Abstract

   Growing evidence underscores the pivotal impact of crosstalk between
   leukemic stem cells (LSCs) and mesenchymal stromal cells (MSCs) within
   their niche on leukemia initiation, progression, and therapy response.
   Although MMP14 plays an important role in inflammation and cancer, the
   regulation and role of MSC-derived MMP14 in acute myeloid leukemia
   (AML) are largely unknown. Here, we found that AML patient-derived MSCs
   (AML-MSCs) were more supportive of AML cell growth compared to healthy
   donor-derived MSCs (HD-MSCs). Moreover, AML-MSCs and HD-MSCs showed
   significant differences in gene expression and protein expression
   profiles. Knockdown of MMP14 in MSCs inhibited the CFU-F ability of MSC
   cells and increased the proportion of cells in the G0 phase, thereby
   inhibiting proliferation. Co-culture with MSCs inhibited the
   proliferation and cell cycle progression of leukemia cells, while
   increasing the apoptosis rate, thus impairing the leukemogenic
   potential of AML cells both in vitro and in vivo. Mechanistic studies
   revealed that MMP14-mediated alterations in the AML stromal
   microenvironment are driven by PGE2 secretion and activation of the
   JAK-STAT pathway, promoting leukemia progression. Notably, inhibition
   of MMP14 can attenuate the chemotherapy resistance of AML cells induced
   by MSCs to cytarabine (Ara-C). Together, our study, for the first time,
   demonstrates the critical role of MSC-derived MMP14 in promoting AML
   progression and chemoresistance. Targeting MMP14 signaling pathways may
   offer novel therapeutic options for AML.

Supplementary Information

   The online version contains supplementary material available at
   10.1186/s40164-025-00635-6.

   Keywords: Acute myeloid leukemia, Mesenchymal stromal cells, MMP14,
   JAK/STAT pathway, PGE2

Background

   The tumor microenvironment plays a crucial role in cancer progression
   and therapy resistance [[47]1–[48]3]. The bone marrow (BM) comprises a
   dynamic network of growth factors, cytokines, and stromal cells,
   fostering a conducive environment for leukemia development and
   progression [[49]4]. Bone marrow-derived mesenchymal stem cells (BMSCs)
   are crucial components of the bone marrow microenvironment, capable of
   differentiating into osteoblasts, chondrocytes, and adipocytes under
   specific conditions [[50]5, [51]6]. MSCs regulate both normal and
   malignant hematopoiesis, yet the underlying molecular mechanisms remain
   poorly understood. MSC function is markedly altered in AML patients. In
   addition to changes in growth kinetics and differentiation potential,
   MSCs from AML patients alter gene expression associated with the
   induction of hematopoietic progenitor cell quiescence [[52]7, [53]8].
   MSC-derived signaling within the bone marrow may elucidate the origin
   of bone marrow-derived changes observed in AML, which confer a leukemic
   advantage over normal hematopoiesis [[54]9]. MSCs can regulate the
   development of leukemia through direct cell-to-cell contact,
   cytokine-receptor interactions, and exosome-mediated communication
   [[55]10–[56]13]. These interactions hinder physiological hematopoiesis
   while promoting LSC survival and proliferation. Therefore, further
   studies on MSCs within the bone marrow microenvironment are crucial for
   advancing therapeutic options for leukemia.

   Matrix metalloproteinases (MMPs) are critical enzymes that degrade the
   extracellular matrix and disrupt tissue barriers, mediating changes in
   the tumor microenvironment during development and playing a crucial
   role in tumor invasion and metastasis [[57]14–[58]16]. As a
   transmembrane protein, MMP14 activates pro-MMP2 to induce tumor cell
   invasion. MMP14 is highly expressed in various tumors, promoting
   inflammation, angiogenesis, cancer cell invasion and metastasis, and
   drug resistance [[59]17–[60]21]. Additionally, aberrant expression of
   MMP family molecules has been studied in leukemia [[61]22–[62]24].
   Despite the well-established significance of MMP14 in cancer, the
   regulatory mechanisms and functions of MSC-derived MMP14 in AML remain
   largely unexplored.

   Here, we demonstrate the dynamic functional and molecular alterations
   of BMMSCs associated with AML burden. We also found that MMP14 is
   highly expressed in AML-MSCs, playing a crucial role in leukemia cell
   proliferation, apoptosis, and chemoresistance. Most importantly,
   depletion of MMP14 in MSCs leads to the inhibition of AML development.
   Mechanistically, MSC-derived MMP14 is a major contributor to the
   activation of the JAK-STAT pathway through PGE2 secretion, contributing
   to increased AML proliferation and chemoresistance. Aberrant expression
   of MMP14 in MSCs may aid in identifying novel therapeutic targets for
   AML treatment.

Methods

Drugs and reagents

   The MMP14 inhibitor NSC 405020 was purchased from MCE (HY-15827,
   China). Cytarabine was purchased from Actavis (Italy). PGE2 was
   obtained from MCE (HY-101952, China).

Cell culture

   MSCs were derived from fresh BM aspirates of patients with newly
   diagnosed AML and healthy controls using a density gradient
   centrifugation with Ficoll-Lymphoprep™ (TBD, China). BMMSCs were
   maintained in Human MSC Growth Medium (Cyagen Biosciences Inc.) with
   regular medium changes every three days. After three days, non-adherent
   cells were removed through complete medium replacement, while the
   adherent cells were persistently cultured. All experiments utilized
   MSCs from passage 3. Surface immunophenotyping via flow cytometry
   characterized the MSCs. Morphological assessment confirmed the
   differentiation of MSCs into adipocytes and osteocyte lineages
   (Fig.[63]S1). MSCs from C57BL/6 mice were acquired from Cyagen
   Biosciences Inc. These cells were propagated in Mouse MSC Growth Medium
   (Cyagen Biosciences Inc.) and utilized within five passages. AML cell
   lines Kasumi-1 and Molm-13, sourced from the Cell Resource Center of
   the Shanghai Institutes for Biological Sciences (Shanghai, China), were
   cultured in RPMI 1640 medium (HyClone, USA). All cells were incubated
   at 37 °C in a humidified atmosphere of 5% CO2.

RNA library construction and sequencing

   Total mRNAs were isolated from AML-MSCs (passage 3, n = 5) and HD-MSCs
   (passage 3, n = 5) using TRIzol reagent (Vazyme, China) and measured
   using NanoDrop (Thermo). RNA library construction and sequencing
   analysis were performed by Novogene (Tianjin, China). Genes with a p
   value < 0.05 were identified as differentially expressed genes (DEGs).

Proteomic analysis of secreted proteins in BMSC culture supernatants

   The MSC medium was replaced with FBS-free DMEM/F12 medium 24 h before
   supernatant extraction. Following 24 h of culture, the supernatant was
   harvested, spun down at 1500 rpm for 20 min to eliminate nonviable and
   dead cells, and preserved at -80 °C for further proteomic analysis.
   Samples derived from MSC supernatants underwent analysis on a Q
   Exactive mass spectrometer (Thermo Fisher Scientific, Bremen, Germany)
   by KangChen Bio-tech (Shanghai, China). MaxQuant software (version
   1.5.6.0) was utilized to process raw data. Proteins exhibiting a
   differential expression p value < 0.05 were deemed significantly
   different.

Immunofluorescence microscopy of BM sections

   Bone marrow biopsy specimens were quickly placed in formalin fixative.
   Following a minimum of 4 h of fixation, the tissue underwent
   decalcification with EDTA for 24 h, proceeded by standard dehydration
   and paraffin embedding, producing sections of 5 μm thickness. Slides
   were treated with MMP14 (380861, Zenbio, China) and CD271 (55014-1-AP,
   Proteintech, China) antibodies for two days, then with
   4’,6-diamidino-2-phenylindole (DAPI) for 12 h.

Real-time quantitative RT-PCR (qRT-PCR)

   Total RNA was isolated using Trizol (Vazyme, China) and converted into
   cDNA with Hifair^® II 1st Strand cDNA Synthesis SuperMix (Yeasen
   Biotechnology, 11120ES60). Quantitative RT-PCR analysis utilized the
   SYBR Green method via the CWBIO Fast SYBR Mixture real-time PCR system
   (CWBIO, China). qRT-PCR primer sequences can be found in Supplemental
   Table 1. Relevant gene expression levels were measured using the 2-ΔΔCt
   method.

Lentivirus infection

   MMP14 shRNA (shMMP14 group) and scrambled shRNA (control group) were
   obtained from Hanheng Biotechnology (Hanheng Biotechnology Co. Ltd.,
   Shanghai, China). Green fluorescent protein (GFP) was used to assess
   transduction efficiency, and puromycin was used to select stably
   transduced cells.

Cell proliferation, apoptosis and cell cycle analysis

   Cell proliferation tests were carried out through flow cytometry to
   measure BrdU incorporation, executed as per the manufacturer’s
   instructions (BD Pharmingen BrdU Flow Kit, BrdU-APC). Apoptosis rates
   were determined by flow cytometry using the Apoptosis Detection Kit
   (Lianke Technology Co., Ltd.). Cell cycle analysis was performed using
   the Cell Cycle Analysis Kit (Lianke Technology Co., Ltd.).

Colony forming Unit-Fibroblast (CFU-F) assay

   The colony forming unit-fibroblast (CFU-F) assay was performed on two
   groups of MSCs by plating 2,000 cells from each group on 6 cm² dishes.
   The cells were cultured in MSC medium for 14 days with medium changes
   every three days. Following 14 days, the cells were preserved with 4%
   paraformaldehyde and dyed with crystal violet. The colony forming
   ability of MSCs was assessed by counting colonies containing more than
   50 cells.

Animal experimentation

   MLL-AF9 retroviral transduction was conducted on CD117 + enriched cells
   from 8–12-week-old C57BL/6 mice, isolated via magnetic bead sorting
   (Miltenyi, 130-097-146, Germany), and subsequently infected with
   MSCV-MLL-AF9-IRES-YFP retroviruses twice amidst 8 µg/mL Polybrene
   (Hanheng Biotechnology Co. Ltd., HB-PB-500), 10 ng/mL IL-3 (PeproTech,
   213 − 13), 10 ng/mL IL-6 (PeproTech, 216 − 16), and 20 ng/mL SCF
   (PeproTech, 250-03) in IMDM medium. Sub-lethally irradiated (7 Gy)
   recipient mice then received the infected cells via intravenous
   transplantation. Sorted by fluorescence-activated cell sorting (FACS),
   the MLL-AF9-expressing cells from primary recipient mice with AML were
   preserved for further transplantation to induce AML. A total of 100,000
   MLL-AF9 cells in 200 µL PBS were intravenously transplanted into
   recipient mice. The MMP14 knockdown and control MSCs (100,000 cells/10
   µL PBS) were then injected into the distal femora of both the right and
   left hind limbs. The mice were subsequently divided into two groups:
   the shMMP14 knockdown group and the control group. NSG mice (female, 5
   weeks old) were procured from Jiangsu Jicui Biotechnology Co., Ltd.
   Primary human leukemia cells were isolated from high-leukemia AML
   patients using leukapheresis and PBMC separation. Each mouse was
   intravenously injected with 1 × 10^7 cells to induce leukemia. The
   successful establishment of the patient-derived xenograft (PDX) model
   was confirmed by observing leukemia cells in BM smears and through flow
   cytometry, with leukemia cell proportions in the BM ≥ 20%. Information
   on the antibodies used in the experiments is provided in Supplemental
   Table 2.

Western blot

   Proteins were isolated from cells using RIPA lysis buffer, which
   included PMSF and phosphatase inhibitors A and B. Protein aliquots were
   subjected to 10% SDS-polyacrylamide gel electrophoresis and then
   transferred onto PVDF membranes (Millipore Corp, Billerica, MA, USA).
   Overnight at 4 °C, the membranes were treated with primary antibodies
   (dilution 1:1000): MMP14 (380861, Zenbio, China), STAT5 (13179-1-AP,
   Proteintech, China), P-STAT5 (80115-1-RR, Proteintech, China), STAT3
   (10253-2-AP, Proteintech, China), P-STAT3 (28945-1-AP, Proteintech,
   China), COX-2 (501253, Zenbio, China), and GAPDH (60004, Proteintech,
   China). Subsequently, for 1 h at room temperature, the membranes were
   treated with either goat anti-rabbit IgG (H + L) antibody (SA00001-2,
   Proteintech, China) or goat anti-mouse IgG (H + L) antibody (SA00004-1,
   Proteintech, China). An imaging system (Tanon, China) was used for
   detection.

ELISA

   Culture supernatants were collected from each well, and PGE2 protein
   levels were measured using the PGE2 ELISA kit (Jianglaibio, Shanghai,
   China). Duplicate wells were used for standard curves and experimental
   samples for each measurement. Concentrations were determined according
   to the manufacturer’s instructions.

Histopathology

   Femurs and tibiae from animal studies were utilized to prepare BM
   smears with Wright-Giemsa staining. Tissues obtained from mice
   underwent sectioning, dewaxing, and rehydration. Some samples received
   hematoxylin-eosin (HE) staining, while others underwent
   immunohistochemical analysis. Sections were treated with antibodies
   against P-STAT3 (CY6501, Abways) and P-STAT5 (CY5569, Abways). All
   slides were examined and photographed through a Nikon microscope
   outfitted with a high-resolution digital camera.

Statistical analysis

   Statistical analyses were performed using GraphPad Prism (8.0), gene
   set enrichment analysis (GSEA), and R scripts. Differences between
   groups were analyzed using two-tailed t-tests. All data are expressed
   as means ± standard deviations. Kaplan-Meier survival curves for mice
   were generated using the log-rank (Mantel-Cox) test in Prism 8.0. A p
   value of < 0.05 was considered statistically significant. ***
   p < 0.001, ** p < 0.01, and * p < 0.05.

Results

Multi-omic analysis identifies MMP14 as a key molecule in AMLMSC supporting
leukemia cell growth

   To investigate the capacity of AML-MSC and HD-MSC to support leukemia
   cell proliferation, primary leukemia cells were co-cultured with
   AML-MSC and HD-MSC for two days in vitro. Flow cytometry revealed a
   lower apoptosis rate in leukemia cells co-cultured with AML-MSC, while
   a Brdu assay indicated a higher proliferation rate in the AML-MSC group
   (Fig. [64]1A-D). Similarly, leukemia cell lines Kasumi-1 and Molm-13
   were cultured with AML-MSC and HD-MSC for two days in vitro, which
   showed that AML-MSC, compared to HD-MSC, promoted leukemia cells
   proliferation and reduced apoptosis rates (Fig. S2). To further
   investigate the mechanisms by which AMLMSCs promote AML cell growth
   compared to HDMSCs, we collected BMMSCs from five patients with primary
   AML diagnosis and five healthy controls for transcriptome sequencing.
   Differentially expressed genes (DEGs) were identified between AML-MSCs
   and HD-MSCs with criteria of P value < 0.05. In total, 668 DEGs were
   detected, comprising 349 upregulated and 319 downregulated genes
   (Fig. [65]1E; Supplemental Table S3, 4). To further elucidate the
   pathways and processes influenced by the identified DEGs, GO and KEGG
   analyses were conducted. KEGG pathway enrichment analysis indicated
   significant enrichment in the PI3K-Akt signaling pathway and focal
   adhesion (Fig. [66]1F). Following the characterization of gene
   properties in AML-MSCs and HD-MSCs, we aimed to identify proteins
   characteristic of the AML-MSC supernatant. To identify secreted
   proteins, we conducted label-free proteomics on AML-MSCs and HD-MSCs. A
   total of 97 differentially expressed proteins were identified,
   comprising 81 upregulated and 16 downregulated proteins (P
   value < 0.05; Fig. [67]1G). Details of the altered protein expressions
   are provided in Supplemental Table S5. We integrated transcriptomic and
   proteomic analyses, revealing overlapping genes (Fig. [68]1H). To
   further screen candidate molecules, OHSU-AML survival analysis
   indicated that MMP14 was associated with poor prognosis in AML
   (Fig. [69]1I). We subsequently verified high MMP14 expression in
   AML-MSCs using RT-qPCR and immunofluorescence in bone marrow MSC
   specimens from AML patients compared to those from healthy donors
   (Fig. [70]1J, K). These data suggest that MMP14 is a key molecule in
   AML-MSCs supporting leukemia cell growth.

Fig. 1.

   [71]Fig. 1
   [72]Open in a new tab

   Combined transcriptomic and proteomic analyses identified MMP14 as a
   key molecule in AMLMSC supporting leukemia cell growth. A, B. Analysis
   of apoptosis in primary leukemia cells after 48 h of co-culture with
   five cases each of AML-MSC and HD-MSC, using the Annexin V APC
   Apoptosis Detection Kit (n = 5). Alone, leukemia cells cultured alone;
   Co, leukemia cells cocultured with MSCs. C, D. Flow BrdU assay for
   assessing the proliferation of primary leukemia cells after 48 h of
   co-culture with 5 AMLMSC and 5 HDMSC cases(n = 5). E. Volcano plots
   displaying the expression of differentially expressed genes between
   five AML-MSC and five HD-MSC samples, with purple representing
   upregulation and blue representing downregulation(n = 5). F. Enrichment
   analysis of differentially expressed genes. G. Proteomic analysis of
   supernatants from 4 AML-MSC and 4 HD-MSC samples, including a heat map
   of differentially expressed proteins(n = 4). H. Intersection analysis
   of differentially expressed genes identified in transcriptomic and
   proteomic analyses. I. Kaplan-Meier survival analysis of OHSU-AML
   dataset revealed that high expression of MMP14 is associated with
   poorer overall survival. J. PCR detection of MMP14 expression in AMLMSC
   (n = 12) compared to HDMSC (n = 6). K. Co-immunohistochemical staining
   of MMP14 (red) and CD271 (green) in AML and healthy control bone marrow
   samples. Scale bar = 20 μm (n = 5)

Knockdown of MSC-derived MMP14 inhibits MSC proliferation and induces
cytokine alterations

   To evaluate the impact of MMP14 upregulation in AMLMSC on MSCs, shRNA
   constructs (shMMP14#1/#2) were employed to inhibit MMP14 expression.
   Both shRNAs significantly reduced MMP14 expression at the protein and
   RNA levels (Fig. [73]2A, B). CFU-F assay revealed that MMP14 knockdown
   decreased the colony-forming ability of MSCs (Fig. [74]2C, D). A
   subsequent BrdU assay demonstrated reduced proliferation of MSCs in the
   MMP14 knockdown group (Fig. [75]2E, F). Cell cycle analysis via flow
   cytometry was conducted to determine if MMP14 knockdown-induced
   inhibition was attributable to cell cycle disruption. The analysis
   indicated a significant increase in the G0 phase of MSCs (Fig. [76]2G,
   H). Studies have demonstrated that MSCs can modulate the bone marrow
   microenvironment through cytokine secretion, which is closely related
   to AML pathogenesis and progression [[77]25, [78]26]. Therefore, we
   investigated whether MMP14 knockdown alters cytokine secretion by MSCs.
   After co-culturing MSCs from the knockdown and control groups with
   Kasumi-1 cells and Molm-13 cells for 48 h, we performed flow cytometric
   sorting of MSCs and PCR detection of leukemia progression-related genes
   (Fig. [79]2I and Fig. S3). The results showed significant reductions in
   TGF-β1, CXCL12, CXCL10, OPN, and COX-2, of which COX-2 and OPN have
   been shown to be associated with promoting leukemia progression
   [[80]10, [81]27], and CXCL12, TGF-β1, and CXCL10 have been reported to
   be associated with leukemia resistance [[82]13, [83]28,
   [84]29](Fig. [85]2J, K).

Fig. 2.

   [86]Fig. 2
   [87]Open in a new tab

   Effects of MMP14 on MSC proliferation and cytokine secretion. A, B.
   Knockdown efficiency of MMP14 in MSCs verified by Western blot and
   RT-qPCR. C, D. Representative plots of CFU-F in the MMP14 knockdown and
   control groups. Quantification of CFU-F colonies in MSCs. E, F. BrdU
   assay showing cell proliferation rates in the MMP14 knockdown and
   control groups. G, H. Cell cycle analysis of the MMP14 knockdown and
   control groups. I-K. After co-culturing Kasumi-1 and Molm-13 cells with
   MSCs from the knockdown and control groups for 48 h, flow sorting was
   performed on the MSCs followed by PCR analysis of TGF-β1, CXCL12,
   CXCL10, OPN, and COX-2 gene expression levels

Knockdown of MSC-derived MMP14 significantly inhibits AML progression in
vitro

   To elucidate the effect of MSC-derived MMP14 on AML progression, we
   co-cultured two AML cell lines, Kasumi-1 and Molm-13, with MSCs
   transduced with either sh-MMP14 or sh-NC. We observed several changes
   in the AML cells, notably an increase apoptosis rate in both Kasumi-1
   and Molm-13 upon MMP14 knockdown (Fig. [88]3A-C). Additionally, cell
   cycle analysis revealed that MMP14 inhibition in MSCs increased the
   proportion of cells in the G0/G1 phase while decreasing the proportion
   in the S phase (Fig. [89]3D-F). Furthermore, a BrdU assay demonstrated
   that MMP14 inhibition led to reduced proliferative capacity in both
   Kasumi-1 and Molm-13 cells (Fig. [90]3G-I). These findings suggest a
   potential role of MSC-derived MMP14 in the pathogenesis of AML.

Fig. 3.

   [91]Fig. 3
   [92]Open in a new tab

   Knockdown of MSC-derived MMP14 significantly inhibits AML progression
   in vitro. A-C. Flow cytometry analysis of apoptosis rates in Kasumi-1
   and Molm-13 cells co-cultured with the knockdown and control groups for
   48 h. D-F. Flow cytometry PI analysis of cell cycle changes in Kasumi-1
   and Molm-13 cells co-cultured with the knockdown and control groups for
   48 h. G-I. Flow cytometry BrdU analysis of cell proliferation rates in
   Kasumi-1 and Molm-13 cells co-cultured with the knockdown and control
   groups for 48 h

Inhibition of MSC-derived MMP14 delays AML progression

   We further established an animal model of AML to investigate the role
   of MSC-derived MMP14 in AML disease progression. Bone marrow
   YFP + cells were isolated from virus-transduced leukemia mice with
   MLL-AF9/YFP. These MLL-AF9 cells were transplanted into 7 Gy-irradiated
   mice via the tail vein and injected intrafemurally with BMMSCs from two
   groups: lentivirally-transfected MMP14 knockdown and control mice
   (Fig. [93]4A). We first examined whether MMP14 deletion in MSCs
   affected the survival of MLL-AF9 AML mice. The results, as shown in
   Fig. [94]4B, indicated that leukemia progression was delayed and
   survival was significantly prolonged following MMP14 knockdown.
   Additionally, tumor progression was monitored weekly using
   bioluminescent imaging, revealing that the tumor burden in the MMP14
   knockdown group was lower than that in the control group(Fig. [95]4C).
   Further comparison of the infiltration of leukemic cells (YFP + cells)
   and leukemic stem cells (LSC YFP + c-Kit + Gr-1-) in the two groups of
   mice revealed that the proportion of bone marrow leukemic cells and
   LSCs in the MMP14 knockout group was lower than in that the control
   group (Fig. [96]4D-G). Additionally, LSC characterization showed
   increased proportions of apoptotic cells and G0 phase cells in the
   MMP14 knockdown group (Fig. [97]4H-K). These results suggest that
   inhibition of MSC-derived MMP14 can effectively delay AML progression.

Fig. 4.

   [98]Fig. 4
   [99]Open in a new tab

   Effect of MMP14 deletion in MSCs on disease progression and leukemic
   stem cells in MLL-AF9 leukemic mice. (A) The experimental layout
   involved transplanting MLL-AF9 cells via the tail vein and MSCs from
   MMP14 knockdown and control groups via intra-femoral injection into
   mice. AML progression was monitored by flow cytometric analysis. (B)
   Survival rates of mice in MMP14 knockdown and control groups after cell
   transplantation (n = 7). (C) Live imaging was performed on days 14, 21,
   and 28 post-transplant to monitor disease progression and
   burden(n = 5). D, E. Flow cytometry plots of BM LSCs and statistical
   analysis of the proportion of LSC in whole BM (n = 4). F. Statistical
   analysis of the proportion of leukemia cells in mice of both groups
   (n = 4). G. Bone marrow cell count statistics in both groups of mice
   (n = 4). H, I. Representative flow cytometry plots of LSC apoptosis and
   statistical analysis of the apoptosis rate of LSCs (n = 4). J, K.
   Representative flow cytometry plots of the LSC cell cycle and
   statistical analysis of the percentage of apoptosis in LSCs (n = 4)

MMP14 exerts its functions by activating the JAK-STAT pathway in AML cells
through the secretion of PGE2

   To elucidate the potential mechanisms of MSC-derived MMP14 in AML
   development, we conducted RNA-seq analysis on Kasumi-1 cells
   co-cultured with MMP14-knockdown and control MSCs, with each group
   comprising three biological replicates. The volcano plot and heat map
   indicate that 507 genes were differentially altered, with 241 genes
   upregulated and 266 genes downregulated (Fig. [100]5A, B). KEGG
   enrichment analysis revealed that MMP14 knockdown in MSCs was
   associated with the TNF signaling pathway, NF-kappa B signaling
   pathway, cytokine-cytokine receptor interaction, and JAK-STAT signaling
   pathway (Fig. [101]5C). GSEA analysis indicated that MMP14 knockdown
   inhibited the JAK-STAT signaling pathway in Kasumi-1 cells
   (Fig. [102]5D). Therefore, we analyzed the expression of STAT3,
   P-STAT3, STAT5, and P-STAT5 in Kasumi-1 and Molm-13 cells co-cultured
   with MMP14-knockdown and control MSCs using Western blotting. As shown
   in Fig. [103]5E, knockdown of MMP14 in MSCs resulted in a decrease in
   the expression of P-STAT3 and P-STAT5 in leukemia cells, with no
   decrease in the expression of total STAT3 and STAT5. Given that studies
   have shown that MMP14 can regulate COX-2 expression, our data verified
   this as well. Western blotting showed that MMP14 knockdown in MSCs
   resulted in decreased expression of COX-2, a key enzyme in PGE2
   production (Fig. S4). Additionally, the literature reports that
   downregulation of PGE2 is associated with inhibition of the JAK-STAT
   pathway [[104]30]. Therefore, we hypothesized that MMP14 knockdown
   resulted in decreased PGE2 secretion by MSCs, which in turn inhibited
   the JAK-STAT pathway in AML cells. To this end, we examined PGE2 levels
   in the supernatants of MSCs from both groups. ELISA results showed that
   MMP14 knockdown in MSCs led to reduced PGE2 levels. Additionally, we
   assessed the PGE2 levels in the serum of mice from the MMP14 knockdown
   and control groups, finding that PGE2 levels were decreased in the
   serum of mice with MMP14 knockdown (Fig. [105]5F-G). To validate
   whether MSC-derived MMP14 exerts its effects through the secretion of
   PGE2, we added PGE2 to the co-culture system and observed that PGE2
   could reverse the inhibition of cell proliferation caused by MMP14
   knockdown (Fig. [106]5H-J).To further delineate the regulatory role of
   PGE2 on the JAK/STAT pathway, the expression of P-STAT3 and P-STAT5 was
   assessed using flow cytometry in the presence or absence of PGE2 within
   the co-culture systems of the knockdown and control groups. The results
   demonstrated that the knockdown of MMP14 in MSCs significantly
   inhibited the JAK/STAT pathway, an effect that was reversed by the
   addition of PGE2(Fig. S5). These data suggest that MSC-derived MMP14
   promotes AML progression by activating the JAK-STAT pathway in AML
   cells through PGE2 secretion.

Fig. 5.

   [107]Fig. 5
   [108]Open in a new tab

   MMP14 exerts its functions by activating the JAK-STAT pathway in AML
   cells through the secretion of PGE2. (A) Transcriptome analysis of
   Kasumi-1 cells co-cultured with MSCs from the MMP14 knockdown and
   control groups after 48 h. Volcano plots are shown. (B) Heat map of
   differentially expressed genes. (C) Enrichment analysis of
   differentially expressed genes. (D) GSEA analysis showing that MMP14
   knockdown in MSCs is associated with inhibition of the JAK-STAT
   pathway. (E) Western blot analysis of STAT3, P-STAT3, STAT5, and
   P-STAT5 protein levels in Kasumi-1 and Molm-13 cells co-cultured with
   MSCs from the MMP14 knockdown and control groups. (F) ELISA assay of
   PGE2 levels in the supernatant of MSCs from the MMP14 knockdown and
   control groups. (G) ELISA assay of PGE2 levels in the serum of mice
   from the MMP14 knockdown group and the control group. H-J.
   Proliferation of Kasumi-1 and Molm-13 cells in the presence or absence
   of PGE2 (1 μm) in the co-culture system of the knockdown and control
   groups

MMP14 confers chemoresistance to Ara-C

   Mesenchymal stem cells play a crucial role in protecting leukemia cells
   from chemotherapeutic agents. Studies have shown that activation of the
   JAK/STAT pathway may be involved in the mechanism of cytarabine
   resistance in AML [[109]31]. Therefore, we investigated the effect of
   MMP14 deletion on chemosensitivity to cytarabine. To this end, we
   established the MLL-AF9 leukemia model by intrafemoral injection of
   MMP14 knockdown or control MSCs, followed by tail vein injection of
   MLL-AF9 cells. The mice were subsequently divided into four
   experimental groups, with or without Ara-C treatment: the shNC group,
   the shMMP14 group, the shNC combined with Ara-C group, and the shMMP14
   combined with Ara-C group (Fig. S6A). Survival analysis revealed that
   survival time was significantly prolonged in the shMMP14 group, the
   shNC combined with Ara-C group, and the shMMP14 combined with Ara-C
   group, with the latter exhibiting the most pronounced survival benefit
   (Fig. S6B). Bone marrow smears and flow cytometric analysis further
   demonstrated that leukemic burden was more effectively suppressed in
   the shMMP14 group treated with Ara-C, compared to the shNC control
   group (Fig. S6C-F). These findings suggest that the inhibition of
   MSC-derived MMP14 enhances the cytotoxic effects of Ara-C in vivo. To
   explore the clinical relevance, we co-cultured Kasumi-1 and Molm-13
   leukemia cell lines with MSCs in vitro, and treated the co-cultures
   with either the specific MMP14 inhibitor NSC-405,020 [[110]32]or Ara-C
   to evaluate the combinatorial effect. Flow cytometric analysis of
   apoptosis revealed that both monotherapies and combination treatments
   induced apoptosis in a concentration-dependent manner, with the
   combination of the MMP14 inhibitor and Ara-C exhibiting a synergistic
   effect (CI < 1) (Fig. S7A-D). Additionally, Western blotting analysis
   of key proteins in the JAK/STAT pathway demonstrated that the
   combination therapy significantly reduced the levels of P-STAT3 and
   P-STAT5, thereby inhibiting the activation of the JAK/STAT signaling
   pathway(Fig. S7E).Furthermore, to investigate the therapeutic potential
   in vivo, we administered MMP14-specific pharmacological inhibitors
   (NSC-405020) to mice, dividing them into four groups: DMSO-treated
   control (Ctrl), MMP14 inhibitor (inhibitor), cytarabine (Ara-C), and
   MMP14 inhibitor combined with cytarabine (inhibitor + Ara-C). The
   experimental design is illustrated in Fig. [111]6A. At the conclusion
   of the experiment, the highest survival rate was observed in the MMP14
   inhibitor combined with cytarabine group (Fig. [112]6B). Additionally,
   the proportion of bone marrow and splenic leukemia cells in mice was
   assessed by flow cytometry. The proportion of bone marrow and splenic
   leukemia cells in mice treated with the combination of MMP14 inhibitor
   and cytarabine was significantly lower than in the control and
   single-agent groups (Fig. [113]6C-E). Furthermore, bone marrow smears
   demonstrated a significant reduction in leukemia cells in the MMP14
   inhibitor combined with cytarabine group, and infiltration of splenic
   leukemia cells was diminished as evidenced by histological H&E staining
   (Fig. [114]6F). Spleen size was significantly reduced in the MMP14
   inhibitor combined with cytarabine group (Fig. [115]6G). Reductions in
   P-STAT3 and P-STAT5 were more pronounced in the combination therapy
   group compared to the monotherapy group (Fig. [116]6H-J). These data
   suggest that MMP14 confers resistance to Ara-C chemotherapy.

Fig. 6.

   [117]Fig. 6
   [118]Open in a new tab

   MMP14 confers chemoresistance to Ara-C. A. Overview of the experimental
   design, dividing mice into four groups based on the presence or absence
   of the inhibitor (NSC-405020) and cytarabine. B. Survival statistics of
   mice (n = 7). C-E. Representative FACS analysis and statistical
   proportions of leukemia cells in the four groups of mice (n = 4). F.
   Representative images of Wright-Giemsa staining of bone marrow cells
   and hematoxylin-eosin staining of spleens. G. Representative images of
   mouse spleens. H-J. Immunohistochemical staining of P-STAT3 and P-STAT5
   in mouse bone marrow and quantification of the staining. Scale bar,
   50 μm

Inhibition of MMP14 suppressed AML patient blasts

   To further elucidate the role of MSC-derived MMP14 in primary leukemia
   cells, we obtained cells from two distinct patients and co-cultured
   them with MMP14 knockdown and control MSCs. Flow cytometry analysis of
   apoptosis revealed that knockdown of MMP14 increased the apoptosis rate
   in primary leukemia cells (Fig. [119]7A-C). Additionally, we employed
   trypan blue staining to count the viable cells after co-culture, and
   the results showed that the viable cell counts of primary cells
   co-cultured with MSCs from the knockdown group were lower than those of
   the control group (Fig. [120]7D, E). We subsequently assessed the
   impact of MSC-derived MMP14 knockdown on normal bone marrow-derived
   CD34 + cells and showed that knockdown of MMP14 did not significantly
   affect apoptosis and proliferation of normal CD34 + cells (Fig. S8). To
   evaluate the in vivo effect of MMP14 knockdown on AML blasts,
   patient-derived cells were transplanted into NSG mice to establish a
   PDX model. This model was stratified into four groups: DMSO-treated
   control (Ctrl), MMP14 inhibitor (inhibitor), cytarabine (Ara-C), and
   MMP14 inhibitor combined with cytarabine (inhibitor + Ara-C)
   (Fig. [121]7F). The results revealed that in vivo inhibition of MMP14
   expression prolonged the survival of PDX mice. More importantly, the
   combined MMP14 inhibitor and cytarabine group had a significant
   survival advantage compared to the inhibitor or cytarabine alone
   (Fig. [122]7G), and the results showed that the inhibitor of MMP14
   group reduced the leukemia burden by flow cytometric analysis of human
   CD45 + cells in the BM of PDX mice, the combination of the administered
   drug group had the lowest proportions and absolute numbers of human
   CD45 + cells, consistent with our findings in the mouse model of
   MLL-AF9 (Fig. [123]7H-J). In conclusion, our findings suggest that
   inhibition of MMP14 suppresses AML patient blasts, underscoring its
   potential as a therapeutic target in leukemia.

Fig. 7.

   [124]Fig. 7
   [125]Open in a new tab

   Inhibition of MMP14 suppressed AML patient blasts. A-C. Flow cytometry
   analysis of apoptosis rates in primary leukemia cells co-cultured with
   the knockdown and control groups for 48 h. Alone, leukemia cells
   cultured alone; Co, leukemia cells cocultured with MSCs. D, E. Viable
   cells were counted using trypan blue staining and assessed with a
   hemocytometer after 48 h of co-culturing. F. Schematic of the PDX model
   construction process. Primary human leukemia cells were isolated from
   high-leukemia AML patients using leukapheresis and PBMC separation.
   These cells were then injected into NSG mice via the tail vein. Four
   weeks later, one mouse per week was randomly sacrificed, and the
   proportion of leukemia cells in the BM was analyzed by flow cytometry.
   Drug treatment commenced when the leukemia cell proportion
   reached ≥ 20%. Following successful PDX model construction, mice were
   randomly allocated to four treatment groups: DMSO-treated control
   (Ctrl), MMP14 inhibitor (inhibitor), cytarabine (Ara-C), and a
   combination of MMP14 inhibitor and cytarabine (inhibitor + Ara-C). G.
   Survival statistics for the mice were compiled (n = 6). H-J. Flow
   cytometric representative plots and statistical analysis detailing the
   proportion and absolute numbers of human CD45 + cells in the four mouse
   groups (n = 5)

Discussion

   The contribution of the bone marrow microenvironment to leukemogenesis
   and progression has been increasingly recognized [[126]33–[127]36].
   However, the specific contribution of BM MSCs to leukemia niche
   formation and progression remains unclear. Therefore, understanding the
   role of BM MSCs in leukemia is crucial for identifying disease-specific
   therapeutic targets. MMP14 is elevated in numerous cancers and
   facilitates tumor angiogenesis, inflammation, and progression [[128]24,
   [129]37]. Although there is substantial evidence that MMP14 is
   aberrantly expressed in leukemia [[130]22, [131]38], the role of
   MSC-derived MMP14 in AML and its impact on the microenvironment remain
   unclear. Here, we demonstrate that AML-MSCs are molecularly and
   functionally distinct from normal donor-derived MSCs. For the first
   time, we demonstrate that MSC-derived MMP14 promotes MSC proliferation
   and accelerates AML progression. Furthermore, in vivo experiments
   demonstrated that inhibiting MMP14 increased chemosensitivity to
   cytarabine in AML mice. Mechanistically, the supportive effect of
   MSC-derived MMP14 on AML may operate through the secretion of PGE2,
   which activates the JAK-STAT pathway in leukemia cells (Fig. [132]8).
   This underscores the pivotal role of MSC-derived MMP14 in AML
   pathogenesis and progression, offering fresh evidence for the
   involvement of MSCs in AML development.

Fig. 8.

   [133]Fig. 8
   [134]Open in a new tab

   Mechanism of action of MSC-derived MMP14 in promoting AML progression.
   The model illustrates how MSC-derived MMP14 promotes AML cell growth
   and chemotherapy resistance through the secretion of PGE2, activating
   the JAK-STAT pathway, leading to AML progression. Additionally,
   MSC-derived MMP14 enhances MSC proliferation and self-renewal, while
   also inducing cytokine alterations

   In this study, we co-cultured AML-MSCs and HD-MSCs with leukemia cells
   in vitro to compare their ability to support AML cells, and found that
   AML-MSCs were more supportive of AML cell growth compared to HD-MSCs.
   Previous studies have reported differences among the MSCs from leukemia
   patients compared to those from healthy donors [[135]39–[136]42].
   Additionally, numerous studies have shown that interactions between
   leukemia cells and cellular components in the bone marrow
   microenvironment are key in chemoresistance, clonal proliferation, and
   cell growth, as well as in disease relapse [[137]10, [138]43]. The
   contribution of underlying molecular alterations in AML-MSCs to
   leukemogenesis and chemoresistance remains controversial, emphasizing
   the need to study these cells further. To elucidate why AML-MSCs
   support leukemia cell growth, we performed transcriptomic and proteomic
   analyses, along with joint OHSU-AML datasets prognostic analysis, which
   preliminarily confirmed that MMP14 plays a key role in AML development
   and progression.

   Accumulating evidence strongly suggests that MSCs play a critical role
   in leukemia progression, primarily through the secretion of specific
   cytokines, growth factors, and chemokines. Carter et al. revealed that
   co-culturing AML cells with MSCs significantly increased COX-2
   expression and PGE2 production in MSCs, thereby promoting leukemia cell
   growth via the β-catenin signaling pathway [[139]10]. Moreover, Xiao et
   al. identified that OPN is significantly upregulated in the AML bone
   marrow microenvironment, contributing to the formation of a leukemic
   niche that supports AML cell survival and proliferation [[140]27].
   Additionally, the CXCL12-CXCR4 axis plays a pivotal role in AML
   progression and drug resistance, with studies indicating that
   disruption of this axis can enhance the efficacy of targeted therapies
   [[141]44, [142]45]. Research indicates that TGF-β1 secreted by MSCs
   enhances the resistance of AML cells to cytarabine therapy through the
   TGF-β/p38/ALDH2 signaling pathway [[143]28]. Furthermore, CXCL10 was
   shown to promote leukemia cell survival and drug resistance by enabling
   cells to evade apoptosis induced by chemotherapeutic agents like
   cytarabine and methotrexate [[144]29]. Building upon these insights, we
   selected several molecules closely associated with leukemia progression
   and examined the impact of MMP14 knockdown on the gene expression of
   cytokines in MSCs. Our results are consistent with published studies,
   though further investigation is warranted to elucidate the functional
   implications. The role of PGE2 as an autocrine and paracrine signaling
   molecule has been well documented in cancer, where it promotes tumor
   cell proliferation, survival, invasion, and migration [[145]46]. The
   JAK/STAT signaling pathway, a downstream pathway of cytokine signaling,
   is involved in crucial biological processes such as cell proliferation,
   differentiation, apoptosis, and immune regulation. It is strongly
   linked to the progression of hematological malignancies [[146]47,
   [147]48].Our data show that MSC-derived MMP14 activates the JAK-STAT
   pathway in leukemia cells by regulating PGE2 secretion.

   Drug resistance remains a central challenge in treating AML. Despite
   recent advances in targeted therapies and the development of drugs for
   specific gene mutations, the efficacy of current treatments continues
   to face significant hurdles [[148]49, [149]50]. Emerging research
   suggests that interactions between LSCs and associated stromal cells
   critically influence leukemia onset, progression, and therapy response.
   Within the bone marrow niche, stromal cells provide a sanctuary that
   allows LSCs to develop drug-resistant phenotypes and evade
   chemotherapy-induced death. Cytarabine serves as a first-line agent in
   AML treatment. While cytarabine-based regimens can achieve complete
   remission in the majority of newly diagnosed AML patients, overall
   clinical outcomes are often unsatisfactory due to persistent drug
   resistance. Multiple mechanisms are strongly associated with resistance
   to cytarabine [[150]51–[151]53]. For instance, targeting the bone
   marrow microenvironment, specifically Lama4, has been shown to mitigate
   Ara-C chemoresistance in AML cells [[152]35]. Our study reveals that
   MSC-derived MMP14 bestows cytarabine resistance in AML LSCs.
   Mechanistically, inhibiting MMP14 in AML models enhances sensitivity to
   cytarabine by disrupting the JAK-STAT pathway.

   Dissecting the crosstalk between leukemia cells and the BM
   microenvironment will not only enhance our understanding of
   leukemogenesis but also help identify new strategies to target
   microenvironment-mediated LSC survival and drug resistance. Future
   studies should explore the effects of MSC-derived MMP14 on bone marrow
   microenvironment components and hematopoiesis. In conclusion, our study
   provides the first evidence of the effect of MSC-derived MMP14 on AML
   progression. Additionally, MSC-secreted MMP14 promotes MSC
   proliferation and self-renewal, while also inducing cytokine
   alterations. MSC-derived MMP14 may promote AML progression and increase
   drug resistance through the secretion of PGE2, which activates the
   JAK-STAT pathway in leukemia cells. Therefore, strategies to inhibit
   MSC-derived MMP14 may represent a promising therapeutic approach to
   inhibit AML and sensitize AML LSCs to chemotherapy, thereby improving
   treatment outcomes.

Conclusions

   Our study provides the first evidence that MSC-derived MMP14 promotes
   AML cell growth and chemotherapy resistance via the secretion of PGE2,
   which activates the JAK-STAT pathway, consequently leading to AML
   progression. Additionally, MSC-derived MMP14 enhances MSC proliferation
   and self-renewal, concurrently inducing alterations in cytokine
   profiles. Targeting the MMP14 signaling pathway may offer new
   therapeutic options for AML.

Electronic supplementary material

   Below is the link to the electronic supplementary material.
   [153]Supplementary Material 1^ (914.8KB, pdf)

Acknowledgements