003eGpc3\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e or N-cad; \u003cem\u003eGpc3\u003c/em\u003e\u003csup\u003e−/− \u003c/sup\u003emice. (c) Representative images of BM sections of N-cad; \u003cem\u003eGpc3\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e (top) and N-cad; \u003cem\u003eGpc3\u003c/em\u003e\u003csup\u003e−/− \u003c/sup\u003e(middle) AML cell localization in the BM of N-cad-TdT mice at day 15 and day 28 respectively after transplantation. Green, GFP\u003csup\u003e+\u003c/sup\u003e AML cells; red, N-cad\u003csup\u003e+\u003c/sup\u003e MSCs; white, Kit\u003csup\u003e+\u003c/sup\u003e; blue, DAPI for nuclei. The dashed box indicates the area of focus. Scale bars：100mm. (bottom) Distance between AML cells and N-cad\u003csup\u003e+\u003c/sup\u003e cells. The numbers on the x axis indicate intervals of 5 mm (5 indicates the interval 0–5; 10 indicates 5–10, and so on). (d) Cell cycle analysis of MA9 L-GMPs in PM, CM and DM areas in N-cad; \u003cem\u003eGpc3\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e or N-cad; \u003cem\u003eGpc3\u003c/em\u003e\u003csup\u003e−/− \u003c/sup\u003emice by flow cytometry, (n = 5 mice per group). (e) Apoptosis analysis of MA9 L-GMPs in PM, CM and DM areas in N-cad; \u003cem\u003eGpc3\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e or N-cad; \u003cem\u003eGpc3\u003c/em\u003e\u003csup\u003e−/− \u003c/sup\u003emice. (n = 5 mice per group) by flow cytometry. (f) Survival curve of control L-GMPs transplanted recipient N-cad; \u003cem\u003eGpc3\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e or N-cad; \u003cem\u003eGpc3\u003c/em\u003e\u003csup\u003e−/− \u003c/sup\u003emice (n= 10 mice; P\u0026lt; 0.0001, log-rank test).\u003c/p\u003e","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-6515437/v1/506bfd4f1aef65df34c96482.png"},{"id":81939870,"identity":"55cd8e82-175a-49a3-928c-0ff15605416e","added_by":"auto","created_at":"2025-05-05 06:51:08","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":490715,"visible":true,"origin":"","legend":"\u003cp\u003eIn the metaphysis, N-cad⁺ MSCs support quiescent leukemic stem cells (LSCs) by creating CXCL12-rich niches. These MSCs express GPC3, which inhibits DPP4 on LSCs, preserving local CXCL12 levels and reinforcing LSC retention and quiescence. In the central marrow, the absence of N-cad⁺ MSCs allows DPP4 to degrade CXCL12, forming a CXCL12 gradient from the central marrow to the metaphysis. This gradient drives LSC migration toward the metaphysis, where they become quiescent, while central marrow LSCs remain active and proliferative. Systemically, low CXCL12 in the marrow and high levels in peripheral tissues (e.g., liver, spleen) promote LSC dissemination into circulation.\u003c/p\u003e","description":"","filename":"floatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-6515437/v1/fdc709958aba00f08a41290f.png"},{"id":107776425,"identity":"0c99fc94-d7a2-4c52-ab25-94a48e516416","added_by":"auto","created_at":"2026-04-25 07:10:45","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":6111650,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6515437/v1/e37aa8a6-c96f-491a-8003-fffc9fd687bd.pdf"},{"id":81939395,"identity":"b42ebd7c-6dd2-495d-a87e-625dd62ed58e","added_by":"auto","created_at":"2025-05-05 06:43:07","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":2309428,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryFigures.docx","url":"https://assets-eu.researchsquare.com/files/rs-6515437/v1/9fc049503a0cf15cfde25950.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Longitudinal Localization of Leukemia Stem Cells Between Metaphysis and Central Marrow Governs Leukemic Stem Cell Behavior","fulltext":[{"header":"Introduction","content":"\u003cp\u003eAcute myeloid leukemia (AML) is an aggressive hematopoietic malignancy driven by leukemic stem cells (LSCs), which are central to disease initiation, progression, and relapse. Thus, targeting LSCs has emerged as a promising strategy to combat AML effectively. LSC behavior is governed by two critical factors: their trafficking through circulation and their localization within specialized bone marrow (BM) niches. However, despite extensive studies on LSC-niche interactions and significant efforts to translate these insights into clinical therapies\u003csup\u003e1\u0026ndash;3\u003c/sup\u003e, no effective treatment has been developed to restrict leukemia dissemination or eradicate LSCs. This highlights an incomplete understanding of the complex interplay between LSCs and their niches.\u003c/p\u003e \u003cp\u003ePrevious research has extensively characterized how niche cells regulate LSCs through single-direction communication\u003csup\u003e3,4\u003c/sup\u003e. However, the dynamic cross talk between LSCs and niche cells in co-creating a leukemia-supportive microenvironment remains poorly understood. Additionally, two BM niches have been well-described in regulating LSC behavior: the endosteal niche, which promotes LSC quiescence and survival, and the sinusoidal niche, which facilitates LSC proliferation and dissemination\u003csup\u003e5\u003c/sup\u003e. The endosteal niche comprise two anatomically distinct regions: the metaphysis (trabecular bone) and the diaphysis (cortical bone). While interactions between LSCs and the diaphyseal endosteum at a local transverse scale (diaphysis vs. sinusoid) have been well-studied, the broader longitudinal scale (metaphysis vs. sinusoid) remains underexplored.\u003c/p\u003e \u003cp\u003eCXCL12 is a well-established chemoattractant for leukemic cells\u003csup\u003e6,7\u003c/sup\u003e, playing a crucial role in their migration and localization. DPP4, a key regulator of CXCL12 activity\u003csup\u003e8,9\u003c/sup\u003e, has been implicated in promoting AML progression\u003csup\u003e10\u003c/sup\u003e. While GPC3 is recognized as an inhibitor of DPP4\u003csup\u003e11,12\u003c/sup\u003e, its specific role in leukemia development remains unexplored.\u003c/p\u003e \u003cp\u003eIn this study, we identified the metaphysis and central marrow as two key LSC niche areas, emphasizing the crucial role of N-cadherin-expressing mesenchymal stromal/stem cells (N-cad\u003csup\u003e+\u003c/sup\u003e MSCs)\u003csup\u003e13\u0026ndash;15\u003c/sup\u003e in the metaphysis for maintaining LSCs. Furthermore, we uncovered a three-way crosstalk between the BM matrix, N-cad\u003csup\u003e+\u003c/sup\u003e MSCs, and LSCs through the CXCL12-DPP4-GPC3 axis that governs LSC behavior. Targeting DPP4 mobilized LSCs from the metaphysis to the central marrow niche, disrupting their protective microenvironment, inducing exhaustion, and limiting their dissemination into peripheral blood. This therapeutic approach significantly enhances the efficacy of targeting LSCs, providing a novel avenue for AML treatment.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e \u003cb\u003eDpp4\u003c/b\u003e \u003cb\u003edeficiency re-distributes AML cells and induces LSC exhaustion\u003c/b\u003e\u003c/p\u003e \u003cp\u003eOur group recently identified DPP4 as a promising therapeutic target for AML due to its selectively high expression in AML cells and its role in blocking AML progression upon depletion\u003csup\u003e10\u003c/sup\u003e. Strikingly, our experiments revealed that \u003cem\u003eDpp4\u003c/em\u003e deletion alters the spatial distribution of AML cells within the BM, as observed in whole BM sections (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea-b). In control AML cells transplanted mice (\u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e AML mice; 2 weeks post-transplantation), AML cells predominantly localize around the metaphysis area, especially the proximal metaphysis (PM). However, in \u003cem\u003eDpp4\u003c/em\u003e KO AML cells transplanted mice (\u003cem\u003eDpp4\u003c/em\u003e\u003csup\u003e\u0026minus;/\u0026minus;\u003c/sup\u003e AML mice; 12 weeks post-transplantation), AML cells exhibited a more even distribution throughout the BM, with increased centralization towards the central marrow (CM) in both MLL-AF9 (MA9) and AML-ETO9a (AE9) models (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea-b). To eliminate potential bias due to differences in AML progression since \u003cem\u003eDpp4\u003c/em\u003e KO AML mice rarely reach moribund stage and sustain minimal AML cells in peripheral blood (PB), we transplanted 1x10\u003csup\u003e6\u003c/sup\u003e control or \u003cem\u003eDpp4\u003c/em\u003e KO primary AML cells (GFP\u003csup\u003e+\u003c/sup\u003e) into C57/B6 recipients and compared AML distribution within the BM when PB GFP\u003csup\u003e+\u003c/sup\u003e cells reach\u0026thinsp;~\u0026thinsp;30% (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec). Notably, this approach controlled for tumor burden, ruling out the possibility that redistribution was secondary to delayed AML progression in \u003cem\u003eDpp4\u003c/em\u003e KO mice. Consistently, \u003cem\u003eDpp4\u003c/em\u003e KO resulted in a similar redistribution of AML cells across the BM. This redistribution correlated with a significant increase in LSC populations (L-GMPs: GFP\u003csup\u003e+\u003c/sup\u003eIL-7R\u003csup\u003e\u0026minus;\u003c/sup\u003eLin\u003csup\u003e\u0026minus;\u003c/sup\u003eSca-1\u003csup\u003e\u0026minus;\u003c/sup\u003ec-Kit\u003csup\u003e+\u003c/sup\u003eCD34\u003csup\u003e+\u003c/sup\u003eFcγRII/III\u003csup\u003e+\u003c/sup\u003e) in the BM of MA9 model (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed-f). Additionally, L-GMPs in \u003cem\u003eDpp4\u003c/em\u003e KO mice exhibited increased cell division (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eg) and higher apoptosis (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eh), suggesting an exhaustion phenotype.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eThe PM (proximal metaphysis) and DM (distal metaphysis) mark distinct niches versus CM for maintaining LSCs\u003c/b\u003e \u003c/p\u003e \u003cp\u003eWhile we observed that L-GMPs exhibit reduced quiescence and increased exhaustion-like behavior after \u003cem\u003eDpp4\u003c/em\u003e KO, it is not clear that whether such phenomenon results from the intracellular effects of \u003cem\u003eDpp4\u003c/em\u003e KO or is associated with altered LSC niche. To understand this, we evaluated L-GMPs across the three anatomical partitions (PM, DM, and CM) of the BM in control AML mice. Our analysis revealed that the PM and DM harbored a significantly higher percentage of L-GMPs compared to the CM (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea-b). Among the L-GMPs, in each niche, those derived from the PM and DM displayed lower cell division activity (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec-d). To functionally assess LSC from PM/DM vs CM niches, we transplanted L-GMPs from the PM, DM or CM into C57/B6 recipient, and found those from PM and DM induced a more aggressive AML progression (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ee). Furthermore, limiting dilution assays show PM/DM-derived AML cells exhibited higher stemness (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ef). These findings strongly suggest that the PM and DM, as opposed to the CM, constitute specialized LSC-supporting niches that promote LSC maintenance and functionality.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eDpp4\u003c/b\u003e \u003cb\u003edeficiency confines AML cells within the BM through reversed BM-to-PB CXCL12 gradient\u003c/b\u003e\u003c/p\u003e \u003cp\u003eAnother hallmark finding in \u003cem\u003eDpp4\u003c/em\u003e KO AML mice is the dramatic decrease in peripheral AML cells and an enrichment of AML cells within the BM after \u003cem\u003eDpp4\u003c/em\u003e deletion in both MA9 and AE9 models (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea). To investigate whether \u003cem\u003eDpp4\u003c/em\u003e KO AML cells exhibit impaired homing or show reduced peripheralization despite successful homing, we compared the homing efficiency of control and \u003cem\u003eDpp4\u003c/em\u003e KO AML cells 16 hours post-transplantation. Our findings indicate comparable homing of control and \u003cem\u003eDpp4\u003c/em\u003e KO AML cells to all hematopoietic organs (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb), suggesting that \u003cem\u003eDpp4\u003c/em\u003e deletion does not impair homing but instead restricts AML cells within the host BM. This restriction persists even in the late stage of AML development (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea) and resulted in prolonged mouse survival (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eHowever, this finding challenges the existing knowledge that the CM, rich in sinusoids, serves as a crucial niche for LSCs and hematopoietic stem cells (HSCs) trafficking to the PB\u003csup\u003e4,5\u003c/sup\u003e. To explore the mechanism behind this intriguing observation, we next focused on MA9 AML model. We first hypothesized that \u003cem\u003eDpp4\u003c/em\u003e KO might impair AML cells' access to blood vessels. However, immunofluorescence analysis revealed that both GFP-labeled control and \u003cem\u003eDpp4\u003c/em\u003e KO AML cells were equidistant from endomucin-labeled vessels throughout the BM, indicating unimpeded vascular access (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ed-f)\u003csup\u003e16,17\u003c/sup\u003e. This prompted us to hypothesize that altered cytokine / chemokine gradients might be responsible for the observed AML cell confinement. We focused on DPP4 substrates and cytokines or chemokines implicated in AML cell trafficking \u003csup\u003e18\u0026ndash;22\u003c/sup\u003e. Among the factors tested, including CCL11, CCL22, CXCL12, IL-6, IL-15, etc., only CXCL12 exhibited a reversed gradient from plasma to BM in \u003cem\u003eDpp4\u003c/em\u003e KO mice (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eg-j). Despite prior reports implicating G-CSF and NPY in HSPC mobilization\u003csup\u003e23\u0026ndash;25\u003c/sup\u003e, their levels were comparable between BM extracellular fluid (BMEF) and plasma in our model, excluding their role in the observed phenotype. Exogenous CXCL12 administration mobilized both control and \u003cem\u003eDpp4\u003c/em\u003e KO AML cells at high doses (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ek), but \u003cem\u003eDpp4\u003c/em\u003e KO cells showed tolerance at lower doses. This was not due to altered CXCR4 expression (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003el), or migration capacity \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003em-n\u003cb\u003e)\u003c/b\u003e, suggesting that the reversed gradient itself\u0026mdash;not cell-intrinsic changes\u0026mdash;drives confinement.\u003c/p\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003cp\u003e\u003cstrong\u003eNiche-specific CXCL12 production by N-cad\u003csup\u003e\u0026nbsp;⁺\u003c/sup\u003e MSCs dictates LSC spatial distribution\u003c/strong\u003e\u003c/p\u003e\u003cp\u003eOur previous findings demonstrated that a reversed BM-to-PB CXCL12 gradient confines \u003cem\u003eDpp4\u003c/em\u003e KO AML cells to the BM. To dissect the mechanism driving AML redistribution, we quantified CXCL12 levels across BM subregions. Strikingly, \u003cem\u003eDpp4\u003c/em\u003e KO mice exhibited a reversed PM/DM-CM CXCL12 gradient, with low CXCL12 in the CM of controls but elevated levels in \u003cem\u003eDpp4\u003c/em\u003e KO mice (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea\u0026ndash;b). This suggested compartmentalized CXCL12 regulation by DPP4, prompting us to identify its cellular source.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo identify the specific niche cells secreting CXCL12, we performed single-cell RNA sequencing (scRNA-seq) on non-hematopoietic PM and DM cells from AML mice three weeks post-transplantation. Through the analysis of 11,512 cells (median of 19,367.5 molecules and 4,391 genes per cell), we identified 14 distinct clusters, spanning mesenchymal stem/stromal cells (MSC), osteolineage cells (OLC), chondrocytes (Chondro) and chondrocytes of possible transitional states (c/r, cycling/resting; pro, progenitor), fibroblasts (Fibro), endothelial cells (EC), pericytes (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec). Consistent with prior reports, \u003cem\u003eCxcl12\u003c/em\u003e was predominantly expressed by the MSC cluster (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ed)\u003csup\u003e1\u003c/sup\u003e. Given the heterogeneity of MSCs, we compared Cxcl12 expression across known subpopulations marked by LepR, N-cadherin (\u003cem\u003eCdh2\u003c/em\u003e), Prx-1, Osx (\u003cem\u003eSp7\u003c/em\u003e), and Nestin \u003csup\u003e13,26\u0026ndash;31\u003c/sup\u003e. Transcriptomic association analysis revealed that \u003cem\u003eCxcl12\u003c/em\u003e was most strongly correlated with \u003cem\u003eLepr⁺\u003c/em\u003e and \u003cem\u003eCdh2⁺\u003c/em\u003e MSCs (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ee). LepR\u003csup\u003e+\u003c/sup\u003e MSCs are known as a heterogenous population of MSCs spanning across the BM, including metaphysis (but mainly perivascular localization in metaphysis\u003csup\u003e28\u003c/sup\u003e) and CM region. The N-cad protein is predominantly detected in the endosteum of metaphysis\u003csup\u003e14\u003c/sup\u003e. In contrast, other MSC subpopulations ( Prx-1\u003csup\u003e+\u003c/sup\u003e, Osx\u003csup\u003e+\u003c/sup\u003e, and Nes\u003csup\u003e+\u003c/sup\u003e MSCs), were minimally represented in our dataset (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ef), aligning with prior studies showing Prx-1\u003csup\u003e+\u003c/sup\u003e MSCs enriched in periosteal region\u003csup\u003e32,33\u003c/sup\u003e, Osx\u003csup\u003e+\u003c/sup\u003e MSCS enrich in the bone tissue\u003csup\u003e34\u003c/sup\u003e, and Nes\u003csup\u003e+\u003c/sup\u003e MSCs enriched in perivascular region\u003csup\u003e31\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eCritically, despite prior reports implicating Nestin\u003csup\u003e+\u003c/sup\u003e MSCs in AML chemoresistance\u003csup\u003e35\u003c/sup\u003e their \u003cem\u003eCxcl12\u003c/em\u003e expression was negligible \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ef\u003cb\u003e)\u003c/b\u003e To rule out functional redundancy, we deleted \u003cem\u003eCxcl12\u003c/em\u003e in Nestin\u003csup\u003e+\u003c/sup\u003e cells (\u003cem\u003eNestin-CreER; Cxcl12\u003c/em\u003e\u003csup\u003e\u003cem\u003efl/fl\u003c/em\u003e\u003c/sup\u003e), and observed no impact on AML progression or BM retention \u003cb\u003e(Supplementary Fig.\u0026nbsp;1a-b)\u003c/b\u003e, confirming Nestin\u003csup\u003e+\u003c/sup\u003e MSCs are dispensable for CXCL12-mediated LSC maintenance in our model.\u003c/p\u003e \u003cp\u003eN-cad⁺ MSCs co-expressed \u003cem\u003eCxcl12\u003c/em\u003e with other LSC-supportive factors (e.g., \u003cem\u003eGas6\u003c/em\u003e, \u003cem\u003eAngpt1\u003c/em\u003e, \u003cem\u003eKitl\u003c/em\u003e; Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eg) \u003csup\u003e4,36\u0026ndash;38\u003c/sup\u003e. Given reports that SCF (\u003cem\u003eKitl)\u003c/em\u003e promotes AML adhesion and survival\u003csup\u003e39\u003c/sup\u003e, we rigorously tested its role. Conditional deletion of \u003cem\u003eScf\u003c/em\u003e in N-cad\u003csup\u003e+\u003c/sup\u003e cells (\u003cem\u003eN-cad-CreER; Scf\u003c/em\u003e \u003csup\u003e\u003cem\u003efl/fl\u003c/em\u003e\u003c/sup\u003e) yielded no difference in mice survival, AML cell engraftment and distribution in the BM (\u003cb\u003eSupplementary Fig.\u0026nbsp;2a\u0026ndash;e\u003c/b\u003e). This negative result underscores that the observed niche effects are CXCL12-specific and not confounded by SCF.\u003c/p\u003e \u003cp\u003eTo spatially assess the relationship between LSCs and N-cad\u003csup\u003e+\u003c/sup\u003e cells, we transplanted control L-GMPs into N-cad-tdTomato (N-cad-TdT) reporter mice, where Tomato\u003csup\u003e+\u003c/sup\u003e cells mark \u003cem\u003eN-cadherin\u003c/em\u003e expression. AML cells (GFP⁺) and \u003cem\u003eN-cad⁺\u003c/em\u003e stromal cells were co-enriched in the proximal metaphysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eh). Strikingly, LSCs (GFP⁺ Kit⁺) were significantly more likely to reside within 5 \u0026micro;m of \u003cem\u003eN-cad⁺\u003c/em\u003e cells (45.3% \u0026plusmn; 5.4%) compared to differentiated AML cells (GFP⁺ Kit⁻; 10.6% \u0026plusmn; 2.2%) (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eh, \u003cb\u003ebottom right\u003c/b\u003e), suggesting active niche-LSC crosstalk.\u003c/p\u003e \u003cp\u003eTo test whether N-cad\u003csup\u003e+\u003c/sup\u003e-derived CXCL12 drives LSC retention, we conditionally deleted \u003cem\u003eCxcl12\u003c/em\u003e in N-cad\u003csup\u003e+\u003c/sup\u003e cells (\u003cem\u003eN-cad-CreER; Cxcl12\u003c/em\u003e\u003csup\u003e\u003cem\u003efl/fl\u003c/em\u003e\u003c/sup\u003e) (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ei) and transplanted control L-GMPs (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ej). N-cad; \u003cem\u003eCxcl12\u003c/em\u003e\u003csup\u003e\u003cem\u003e\u0026minus;/\u003c/em\u003e\u0026minus;\u003c/sup\u003e mice recapitulated the \u003cem\u003eDpp4\u003c/em\u003e KO phenotype, with AML cells redistributed from PM/DM to CM (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ek-l).\u003c/p\u003e \u003cp\u003e \u003cb\u003eShared LSC behavior and transcriptomic Signatures in N-cad;\u003c/b\u003e \u003cb\u003eCxcl12\u003c/b\u003e\u003csup\u003e\u003cb\u003e\u0026minus;/\u0026minus;\u003c/b\u003e\u003c/sup\u003e \u003cb\u003eand\u003c/b\u003e \u003cb\u003eDpp4\u003c/b\u003e \u003cb\u003eKO AML mice\u003c/b\u003e\u003c/p\u003e \u003cp\u003eThe redistribution of LSCs from PM/DM to CM in \u003cem\u003eN-cad; Cxcl12\u003c/em\u003e\u003csup\u003e\u003cem\u003e\u0026minus;/\u0026minus;\u003c/em\u003e\u003c/sup\u003e mice mirrored the phenotype observed in \u003cem\u003eDpp4\u003c/em\u003e KO AML, prompting us to investigate whether these models shared LSC properties and underlying molecular mechanisms. CM-localized AML cells from \u003cem\u003eN-cad; Cxcl12\u003c/em\u003e\u003csup\u003e\u003cem\u003e\u0026minus;/\u0026minus;\u003c/em\u003e\u003c/sup\u003e mice exhibited hallmark features of LSC exhaustion, including increased cell cycle activity and apoptosis, alongside impaired self-renewal capacity (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea\u0026ndash;d). These phenotypic changes were accompanied by delayed AML progression, as evidenced by prolonged survival and reduced organ infiltration (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ee-f).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo determine whether the shared spatial redistribution phenotype reflected convergent molecular reprogramming, we performed transcriptomic analyses. Principal component analysis (PCA) revealed striking similarity between \u003cem\u003eDpp4\u003c/em\u003e KO AML cells and control AML cells from \u003cem\u003eN-cad; Cxcl12\u003c/em\u003e\u003csup\u003e\u003cem\u003e\u0026minus;/\u0026minus;\u003c/em\u003e\u003c/sup\u003e recipients, with both populations clustering distinctly from their respective controls (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eg). RNA sequencing further demonstrated consistent transcriptional changes, with upregulation of cell cycle and metabolic genes and downregulation of stemness-associated pathways and migration (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eh\u0026ndash;i).\u003c/p\u003e \u003cp\u003eTogether, these findings demonstrate that disruption of the CXCL12-DPP4 axis induces convergent transcriptomic rewiring that drives LSC exhaustion and impairs disease progression.\u003c/p\u003e \u003c/div\u003e\n\u003cp\u003e\u003cstrong\u003eN-cad\u003csup\u003e+\u003c/sup\u003e cells support LSCs through GPC3 mediated attraction\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWhile we have shown that DPP4, by deactivating CXCL12, specifically create an intra-BM CXCL12 gradient that favors LSC localization to the PM/DM for maintenance, two key questions remain unanswered: 1) Why does DPP4 selectively deactivate CXCL12 in the CM while preserving CXCL12 in the PM and DM in control AML mice (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb)? 2) How do control LSCs achieve close proximity to N-cad\u003csup\u003e+\u003c/sup\u003e niche cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eh)?\u003c/p\u003e \u003cp\u003ePrior studies have identified CXCL12 hotspots as critical HSC niches that attract HSCs to close proximity for their maintenance\u003csup\u003e40\u003c/sup\u003e. Consistent with this, our scRNA-seq analysis revealed that N-cad\u003csup\u003e+\u003c/sup\u003e cells express high levels of \u003cem\u003eCxcl12\u003c/em\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ee-f). To further investigate, we compared the spatial relationship between LSC-enriched populations (GFP\u003csup\u003e+\u003c/sup\u003e Kit\u003csup\u003e+\u003c/sup\u003e) and N-cad\u003csup\u003e+\u003c/sup\u003e cells in control and \u003cem\u003eDpp4\u003c/em\u003e KO AML mice. Interestingly, \u003cem\u003eDpp4\u003c/em\u003e KO LSCs resided significantly farther from N-cad\u003csup\u003e+\u003c/sup\u003e cells than control LSCs, with only 10.5% \u0026plusmn; 2.1% of \u003cem\u003eDpp4\u003c/em\u003e KO LSCs located within 5 \u0026micro;m of N-cad\u003csup\u003e+\u003c/sup\u003e cells, compared to 58.6% \u0026plusmn; 7.3% of control LSCs (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eh, \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ea-b). Additionally, in the PM/DM, \u003cem\u003eDpp4\u003c/em\u003e KO L-GMPs exhibited exhaustion-like behavior (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ec), underscoring the importance of close proximity to N-cad\u003csup\u003e+\u003c/sup\u003e cells for LSC maintenance. As such, we hypothesized that an N-cad\u003csup\u003e+\u003c/sup\u003e cell-derived factor interacts with DPP4 to establish the CXCL12 gradient at both the macroscopic (PM/DM-CM) and the microscale level. To test this, we conducted transcriptional profiling of 10 known CXCL12 or DPP4 regulators (e.g., DPP8, elastase, MMPs, cathepsin G, TFPI, and GPC3)\u003csup\u003e11,41,42\u003c/sup\u003e in both AML cells and N-cad\u003csup\u003e+\u003c/sup\u003e cells from the PM/DM. Notably, Glypican-3 (GPC3) was highly expressed in N-cad\u003csup\u003e+\u003c/sup\u003e cells compared to other factors (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ed-e). GPC3 is known as an inhibitor of DPP4, suggesting its potential role in suppressing DPP4 activity and preserving CXCL12 near N-cad\u003csup\u003e+\u003c/sup\u003e cells. Indeed, in vitro experiments demonstrated that GPC3 binds to DPP4 on AML cells and inhibits DPP4 enzymatic activity (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ef-g). Immunostaining further revealed significantly higher GPC3 expression in N-cad\u003csup\u003e+\u003c/sup\u003e cells (79.8% \u0026plusmn; 3.3%) compared to N-cad\u003csup\u003e\u0026minus;\u003c/sup\u003e cells (10.5% \u0026plusmn; 2.1%) in the PM region (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eh-i), with 73.5% of GPC3 and DPP4 co-localized between N-cad\u003csup\u003e+\u003c/sup\u003e cells and LSCs (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ej). These data indicate that GPC3, expressed by N-cad\u003csup\u003e+\u003c/sup\u003e cells, inhibits DPP4 and sustains CXCL12 at the microscale, thereby attracting LSCs to CXCL12 hotspots for their maintenance. Additionally, at the macroscopic level, GPC3-mediated DPP4 inhibition contributes to the intra-BM CXCL12 gradient between the PM/DM and CM, further facilitating LSC localization to supportive niche environments.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eGpc3\u003c/b\u003e \u003cb\u003eknockout in N-cad\u003c/b\u003e\u003csup\u003e\u003cb\u003e+\u003c/b\u003e\u003c/sup\u003e \u003cb\u003eredistributes AML cells and impairs LSCs\u003c/b\u003e\u003c/p\u003e \u003cp\u003eTo functionally assess the role of the GPC3-DPP4 interaction in LSC maintenance, we conditionally deleted \u003cem\u003eGpc3\u003c/em\u003e from N-cad\u003csup\u003e+\u003c/sup\u003e cells (N-cad; \u003cem\u003eGpc3\u003c/em\u003e\u003csup\u003e\u0026minus;/\u0026minus;\u003c/sup\u003e) and transplanted LSCs into both N-cad; \u003cem\u003eGpc3\u003c/em\u003e\u003csup\u003e\u0026minus;/\u0026minus;\u003c/sup\u003e and control mice (N-cad; \u003cem\u003eGpc3\u003c/em\u003e\u003csup\u003e+/+\u003c/sup\u003e) (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ea). Loss of GPC3 in N-cad\u003csup\u003e+\u003c/sup\u003e cells led to a significant reduction in CXCL12 levels in the PM and DM (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eb), resulting in an altered AML cell distribution pattern similar to that observed in \u003cem\u003eDpp4\u003c/em\u003e KO and N-cad; \u003cem\u003eCxcl12\u003c/em\u003e\u003csup\u003e\u0026minus;/\u0026minus;\u003c/sup\u003e mice (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ec). This redistribution was accompanied by LSC exhaustion, as evidenced by increased L-GMPs division and apoptosis (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ed-e). The disruption of the CXCL12 gradient and loss of proximity to supportive N-cad\u003csup\u003e+\u003c/sup\u003e cells impaired LSC maintenance, ultimately prolonging mouse survival (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ef), confirming the role of GPC3 in regulating LSC localization.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eOur study identified two pivotal findings: the redistribution of LSCs, leading to their exhaustion, and the confinement of AML cells within the BM. Both phenomena are fundamentally mediated by CXCL12, a key chemoattractant for AML cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e)\u003csup\u003e6,7\u003c/sup\u003e. While CXCL12's role in LSC biology has been controversial\u003csup\u003e43\u0026ndash;46\u003c/sup\u003e, our data reconcile these discrepancies by demonstrating that spatial compartmentalization\u0026mdash;not absolute CXCL12 levels\u0026mdash;governs LSC fate. Previous in vitro studies suggest that CXCL12 supports LSC survival, either independently or via MSCs\u003csup\u003e45,47,48\u003c/sup\u003e, while in vivo evidence remained conflicting. For instance, studies utilizing mouse models have shown that knocking out \u003cem\u003eCxcl12\u003c/em\u003e in Prx-expressing MSCs or Tek-expressing endothelial cells had no significant effect on LSC survival in MLL-AF9-driven AML models\u003csup\u003e43\u003c/sup\u003e. Conversely, other work has demonstrated that global or conditional deletion of \u003cem\u003eCxcl12\u003c/em\u003e in chronic myeloid leukemia (CML) models enhanced survival and chemotherapy efficacy\u003csup\u003e2,49\u003c/sup\u003e. Our work resolves this paradox by demonstrating that the functional impact of CXCL12 depends critically on its anatomical source and spatial gradient within the BM niche.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eA key advance of our study is the identification of N-cadherin\u003csup\u003e+\u003c/sup\u003e MSCs as the functionally relevant CXCL12 source for LSC maintenance. This contrasts with previous work focusing on Prx\u003csup\u003e+\u003c/sup\u003e periosteal \u003csup\u003e32,33\u003c/sup\u003e or Tek\u003csup\u003e+\u003c/sup\u003e endothelial niches\u003csup\u003e43\u003c/sup\u003e, which our scRNA-seq revealed are minimally involved in metaphyseal LSC maintenance (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ef). Other well-known MSC markers, such as Lepr, Nestin, and Osx, are either widespread expression across the BM\u003csup\u003e28\u003c/sup\u003e, enriches in perivascular region\u003csup\u003e31\u003c/sup\u003e, or localized in the bone tissue\u003csup\u003e34\u003c/sup\u003e. Importantly, we confirmed through genetic deletion that neither Nestin\u003csup\u003e+\u003c/sup\u003e MSC-derived CXCL12 (despite their reported role in chemoresistance) nor N-cad\u003csup\u003e+\u003c/sup\u003e MSC-derived SCF contribute meaningfully to the phenotypes we observed (\u003cb\u003eSupplementary Figs.\u0026nbsp;1\u0026ndash;2\u003c/b\u003e), underscoring the specificity of the N-cadherin\u003csup\u003e+\u003c/sup\u003e MSC-CXCL12 axis in our model.\u003c/p\u003e \u003cp\u003eThe discovery of a reverse PM/DM-CM CXCL12 gradient in control AML mice provides a mechanistic basis for LSC niche specificity. This gradient depends on regional regulation of DPP4 activity: while DPP4 degrades CXCL12 in the CM, GPC3 from N-cad\u003csup\u003e+\u003c/sup\u003e MSCs inhibits DPP4 in the PM/DM, creating protected CXCL12 microdomains. This explains why prior studies using high-resolution mapping in normal hematopoiesis failed to detect long-range CXCL12 gradients\u003csup\u003e40\u003c/sup\u003e - the AML-specific expression of DPP4 creates a pathological gradient not present in steady state\u003csup\u003e50\u003c/sup\u003e. Collectively, our work identifies the CXCL12-DPP4-GPC3 axis as a master regulator of LSC niche interactions, where: GPC3 from N-cad\u003csup\u003e+\u003c/sup\u003e MSCs locally preserves CXCL12 to anchor LSCs in protective metaphyseal niches, DPP4 from AML cells degrades CXCL12 systemically to enable dissemination.\u003c/p\u003e \u003cp\u003eThis axis offers immediate clinical promise, as DPP4 inhibitors could exploit this mechanism: (1) by reversing the BM-PB CXCL12 gradient to confine AML cells, reducing life-threatening extramedullary complications\u0026mdash;such as leukocytosis, clotting abnormalities, respiratory distress, and stroke\u0026mdash;thereby reducing leukemia-associated morbidity\u003csup\u003e51,52\u003c/sup\u003e ; (2) by disrupting protective niche interactions to sensitize LSCs to chemotherapy; and (3) through direct induction of LSC exhaustion. The established safety profile of DPP4 inhibitors (e.g., sitagliptin) underscores the translational potential of this approach\u003csup\u003e10,53\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eImportant remaining questions include whether other DPP4 substrates contribute to these effects, and how metabolic changes in redistributed LSCs influence their exhaustion. Nevertheless, our work establishes the CXCL12-DPP4-GPC3 axis as a critical regulator of LSC spatial organization and viability, providing a strong rationale for niche-targeted therapies in AML (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e).\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eMice\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eC57BL/6 mice were obtained from Charles River, Inc. DPP4 \u003csup\u003eflox/flox\u003c/sup\u003e mice were generated by breeding targeted C57Bl/6NTac-DPP4tm1a Wtsi/Ics mice (European Mouse Mutant Cell Repository, EUCOMM) with 129S4/Bl6-Gt (ROSA) 26Sortm2(FLP*) Sor/J (stock #012930, The Jackson Laboratory, Bar Harbor, ME). The offspring were further crossed with Vav-iCre mice (stock #018968, The Jackson Laboratory, Bar Harbor, ME) \u003csup\u003e54\u003c/sup\u003e, to generate DPP4 \u003csup\u003efl/fl\u003c/sup\u003e；Vav-Cre mice\u003csup\u003e10\u003c/sup\u003e. N-cad-tdTomato (N-cad-TdT) and N-cad-CreER strains were generated by Dr. Linheng Li’s lab\u003csup\u003e13,15\u003c/sup\u003e. \u003cem\u003eCxcl12\u003c/em\u003e \u003csup\u003efl/fl\u003c/sup\u003e were purchased from Jackson Lab. To induce expression of Cre-ER recombinase, mice received tamoxifen via intraperitoneal injection (Sigma, 75 mg tamoxifen/kg body weight) as described\u0026nbsp;\u003csup\u003e55\u003c/sup\u003e. All mouse strains used in this study had a C57BL/6J genetic background. Animals were randomly assigned to experimental groups based on genotyping results. Investigators were blinded to group allocation during data analysis but not during experimental procedures. Sample sizes for each experiment are detailed in figure legends. All animal procedures were conducted in accordance with protocols approved by the Institutional Animal Care and Use Committee (IACUC) of the University of Missouri.\u003c/p\u003e\n\u003cp\u003eAML transplantation\u003c/p\u003e\n\u003cp\u003eFor Fig. 1a, we transplanted infected \u003cem\u003eDpp4\u003c/em\u003e KO or control Lin\u003csup\u003e−\u003c/sup\u003e cells (2 x 10\u003csup\u003e5\u003c/sup\u003e; infection efficiency consistently 40-50%) into lethally irradiated (1,000 cGy) C57/B6 mice (6–8 weeks old) and evaluated 2-50 weeks after transplantation. For other \u003cem\u003eDpp4\u003c/em\u003e KO vs control AML models, 1 × 10\u003csup\u003e6\u003c/sup\u003e primary AML cells were transplanted C57/B6 recipients. For homing assay, GFP\u003csup\u003e+\u003c/sup\u003e AML cells were evaluated at 16 hours. Engraftment was assessed 2- and 4-weeks post-transplantation for MLL-AF9 and AML-ETO9a models, respectively. For control AML cells transplantation into N-cad\u003csup\u003e+\u003c/sup\u003e mice with or without \u003cem\u003eCxcl12\u003c/em\u003e or \u003cem\u003eGpc3\u003c/em\u003e, 500 L-GMPs were transplanted into sublethally irradiated (500cGy) mice and engraftment assessed at 2-8 weeks post-transplantation. To preserve niche cell integrity, all mice were sub-lethally irradiated (500 cGy) before transplantation.\u003c/p\u003e\n\u003cp\u003eFlow cytometry\u003c/p\u003e\n\u003cp\u003ePeripheral blood (PB), bone marrow (BM), spleen, and liver hematopoietic cells were labeled with the following antibodies (all from BioLegend unless otherwise noted): Anti-CD3e-PE/Cyanine5 (#100310, 1:200), Anti-Ly-6G/Ly-6C (Gr-1)-PE/Cyanine5 (#108410, 1:200), Anti-CD11b-PE/Cyanine5 (#101210, 1:200), Anti-CD45R-PE/Cyanine5 (#103210, 1:200), Anti-Ter-119-PE/Cyanine5 (#116210, 1:200), Anti-CD117 (c-Kit)-APC (#105812, 1:200), Anti-Sca-1-PE-Cy7 (#108114, 1:200), Anti-CD150-PE (#115904, 1:200), Anti-CD48-APC/Cyanine7 (#103432, 1:200), Anti-Ki67-FITC (#652410, 1:200), Hoechst 34580 (BD Pharmingen, #565877), Anti-CD16/32-PE (#101308, 1:200), Anti-CD34-FITC (eBioscience, #11-0341-82, 1:200), Anti-CD127-APC/Cyanine7 (#135040, 1:200), Anti-CD135-Brilliant Violet421 (#135314, 1:200), Annexin V (#640941, 1:20), and PI (#421301, 1:50). Intracellular staining was performed using the Foxp3/Transcription Factor Staining Kit (eBioscience) according to the manufacturer’s protocol. Flow cytometry analyses were performed independently in triplicate, with biological replicates from at least five mice per condition. Technical replicates were included for measurement accuracy.\u003c/p\u003e\n\u003cp\u003eImmunofluorescence staining and quantification\u003c/p\u003e\n\u003cp\u003eFemurs were perfused with PBS, fixed with 4% paraformaldehyde,\u0026nbsp;and subjected to frozen sectioning. Antigen retrieval was performed with 1 μg/ml proteinase K in TE buffer (100 mM Tris-HCl, pH 8.0, 50 mM EDTA) at 37°C for 30 minutes. Sections were blocked with Universal Blocking Reagent (BioGenex) and incubated overnight at 4°C with primary antibodies, including anti-Endomucin, biotin-lineage antibody mixture (BioLegend, 1:200), IL-7R, Sca-1 (BioLegend, 1:200), PE-CD34 (BioLegend, 1:200), Alexa Fluor 647-Kit (Invitrogen, 1:200), and anti-GPC3. Secondary staining was performed with donkey anti-goat Alexa Flour® 555 (Invitrogen; 1:500), goat anti-rabbit Alexa Fluor® 750 (Invitrogen; 1:500) and Brilliant violet 421® Streptavidin (Biolegend; 1:500) at room temperature for 1 hour. DAPI stock solution was diluted to 300 nM in PBS and 300 mL was added to the coverslip preparation for 1 minute. Sections were rinsed 3 times in PBS, excess buffer drained from the coverslip and mounted with Shandon™ Immu-Mount™ (Fisher Scientific). Image stitching was done to capture the entire specimen at high magnification and seamlessly create a single high-resolution image. Sections were imaged using a Keyence BZ-X800 fluorescence microscope at 20× magnification (resulting in 200× magnification) and 60× magnification (resulting in 600× magnification). Quantification was performed using Keyence BZ-X800 analyzer software, assessing GFP\u003csup\u003e+\u003c/sup\u003e AML cell distribution across BM areas (PM, CM, DM), L-GMP localization, and apoptosis. Distances between GFP\u003csup\u003e+\u003c/sup\u003e AML cells and N-cad+ cells were measured using a minimum of 100 GFP\u003csup\u003e+\u0026nbsp;\u003c/sup\u003eKit\u003csup\u003e+\u003c/sup\u003e and 80 GFP\u003csup\u003e+\u003c/sup\u003e Kit\u003csup\u003e−\u003c/sup\u003e AML cells per dataset. A minimum of five mice per condition were used for each quantification dataset.\u003c/p\u003e\n\u003cp\u003eCytokine analyses\u003c/p\u003e\n\u003cp\u003eCytokine quantification in plasma and bone marrow extracellular fluid (BMEF) were determined using the LEGENDplex Multi-Analyte Flow Assay Kit (BioLegend, San Diego CA), a bead-based immunoassay that quantifies multiple cytokines simultaneously via flow cytometry. Briefly, a custom mouse cytokine and chemokine panel was employed to measure the concentrations of the designed cytokines/chemokines. The LSRFortessa X-20 Cell Analyzer (BD Biosciences) was used for data acquisition, and results were analyzed using the LEGENDplex Data Analysis software. Assays were performed in 96-well plates following manufacturer protocols, with data recorded using a Fisherbrand microplate photometer. BMEF and plasma were used for ELISA assay collected by using the mouse SDF-1 alpha ELISA Kit (Invitrogen) following the manufacturer’s protocol.\u003c/p\u003e\n\u003cp\u003eMigration assay\u003c/p\u003e\n\u003cp\u003eIn vitro:\u0026nbsp;The Transwell migration assay was utilized to assess cell migration. DPP4\u003csup\u003e+/+\u003c/sup\u003e and DPP4\u003csup\u003e−/−\u003c/sup\u003e AML cells were cultured and seeded in serum-free medium into the upper chamber of Transwell inserts (Corning, Inc.) with an 8 μm pore-size Matrigel-coated membrane. The upper wells contained CXCL12 at concentrations of 0 ng/ml. The lower chambers were filled with culture medium containing 100 ng/ml CXCL12. After a 4-hour incubation at 37°C and 5% CO2, non-migratory cells on the upper membrane surface were removed using a cotton swab. Migratory cells on the lower membrane surface were visualized and quantified in multiple random fields under a microscope. Migration rates and statistical significance were analyzed accordingly.\u003c/p\u003e\n\u003cp\u003eIn vivo:\u0026nbsp;Mice (n=5) were intravenously injected with PBS or CXCL12 (0-500 ng/g per mouse as indicated in figure). The percentage of GFP\u003csup\u003e+\u003c/sup\u003e AML cells in peripheral blood was measured before and 17 hours after injection.\u003c/p\u003e\n\u003cp\u003eColony assays\u003c/p\u003e\n\u003cp\u003eMouse AML cells were diluted to the indicated concentration in IMDM with 2% FBS and were then seeded into methylcellulose medium M3534 (STEMCELL Technologies, Cambridge MA) for myeloid colony formation analysis, as previously described\u003csup\u003e56\u003c/sup\u003e. Each colony-forming unit (CFU) assay was performed in triplicate using independent biological replicates for AML cells obtained from distinct mice. The consistency of clone formation was validated across all replicates.\u003c/p\u003e\n\u003cp\u003eDPP4 activity assay\u003c/p\u003e\n\u003cp\u003eDPP4 activity was measured in plasma-EDTA, BMEF, and cell lysates. For each assay, 20 μl of serum or BMEF, or 100 nM of GPC3 protein, was diluted in DPP4 assay buffer (Tris-HCl [pH 8.0], 150 mM NaCl, and protease inhibitor cocktail) in a black 96-well plate to a final volume of 50 μl. An equal volume (50 μl) of 200 mM H-Ala-Pro-AFC substrate (I-1680; Bachem Americas, Torrance, CA) was added, and the plate was incubated for 10 minutes at room temperature in the dark. Fluorescence was measured using a Synergy Microplate Reader at excitation/emission wavelengths of 405/535 nm, and results were reported as relative light units (RLUs).\u003c/p\u003e\n\u003cp\u003eLigand binding assay\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eRecombinant His-tagged GPC3 binding to DPP4+/+ AML cells was assessed as similarly described\u003csup\u003e57\u003c/sup\u003e. Briefly, 1 × 10\u003csup\u003e6\u003c/sup\u003e \u003cem\u003eDpp4\u003csup\u003e+/+\u003c/sup\u003e\u003c/em\u003e AML\u0026nbsp;cells were incubated with or without His-GPC3 (100 nM) in 200 ml PBS/1% BSA for 3 h at 25°C. Nonspecific binding was subtracted. After incubation, cells were washed twice by centrifugation, resuspended in ice-cold PBS/1% BSA, and stained with Alexa Fluor® 488 anti-His Tag Antibody for flow cytometry analysis.\u003c/p\u003e\n\u003cp\u003eQuantitative RT-PCR\u003c/p\u003e\n\u003cp\u003eRT-PCR was performed using 5 ng total RNA, gene-specific primers, and a QIAGEN One Step RT-PCR kit (210210; Qiagen, Germantown, MD) following the manufacturer’s instructions. 18s rRNA was used as an internal control for normalization. The primer sequences used are listed below: TFPI: forward(5′-GGG CTC CGT TCT TGG TCT C-3′) and reverse(5′- TTG AAT CTG CGG CAC TTT TGC-3′), \u0026nbsp;MMP9: forward(5′- CTG GAC AGC CAG ACA CTA AAG-3′) and reverse (5′- CTC GCG GCA AGT CTT CAG AG-3′), MMP2: forward(5′- CAA GTT CCC CGG CGA TGT C -3′) and reverse (5′- TTC TGG TCA AGG TCA CCT GTC-3′), MMP3: forward(5′- ACA TGG AGA CTT TGT CCC TTT TG-3′) and reverse (5′- TTG GCT GAG TGG TAG AGT CCC -3′), MMP13: forward(5′- CTT CTT CTT GTT GAG CTG GAC TC -3′) and reverse (5′- CTG TGG AGG TCA CTG TAG ACT-3′), MMP14: forward(5′- CAG TAT GGC TAC CTA CCT CCA G-3′) and reverse (5′- GCC TTG CCT GTC ACT TGT AAA-3′), cathepsin G: forward(5′- AGG GTT TCT GGT GCG AGA AG-3′) and reverse (5′- GTT CTG CGG ATT GTA ATC AGG AT-3′), Elastase: forward(5′- AGC AGT CCA TTG TGT GAA CGG-3′) and reverse (5′- CAC AGC CTC CTC GGA TGA AG-3′), DPP8: forward (5′- GGG AAA TGG TGA ATC ACA GGA C-3′) and reverse (5′- ATG TAG CCG TGG TAT TTT CTG G-3′), GPC3: forward (5′- CAG CCC GGA CTC AAA TGG G -3′) and reverse (5′- CAG CCG TGC TGT TAG TTG GTA -3′).\u003c/p\u003e\n\u003cp\u003eCell preparation for single-cell RNA-sequencing\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTissue harvesting: Femurs were collected post-euthanasia and immediately placed in ice-cold PBS. The PM and DM regions of the bone were isolated. Bone marrow was extracted by crushing the bones with a mortar and pestle, followed by enzymatic digestion with collagenase/dispase at 37°C for 45 minutes. The resulting cell suspension was washed, lysed, and filtered through a 100 μm strainer. FACS isolation of non-hematopoietic cells: Bone marrow cells were stained for CD45, and viable triple-negative cells were sorted using a BD FACSAria™.\u003c/p\u003e\n\u003cp\u003eSingle cell sequencing\u003c/p\u003e\n\u003cp\u003eSingle cells were encapsulated into emulsion droplets using Chromium Controller (10x Genomics). scRNA-seq libraries were constructed using Chromium Single Cell 3’ v2 Reagent Kit according to the manufacturer’s protocol. Briefly, post sorting sample volume was decreased, and cells were examined under a microscope and counted with a hemocytometer. Cells were then loaded in each channel with a target output of ~4,000 cells. Reverse transcription and library preparation were performed on C1000 Touch Thermal cycler with 96-Deep Well Reaction Module (Bio-Rad). Amplified cDNA and final libraries were evaluated on an Agilent Bioanalyzer using a High Sensitivity DNA Kit (Agilent Technologies). Individual libraries were diluted to 4nM and pooled for sequencing. Pools were sequenced with 75 cycle run kits (26bp Read1, 8bp Index1 and 55bp Read2) on the Novaseq 5000 Sequencing System (Illumina).\u003c/p\u003e\n\u003cp\u003eSingle cell RNA-seq analysis\u003c/p\u003e\n\u003cp\u003eSingle-cell expression was analyzed using the Cell Ranger Single Cell Software Suite (v3.0.2) to perform quality control, sample demultiplexing, barcode processing, and single-cell 3′ gene counting. Sequencing reads were aligned to the refdata-cellranger-mm39-3.0.0 transcriptome using the Cell Ranger suite with default parameters. Single cell data were imported into Seurat 3 package in R statistical software. Only genes detected in at least two cells were considered for further analysis. Apoptotic cells were filtered out by any cell containing \u0026gt; 5% mitochondrial UMI counts. To detect changes in gene signatures more robustly, any cell with \u0026lt;1,500 genes was filtered out. Cell expression levels were normalized and scaled using default settings in Seurat. Principal component analysis for dimensionality reduction was then run on the normalized gene–barcode matrix. TSNE embedding was applied to visualize the cells in the two-dimensional space after selecting the first 20 principal components for the clustering analysis with a resolution parameter of 0.5. Cluster-specific genes were identified by running the Seurat “FindMarkers/FindAllMarkers” function with the Wilcoxon rank-sum test. Plots were made using the Seurat package and R.\u003c/p\u003e\n\u003cp\u003eQuantification and statistical analysis\u003c/p\u003e\n\u003cp\u003eData are expressed as mean±SE mean (SEM). Statistical analyses were performed using GraphPad Prism Version 9.0 (Graph Pad Prism Software Inc, San Diego, CA). For continuous variables, normality and homogeneity of variance were assessed by the Shapiro-Wilk and Brown-Forsythe tests, respectively. After confirming homogeneous variances and normality, 2-group comparisons for means were performed using the 2-sided Student t test, and multi-group comparisons for means were performed using 2-way ANOVA with Holm-Sidak multiple comparison test. For data that did not pass either normality or equal variance test, 2-group comparisons were performed using Mann-Whitney Rank-sum test, and multi-group comparisons were performed using the Kruskal-Wallis 1-way ANOVA on ranks test with Dunn post hoc test. P\u0026lt;0.05 was considered statistically significant. Randomization and blinded analyses were performed whenever possible.\u003c/p\u003e"},{"header":"Declarations","content":"\n\u003cp\u003e\u003cstrong\u003eAcknowledgments\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe thank the University of Missouri Genomics Core for their support in generating data on the scRNA-seq.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eResearch reported in this publication was supported by the National Cancer Institute of the National Institutes of Health under Award Number R37CA241603 and the American Cancer Society under Award Number RSG-23-1152630.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eAuthor Contributions\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eX.K. conceived and supervised the study; C.W. performed experiments and analyzed data; Y.P., C.W., W.Z., and X.M. conceived the experiments; C.W., W.Z. performed scRNA-seq and bulk RNA-seq. Y.P. performed bioinformatics analysis. C.W. performed immunofluorescence staining and imaging data analysis. C.W., Y.P., W.Z. and X.M. verified the reproducibility of results; R.D.H. contributes to BM imaging data analysis; L.L., R.D., R.N. provided technical assistance and contribute to data analysis. Y.P., C.W. and X.K. wrote the original draft.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003cstrong\u003eDeclaration of interest\u003c/strong\u003e: The authors declare no competing interests.\u003c/p\u003e\n"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eBaryawno, N.\u003cem\u003e et al.\u003c/em\u003e A Cellular Taxonomy of the Bone Marrow Stroma in Homeostasis and Leukemia. \u003cem\u003eCell\u003c/em\u003e \u003cstrong\u003e177\u003c/strong\u003e, 1915-1932 e1916 (2019). https://doi.org/10.1016/j.cell.2019.04.040\u003c/li\u003e\n\u003cli\u003eAgarwal, P.\u003cem\u003e et al.\u003c/em\u003e Mesenchymal Niche-Specific Expression of Cxcl12 Controls Quiescence of Treatment-Resistant Leukemia Stem Cells. \u003cem\u003eCell Stem Cell\u003c/em\u003e \u003cstrong\u003e24\u003c/strong\u003e, 769-784 e766 (2019). https://doi.org/10.1016/j.stem.2019.02.018\u003c/li\u003e\n\u003cli\u003eGrockowiak, E.\u003cem\u003e et al.\u003c/em\u003e Different niches for stem cells carrying the same oncogenic driver affect pathogenesis and therapy response in myeloproliferative neoplasms. \u003cem\u003eNat Cancer\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 1193-1209 (2023). https://doi.org/10.1038/s43018-023-00607-x\u003c/li\u003e\n\u003cli\u003eMendez-Ferrer, S.\u003cem\u003e et al.\u003c/em\u003e Bone marrow niches in haematological malignancies. \u003cem\u003eNat Rev Cancer\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 285-298 (2020). https://doi.org/10.1038/s41568-020-0245-2\u003c/li\u003e\n\u003cli\u003eMorrison, S. J. \u0026amp; Scadden, D. T. The bone marrow niche for haematopoietic stem cells. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e505\u003c/strong\u003e, 327-334 (2014). https://doi.org/10.1038/nature12984\u003c/li\u003e\n\u003cli\u003eAiuti, A., Webb, I. J., Bleul, C., Springer, T. \u0026amp; Gutierrez-Ramos, J. C. The chemokine SDF-1 is a chemoattractant for human CD34+ hematopoietic progenitor cells and provides a new mechanism to explain the mobilization of CD34+ progenitors to peripheral blood. \u003cem\u003eJ Exp Med\u003c/em\u003e \u003cstrong\u003e185\u003c/strong\u003e, 111-120 (1997). https://doi.org/10.1084/jem.185.1.111\u003c/li\u003e\n\u003cli\u003eWright, D. E., Bowman, E. P., Wagers, A. J., Butcher, E. C. \u0026amp; Weissman, I. L. Hematopoietic stem cells are uniquely selective in their migratory response to chemokines. \u003cem\u003eThe Journal of experimental medicine\u003c/em\u003e \u003cstrong\u003e195\u003c/strong\u003e, 1145-1154 (2002). https://doi.org/10.1084/jem.20011284\u003c/li\u003e\n\u003cli\u003eHoffmann, T., Faust, J., Neubert, K. \u0026amp; Ansorge, S. Dipeptidyl peptidase IV (CD 26) and aminopeptidase N (CD 13) catalyzed hydrolysis of cytokines and peptides with N-terminal cytokine sequences. \u003cem\u003eFEBS letters\u003c/em\u003e \u003cstrong\u003e336\u003c/strong\u003e, 61-64 (1993). \u003c/li\u003e\n\u003cli\u003eChristopherson, K. W., 2nd, Hangoc, G., Mantel, C. R. \u0026amp; Broxmeyer, H. E. Modulation of hematopoietic stem cell homing and engraftment by CD26. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e305\u003c/strong\u003e, 1000-1003 (2004). https://doi.org/10.1126/science.1097071\u003c/li\u003e\n\u003cli\u003eWang, C.\u003cem\u003e et al.\u003c/em\u003e Dipeptidylpeptidase 4 promotes survival and stemness of acute myeloid leukemia stem cells. \u003cem\u003eCell Rep\u003c/em\u003e \u003cstrong\u003e42\u003c/strong\u003e, 112105 (2023). https://doi.org/10.1016/j.celrep.2023.112105\u003c/li\u003e\n\u003cli\u003eDavoodi, J., Kelly, J., Gendron, N. H. \u0026amp; MacKenzie, A. E. The Simpson-Golabi-Behmel syndrome causative glypican-3, binds to and inhibits the dipeptidyl peptidase activity of CD26. \u003cem\u003eProteomics\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 2300-2310 (2007). https://doi.org/10.1002/pmic.200600654\u003c/li\u003e\n\u003cli\u003eDe Cat, B.\u003cem\u003e et al.\u003c/em\u003e Processing by proprotein convertases is required for glypican-3 modulation of cell survival, Wnt signaling, and gastrulation movements. \u003cem\u003eJ Cell Biol\u003c/em\u003e \u003cstrong\u003e163\u003c/strong\u003e, 625-635 (2003). https://doi.org/10.1083/jcb.200302152\u003c/li\u003e\n\u003cli\u003eZhang, J.\u003cem\u003e et al.\u003c/em\u003e Identification of the haematopoietic stem cell niche and control of the niche size. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e425\u003c/strong\u003e, 836-841 (2003). https://doi.org/10.1038/nature02041\u003c/li\u003e\n\u003cli\u003eXie, Y.\u003cem\u003e et al.\u003c/em\u003e Detection of functional haematopoietic stem cell niche using real-time imaging. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e457\u003c/strong\u003e, 97-101 (2009). https://doi.org/10.1038/nature07639\u003c/li\u003e\n\u003cli\u003eZhao, M.\u003cem\u003e et al.\u003c/em\u003e N-Cadherin-Expressing Bone and Marrow Stromal Progenitor Cells Maintain Reserve Hematopoietic Stem Cells. \u003cem\u003eCell reports\u003c/em\u003e \u003cstrong\u003e26\u003c/strong\u003e, 652-669.e656 (2019). https://doi.org/10.1016/j.celrep.2018.12.093\u003c/li\u003e\n\u003cli\u003eWeidenaar, A. C.\u003cem\u003e et al.\u003c/em\u003e High acute myeloid leukemia derived VEGFA levels are associated with a specific vascular morphology in the leukemic bone marrow. \u003cem\u003eCell Oncol (Dordr)\u003c/em\u003e \u003cstrong\u003e34\u003c/strong\u003e, 289-296 (2011). https://doi.org/10.1007/s13402-011-0017-9\u003c/li\u003e\n\u003cli\u003ePadr\u0026oacute;, T.\u003cem\u003e et al.\u003c/em\u003e Increased angiogenesis in the bone marrow of patients with acute myeloid leukemia. \u003cem\u003eBlood\u003c/em\u003e \u003cstrong\u003e95\u003c/strong\u003e, 2637-2644 (2000). \u003c/li\u003e\n\u003cli\u003eKupsa, T., Vanek, J., Pavel, Z., Jebavy, L. \u0026amp; Horacek, J. M. Serum levels of soluble adhesion molecules in newly diagnosed acute myeloid leukemia and in complete remission suggest endothelial cell activation by myeloblasts. \u003cem\u003eBiomed Pap Med Fac Univ Palacky Olomouc Czech Repub\u003c/em\u003e \u003cstrong\u003e161\u003c/strong\u003e, 92-99 (2017). https://doi.org/10.5507/bp.2016.054\u003c/li\u003e\n\u003cli\u003eTinoco, A. D., Tagore, D. M. \u0026amp; Saghatelian, A. Expanding the dipeptidyl peptidase 4-regulated peptidome via an optimized peptidomics platform. \u003cem\u003eJ Am Chem Soc\u003c/em\u003e \u003cstrong\u003e132\u003c/strong\u003e, 3819-3830 (2010). https://doi.org/10.1021/ja909524e\u003c/li\u003e\n\u003cli\u003eCampbell, T. B. \u0026amp; Broxmeyer, H. E. CD26 inhibition and hematopoiesis: a novel approach to enhance transplantation. \u003cem\u003eFront Biosci\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 1795-1805 (2008). \u003c/li\u003e\n\u003cli\u003eKei Ohnuma, C. M. DPP4. \u003cem\u003eAtlas of Genetics and Cytogenetics \u003cem\u003ein Oncology and Haematology\u003c/em\u003e (2012). \u003c/em\u003e\u003c/li\u003e\n\u003cli\u003eBroxmeyer, H. E.\u003cem\u003e et al.\u003c/em\u003e Dipeptidylpeptidase 4 negatively regulates colony-stimulating factor activity and stress hematopoiesis. \u003cem\u003eNat Med\u003c/em\u003e \u003cstrong\u003e18\u003c/strong\u003e, 1786-1796 (2012). https://doi.org/10.1038/nm.2991\u003c/li\u003e\n\u003cli\u003ePeled, A.\u003cem\u003e et al.\u003c/em\u003e Dependence of human stem cell engraftment and repopulation of NOD/SCID mice on CXCR4. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e283\u003c/strong\u003e, 845-848 (1999). \u003c/li\u003e\n\u003cli\u003ePetit, I.\u003cem\u003e et al.\u003c/em\u003e G-CSF induces stem cell mobilization by decreasing bone marrow SDF-1 and up-regulating CXCR4. \u003cem\u003eNature immunology\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 687-694 (2002). https://doi.org/10.1038/ni813\u003c/li\u003e\n\u003cli\u003eSingh, P.\u003cem\u003e et al.\u003c/em\u003e Neuropeptide Y regulates a vascular gateway for hematopoietic stem and progenitor cells. \u003cem\u003eJ Clin Invest\u003c/em\u003e \u003cstrong\u003e127\u003c/strong\u003e, 4527-4540 (2017). https://doi.org/10.1172/jci94687\u003c/li\u003e\n\u003cli\u003eWei, Q. \u0026amp; Frenette, P. S. Niches for Hematopoietic Stem Cells and Their Progeny. \u003cem\u003eImmunity\u003c/em\u003e \u003cstrong\u003e48\u003c/strong\u003e, 632-648 (2018). https://doi.org/10.1016/j.immuni.2018.03.024\u003c/li\u003e\n\u003cli\u003eCalvi, L. M.\u003cem\u003e et al.\u003c/em\u003e Osteoblastic cells regulate the haematopoietic stem cell niche. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e425\u003c/strong\u003e, 841-846 (2003). https://doi.org/10.1038/nature02040\u003c/li\u003e\n\u003cli\u003eZhou, B. O., Yue, R., Murphy, M. M., Peyer, J. G. \u0026amp; Morrison, S. J. Leptin-receptor-expressing mesenchymal stromal cells represent the main source of bone formed by adult bone marrow. \u003cem\u003eCell Stem Cell\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 154-168 (2014). https://doi.org/10.1016/j.stem.2014.06.008\u003c/li\u003e\n\u003cli\u003eDing, L. \u0026amp; Morrison, S. J. Haematopoietic stem cells and early lymphoid progenitors occupy distinct bone marrow niches. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e495\u003c/strong\u003e, 231-235 (2013). https://doi.org/10.1038/nature11885\u003c/li\u003e\n\u003cli\u003eGreenbaum, A.\u003cem\u003e et al.\u003c/em\u003e CXCL12 in early mesenchymal progenitors is required for haematopoietic stem-cell maintenance. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e495\u003c/strong\u003e, 227-230 (2013). https://doi.org/10.1038/nature11926\u003c/li\u003e\n\u003cli\u003eMendez-Ferrer, S.\u003cem\u003e et al.\u003c/em\u003e Mesenchymal and haematopoietic stem cells form a unique bone marrow niche. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e466\u003c/strong\u003e, 829-834 (2010). https://doi.org/10.1038/nature09262\u003c/li\u003e\n\u003cli\u003eLu, X.\u003cem\u003e et al.\u003c/em\u003e Identification of the homeobox protein Prx1 (MHox, Prrx-1) as a regulator of osterix expression and mediator of tumor necrosis factor alpha action in osteoblast differentiation. \u003cem\u003eJ Bone Miner Res\u003c/em\u003e \u003cstrong\u003e26\u003c/strong\u003e, 209-219 (2011). https://doi.org/10.1002/jbmr.203\u003c/li\u003e\n\u003cli\u003eEsposito, A., Wang, L., Li, T., Miranda, M. \u0026amp; Spagnoli, A. Role of Prx1-expressing skeletal cells and Prx1-expression in fracture repair. \u003cem\u003eBone\u003c/em\u003e \u003cstrong\u003e139\u003c/strong\u003e, 115521 (2020). https://doi.org/10.1016/j.bone.2020.115521\u003c/li\u003e\n\u003cli\u003eMizoguchi, T.\u003cem\u003e et al.\u003c/em\u003e Osterix marks distinct waves of primitive and definitive stromal progenitors during bone marrow development. \u003cem\u003eDev Cell\u003c/em\u003e \u003cstrong\u003e29\u003c/strong\u003e, 340-349 (2014). https://doi.org/10.1016/j.devcel.2014.03.013\u003c/li\u003e\n\u003cli\u003eForte, D.\u003cem\u003e et al.\u003c/em\u003e Bone Marrow Mesenchymal Stem Cells Support Acute Myeloid Leukemia Bioenergetics and Enhance Antioxidant Defense and Escape from Chemotherapy. \u003cem\u003eCell Metab\u003c/em\u003e \u003cstrong\u003e32\u003c/strong\u003e, 829-843.e829 (2020). https://doi.org/10.1016/j.cmet.2020.09.001\u003c/li\u003e\n\u003cli\u003ePinho, S.\u003cem\u003e et al.\u003c/em\u003e VCAM1 confers innate immune tolerance on haematopoietic and leukaemic stem cells. \u003cem\u003eNat Cell Biol\u003c/em\u003e \u003cstrong\u003e24\u003c/strong\u003e, 290-298 (2022). https://doi.org/10.1038/s41556-022-00849-4\u003c/li\u003e\n\u003cli\u003eMa, D.\u003cem\u003e et al.\u003c/em\u003e Intracellular angiopoietin-1 promotes TKI-resistance via activation of JAK/STAT5 pathway in chronic myeloid leukemia. \u003cem\u003eOncogene\u003c/em\u003e \u003cstrong\u003e42\u003c/strong\u003e, 124-137 (2023). https://doi.org/10.1038/s41388-022-02536-y\u003c/li\u003e\n\u003cli\u003eCarlesso, N.\u003cem\u003e et al.\u003c/em\u003e Human recombinant stem cell factor stimulates in vitro proliferation of acute myeloid leukemia cells and expands the clonogenic cell pool. \u003cem\u003eLeukemia\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 642-648 (1992). \u003c/li\u003e\n\u003cli\u003eBendall, L. J.\u003cem\u003e et al.\u003c/em\u003e Stem cell factor enhances the adhesion of AML cells to fibronectin and augments fibronectin-mediated anti-apoptotic and proliferative signals. \u003cem\u003eLeukemia\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 1375-1382 (1998). https://doi.org/10.1038/sj.leu.2401136\u003c/li\u003e\n\u003cli\u003eKunz, L. \u0026amp; Schroeder, T. A 3D Tissue-wide Digital Imaging Pipeline for Quantitation of Secreted Molecules Shows Absence of CXCL12 Gradients in Bone Marrow. \u003cem\u003eCell Stem Cell\u003c/em\u003e \u003cstrong\u003e25\u003c/strong\u003e, 846-854 e844 (2019). https://doi.org/10.1016/j.stem.2019.10.003\u003c/li\u003e\n\u003cli\u003eJanssens, R., Struyf, S. \u0026amp; Proost, P. The unique structural and functional features of CXCL12. \u003cem\u003eCell Mol Immunol\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 299-311 (2018). https://doi.org/10.1038/cmi.2017.107\u003c/li\u003e\n\u003cli\u003eCui, X. Y.\u003cem\u003e et al.\u003c/em\u003e Tissue factor pathway inhibitor upregulates CXCR7 expression and enhances CXCL12-mediated migration in chronic lymphocytic leukemia. \u003cem\u003eSci Rep\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 5127 (2021). https://doi.org/10.1038/s41598-021-84695-8\u003c/li\u003e\n\u003cli\u003eRamakrishnan, R.\u003cem\u003e et al.\u003c/em\u003e CXCR4 Signaling Has a CXCL12-Independent Essential Role in Murine MLL-AF9-Driven Acute Myeloid Leukemia. \u003cem\u003eCell Rep\u003c/em\u003e \u003cstrong\u003e31\u003c/strong\u003e, 107684 (2020). https://doi.org/10.1016/j.celrep.2020.107684\u003c/li\u003e\n\u003cli\u003eAnderson, N. R.\u003cem\u003e et al.\u003c/em\u003e CXCL12 Knock-out Enhances Leukemia Stem Cell Response to Combination Chemotherapy Plus Tyrosine Kinase Inhibition in Flt3-ITD Acute Myeloid Leukemia. \u003cem\u003eBlood\u003c/em\u003e \u003cstrong\u003e136\u003c/strong\u003e, 7-8 (2020). https://doi.org/https://doi.org/10.1182/blood-2020-142579\u003c/li\u003e\n\u003cli\u003eZeng, Z.\u003cem\u003e et al.\u003c/em\u003e Targeting the leukemia microenvironment by CXCR4 inhibition overcomes resistance to kinase inhibitors and chemotherapy in AML. \u003cem\u003eBlood\u003c/em\u003e \u003cstrong\u003e113\u003c/strong\u003e, 6215-6224 (2009). https://doi.org/10.1182/blood-2008-05-158311\u003c/li\u003e\n\u003cli\u003eAgarwal, P.\u003cem\u003e et al.\u003c/em\u003e Mesenchymal Niche-Specific Expression of Cxcl12 Controls Quiescence of Treatment-Resistant Leukemia Stem Cells. \u003cem\u003eCell Stem Cell\u003c/em\u003e \u003cstrong\u003e24\u003c/strong\u003e, 769-784.e766 (2019). https://doi.org/10.1016/j.stem.2019.02.018\u003c/li\u003e\n\u003cli\u003eSchelker, R. C.\u003cem\u003e et al.\u003c/em\u003e TGF-beta1 and CXCL12 modulate proliferation and chemotherapy sensitivity of acute myeloid leukemia cells co-cultured with multipotent mesenchymal stromal cells. \u003cem\u003eHematology\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, 337-345 (2018). https://doi.org/10.1080/10245332.2017.1402455\u003c/li\u003e\n\u003cli\u003eTavor, S.\u003cem\u003e et al.\u003c/em\u003e CXCR4 regulates migration and development of human acute myelogenous leukemia stem cells in transplanted NOD/SCID mice. \u003cem\u003eCancer Res\u003c/em\u003e \u003cstrong\u003e64\u003c/strong\u003e, 2817-2824 (2004). https://doi.org/10.1158/0008-5472.can-03-3693\u003c/li\u003e\n\u003cli\u003eAnderson, N. R.\u003cem\u003e et al.\u003c/em\u003e Microenvironmental CXCL12 deletion enhances Flt3-ITD acute myeloid leukemia stem cell response to therapy by reducing p38 MAPK signaling. \u003cem\u003eLeukemia\u003c/em\u003e \u003cstrong\u003e37\u003c/strong\u003e, 560-570 (2023). https://doi.org/10.1038/s41375-022-01798-5\u003c/li\u003e\n\u003cli\u003eNamburi, S., Broxmeyer, H. E., Hong, C. S., Whiteside, T. L. \u0026amp; Boyiadzis, M. DPP4(+) exosomes in AML patients\u0026apos; plasma suppress proliferation of hematopoietic progenitor cells. \u003cem\u003eLeukemia\u003c/em\u003e \u003cstrong\u003e35\u003c/strong\u003e, 1925-1932 (2021). https://doi.org/10.1038/s41375-020-01047-7\u003c/li\u003e\n\u003cli\u003eCline, A., Jajosky, R., Shikle, J. \u0026amp; Bollag, R. Comparing leukapheresis protocols for an AML patient with symptomatic leukostasis. \u003cem\u003eJ Clin Apher\u003c/em\u003e \u003cstrong\u003e33\u003c/strong\u003e, 396-400 (2018). https://doi.org/10.1002/jca.21588\u003c/li\u003e\n\u003cli\u003eDel Prete, C., Kim, T., Lansigan, F., Shatzel, J. \u0026amp; Friedman, H. The Epidemiology and Clinical Associations of Stroke in Patients With Acute Myeloid Leukemia: A Review of 10,972 Admissions From the 2012 National Inpatient Sample. \u003cem\u003eClin Lymphoma Myeloma Leuk\u003c/em\u003e \u003cstrong\u003e18\u003c/strong\u003e, 74-77 e71 (2018). https://doi.org/10.1016/j.clml.2017.09.008\u003c/li\u003e\n\u003cli\u003eHavale, S. H. \u0026amp; Pal, M. Medicinal chemistry approaches to the inhibition of dipeptidyl peptidase-4 for the treatment of type 2 diabetes. \u003cem\u003eBioorg Med Chem\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, 1783-1802 (2009). https://doi.org/10.1016/j.bmc.2009.01.061\u003c/li\u003e\n\u003cli\u003ede Boer, J.\u003cem\u003e et al.\u003c/em\u003e Transgenic mice with hematopoietic and lymphoid specific expression of Cre. \u003cem\u003eEur J Immunol\u003c/em\u003e \u003cstrong\u003e33\u003c/strong\u003e, 314-325 (2003). https://doi.org/10.1002/immu.200310005\u003c/li\u003e\n\u003cli\u003eGothert, J. R.\u003cem\u003e et al.\u003c/em\u003e In vivo fate-tracing studies using the Scl stem cell enhancer: embryonic hematopoietic stem cells significantly contribute to adult hematopoiesis. \u003cem\u003eBlood\u003c/em\u003e \u003cstrong\u003e105\u003c/strong\u003e, 2724-2732 (2005). https://doi.org/10.1182/blood-2004-08-3037\u003c/li\u003e\n\u003cli\u003eKang, X. \u0026amp; Wang, C. CAMKs support development of acute myeloid leukemia. \u003cem\u003eThe Journal of experimental medicine\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 30 (2018). https://doi.org/10.1084/jem.20171477 10.1186/s13045-018-0574-8\u003c/li\u003e\n\u003cli\u003eZheng, J.\u003cem\u003e et al.\u003c/em\u003e Inhibitory receptors bind ANGPTLs and support blood stem cells and leukaemia development. \u003cem\u003eNature\u003c/em\u003e (2012). \u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-6515437/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6515437/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eLeukemic stem cells (LSCs) reside in protective bone marrow (BM) niches, evading chemotherapy and driving relapse. Disrupting LSC-niche represents a promising therapeutic strategy, yet effective approach remains elusive. Our study characterized longitudinal BM niches in support of LSC survival and proliferation, i.e. metaphysis vs. central marrow, compared to well-accepted endosteum vs. sinusoid niches. Quiescent LSCs mostly localize to metaphysis and show reduced stemness and aggressiveness when mobilized to the central marrow, the composition of both endosteum and sinusoid in the central marrow. Building on this, we developed an approach to restrict LSCs within the BM and induce LSC apoptosis by targeting DPP4 in AML cells. Genetic deletion of \u003cem\u003eDpp4\u003c/em\u003e in AML cells alters CXCL12 gradient across three scales: \u003cb\u003e1) System-wide\u003c/b\u003e: A reversed CXCL12 gradient between the BM and peripheral blood confines AML cells within the BM, limiting their circulation. \u003cb\u003e2) BM level\u003c/b\u003e: Perturbation of the CXCL12 gradient between the metaphysis and central marrow mobilizes LSCs out of their protective metaphysis niche, leading to exhaustion in the sinusoidal region, despite the higher CXCL12 level in the sinusoidal region. \u003cb\u003e3) Microscale within the metaphysis\u003c/b\u003e: Loss of the CXCL12 gradient between N-cadherin\u003csup\u003e+\u003c/sup\u003e mesenchymal stromal cells and the surrounding matrix impairs LSC recruitment to N-cad\u003csup\u003e+\u003c/sup\u003e cells, further driving their exhaustion. These alterations stem from the CXCL12-DPP4-GPC3 axis. DPP4, highly expressed by AML cells, deactivates CXCL12, while GPC3, enriched in N-cad\u003csup\u003e+\u003c/sup\u003e cells, inhibits DPP4's enzymatic activity. This axis establishes a favorable CXCL12 gradient that attracts LSCs to N-cad\u003csup\u003e+\u003c/sup\u003e cell-rich niches in the metaphysis and facilitates their dissemination via circulation. These findings highlight the therapeutic potential of targeting CXCL12-DPP4-GPC3 axis to disrupt LSC niches and enhance AML treatment.\u003c/p\u003e","manuscriptTitle":"Longitudinal Localization of Leukemia Stem Cells Between Metaphysis and Central Marrow Governs Leukemic Stem Cell Behavior","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-05 06:43:03","doi":"10.21203/rs.3.rs-6515437/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-cell-biology","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"ncb","sideBox":"Learn more about [Nature Cell Biology](http://www.nature.com/ncb/)","snPcode":"","submissionUrl":"","title":"Nature Cell Biology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Research","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"83719968-9a8a-42b3-a975-50b22ecc9e67","owner":[],"postedDate":"May 5th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":47649700,"name":"Health sciences/Diseases/Cancer/Cancer microenvironment"},{"id":47649701,"name":"Health sciences/Diseases/Cancer/Cancer stem cells"},{"id":47649702,"name":"Biological sciences/Cancer/Cancer microenvironment"},{"id":47649703,"name":"Biological sciences/Cancer/Cancer stem cells"}],"tags":[],"updatedAt":"2026-04-25T07:10:37+00:00","versionOfRecord":{"articleIdentity":"rs-6515437","link":"https://doi.org/10.1038/s41556-026-01939-3","journal":{"identity":"nature-cell-biology","isVorOnly":false,"title":"Nature Cell Biology"},"publishedOn":"2026-04-24 04:00:00","publishedOnDateReadable":"April 24th, 2026"},"versionCreatedAt":"2025-05-05 06:43:03","video":"","vorDoi":"10.1038/s41556-026-01939-3","vorDoiUrl":"https://doi.org/10.1038/s41556-026-01939-3","workflowStages":[]},"version":"v1","identity":"rs-6515437","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6515437","identity":"rs-6515437","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}