Human umbilical cord mesenchymal stem cells improve bone marrow hematopoiesis through regulation of bone marrow adipose tissue

Abstract

Bone marrow adipose tissue (BMAT) exhibits a multitude of biological functionalities and influences hematopoiesis. The adiposity status of the bone marrow may play a role in the decline of hematopoietic function. Mesenchymal stem cells (MSCs) constitute crucial regulators within the bone marrow microenvironment; however, their precise role in modulating BMAT and the subsequent implications for hematopoiesis remain poorly understood. We conducted in vivo studies to observe the effects of human umbilical cord mesenchymal stem cells (hucMSCs) on BMAT accumulation and restoration of hematopoietic function in mice with drug-induced hematopoietic impairment. Concurrently, in vitro co-culture experiments were used to investigate the impact of hucMSCs on preadipocytes and mature adipocytes, and the potential subsequent consequences for hematopoietic cells. Moreover, we explored the potential mechanisms underlying these interactions. Our findings reveal that hucMSCs concomitantly mitigate BMAT accumulation and facilitate the recovery of hematopoietic function in mouse models with drug-induced hematopoietic impairment. In vitro, hucMSCs potentially impede adipogenic differentiation of 3T3-L1 preadipocytes through interference with the JAK2/STAT3 signaling pathway and affect the functionality of mature adipocytes, thus mitigating the detrimental effects of adipocytes on hematopoietic stem cells (HSCs). Furthermore, we demonstrate that hucMSCs may protect hematopoietic cells from adipocyte-induced damage by protecting antioxidative mechanisms. These results suggest that hucMSCs exhibit an inhibitory effect on the excessive expansion of adipose tissue and modulate adipose tissue function, which may potentially contribute to the regulation of the bone marrow microenvironment and favorably influence hematopoietic function improvement.

Supplementary Information

The online version contains supplementary material available at 10.1007/s11010-024-05156-0.

Introduction

Bone marrow failure (BMF) comprises a group of diseases characterized by reduced number and function of hematopoietic stem cells (HSCs), leading to a decrease in one or more lineages of blood cells within the bone marrow (BM) and peripheral circulation. A multitude of factors can precipitate BMF, encompassing not only age, immunological aspects, but also the frequently iatrogenic side effects of radiotherapy and chemotherapy, which are often accompanied by impairment of hematopoietic function [1, 2]. Although chemotherapy-induced hematopoietic impairment is generally limited and capable of self-recovery, the extent of recovery is intimately tied to the treatment efficacy and prognosis of diseases [3]. Beyond the direct cytotoxic impact of chemotherapeutic agents on hematopoietic cells, these treatments also exert effects on other cellular constituents of the bone marrow niche, thereby posing additional impediments to the restoration of hematopoietic function [4].

The niche that hematopoietic cells reside within the BM is known as the BM microenvironment (BMM). The BMM constitutes a sophisticated system composed of a diverse array of cellular components, including hematopoietic cells, mesenchymal stem cells (MSCs), bone marrow adipocytes, osteoblasts, and fibroblasts, among others. Their quantities, as well as the cytokines and exosomes they release, collectively influence the stability of the BMM [5]. Another research demonstrates that this relatively stable microenvironment provides essential support for hematopoietic activities [6], and aberrant alterations in the BMM are commonly observed in BMF [7]. Similar to BMF resulting from other etiologies, chemotherapy-induced hematopoietic injury also manifests characteristic pathological alterations of marrow adiposity [8]. The expansion of adipose tissue in the BM is perceived as detrimental to hematopoiesis [9] and is concurrently implicated in drug resistance and metastasis associated with certain malignancies [10].

Bone marrow adipose tissue (BMAT), accounting for approximately 10% of total body adipose tissue, primarily originates from bone marrow mesenchymal stem cells (BMSCs) [11]. Bone marrow adipocytes display distinct biological properties relative to adipocytes at other sites [12]. Under physiological conditions, BMAT serves as a crucial participant and regulator in the BM microenvironment, providing energy to neighboring hematopoietic cells and releasing free fatty acids, stem cell factors, and some adipokines such as adiponectin, all of which contribute significantly to hematopoietic support [13, 14]. However, excessive expansion of BMAT compresses the living space for hematopoietic cells [15]. Simultaneously, the excessive expansion of BMAT not only fosters a transition toward a white adipocyte phenotype but also induces alterations in the secretory profile, characterized by elevated levels of pro-inflammatory adipokines [16]. Moreover, the aforementioned alteration of BMAT also leads to a reduction in the numbers of BMSCs and osteoblasts, impairing their supportive roles in hematopoiesis [17]. Some researchers are currently investigating the potential benefits of modulating BMAT in the context of hematopoiesis, with initial encouraging results suggesting that BMAT represents a viable therapeutic target in BMF [18].

Mesenchymal stem cellss are a type of multipotent stem cells ubiquitously present in various tissues. They have garnered increased attention owing to their low immunogenicity and demonstrated potential in anti-inflammatory, immunomodulatory, and regenerative applications [19]. Umbilical cord mesenchymal stem cells (ucMSCs), derived from the Wharton's jelly of the umbilical cord, exhibit analogous biological characteristics to BMSCs. Due to their hematopoietic support and immunomodulatory potential, as well as their easier obtainment compared to BMSCs, ucMSCs have become a promising option for treating hematological disorders [20]. Currently, ucMSCs have been employed in facilitating hematopoietic recovery post-transplantation and the prophylaxis of graft-versus-host disease (GVHD), with favorable clinical outcomes [21]. Moreover, several researchers have tried to utilize ucMSCs in treating BMF resulting from different causes. These studies have revealed encouraging outcomes, especially in studies involving aplastic anemia (AA) and radiation-induced hematopoietic impairment [22, 23]. Although the therapeutic efficacies observed in these contexts are primarily attributed to the immunomodulatory and antioxidative properties of ucMSCs [24, 25], we posit that there may be other synergistic mechanisms involved. Concurrently with the improvement of hematopoietic function by hucMSCs, changes in BMAT were observed [26]. This suggests a possible correlation between the two mechanisms; however, the exploration of whether hucMSCs modulate BMAT to contribute to the restoration of hematopoietic function during their therapeutic action in hematopoietic impairment is still pending.

To investigate whether hucMSCs participate in the treatment of hematopoietic injury by modulating BMAT, we explored the regulatory effects of hucMSCs on adipocytes and the related mechanisms. Additionally, we induced hematopoietic injury models in vitro and in vivo using chemical reagents to evaluate the impact of hucMSCs on hematopoietic recovery through BMAT. Our data indicate that hucMSCs affect the biological characteristics and functions of adipocytes, thereby providing support for the recovery from hematopoietic injury. This work aims to provide novel theoretical foundations for the application of hucMSCs in the treatment of hematopoietic impairment.

Materials and methods

Isolation, culture, and characterization of hucMSCs

Umbilical cord tissue was obtained from healthy donors delivering at the Second Hospital of Shanxi Medical University, who provided written informed consent. The study was approved by the Ethics Committee of the Second Hospital of Shanxi Medical University. The isolation of MSCs from umbilical cords was performed as previously described in the literature [27]. Briefly, the epidermis of fresh human umbilical cord was incised under sterile conditions, and Wharton's jelly was blunt-dissected after the removal of blood vessels. The isolated Wharton's jelly was minced and seeded onto the bottom of culture dishes containing Dulbecco's Modified Eagle Medium/Nutrient Mixture F-12 (DMEM/F12, Gibco) supplemented with 10% fetal bovine serum (FBS, Gibco). The culture dishes were incubated at 37 °C and 5% CO₂, until the adherent cells reached 80% confluence, at which point they were passaged at a 1:3 ratio. The hucMSCs from passages 3 to 5 were utilized for subsequent experiments [28].

Flow cytometry was used to examine the immunophenotype of hucMSCs. We used a Navios Flow Cytometer (Beckman Coulter) to identify the surface molecular markers of hucMSCs. Details of antibodies are listed in Table S1. Furthermore, we subjected hucMSCs to adipogenic and osteogenic differentiation induction, followed by identification using Oil Red O (Solarbio) or Alizarin Red S (Solarbio) staining.

Mouse BMF model and treatment

Female Institute of Cancer Research Mice (ICR mice), purchased from the Animal Experiment Center of Shanxi Medical University, were housed in a barrier environment under standard management practices. All animals had free access to food and water and were used at 8–12 weeks of age. All animal studies were approved by the Animal Ethics Committee of the Second Hospital of Shanxi Medical University.

Eight-week-old female mice were randomly assigned to a normal control group (NC group, n = 6) and a model group (n = 15). One week prior to the experiment, all mice underwent pre-adaptation feeding. On 1st, 3rd, 5th, and 7th days, the model group received intraperitoneal injections of busulfan (BUS, 20 mg/kg, Sigma-Aldrich) and cyclophosphamide (CTX, 80 mg/kg, Sigma-Aldrich) to induce BMF, while the control group received intraperitoneal injections of an equivalent volume of saline. On day 9, peripheral blood cell counts were assessed, and mice with significantly reduced total blood cells relative to controls were considered successfully modeled.

The modeled mice were then randomly allocated to the BMF group (n = 6) and the BMF + MSC group (n = 6). On 9th and 17th days, respectively, the BMF + MSC group received tail vein injections of Dulbecco's Phosphate-Buffered Saline (DPBS, Gibco) containing 5 × 10⁵ hucMSCs per mouse, while the NC group and the BMF group received an equivalent volume of DPBS (Gibco). On day 23, peripheral blood was collected from the retro-orbital venous plexus and BM was harvested following euthanasia under anesthesia, for further analysis.

Peripheral blood count and bone marrow histology

Peripheral blood samples were collected in Eppendorf tubes containing Ethylene Diamine Tetraacetic Acid (EDTA). White blood cell (WBC), red blood cell (RBC), hemoglobin (HGB), and platelet (PLT) counts were determined using a XE-2100 automated hematology analyzer (Sysmex).

Femurs were fixed in 4% paraformaldehyde (Solarbio) and decalcified in EDTA before being embedded in paraffin. Paraffin blocks were sectioned at a thickness of 5 μm and were subjected to routine hematoxylin and eosin (H&E) staining. Slides were examined under an Olympus CKX53 microscope and images were acquired using the Pannoramic MIDI (3DHISTECH) scanning system.

Differentiation and co-culture of 3T3-L1 cells

Adipogenic differentiation of 3T3-L1 cells (Boster) was induced by using a conventional cocktail protocol. Briefly, 3T3-L1 cells were seeded onto gelatin-coated culture plates. Upon reaching full confluence, cells were exposed to an adipogenic induction medium composed of Dulbecco's Modified Eagle Medium (DMEM, Gibco) supplemented with 10% FBS (Gibco), 2 μmol/L rosiglitazone (Sigma-Aldrich), 1 μmol/L dexamethasone (Sigma-Aldrich), 0.5 mmol/L 3-Isobutyl-1-Methylxanthine (IBMX, Sigma-Aldrich), and 10 mg/L insulin (Sigma-Aldrich). The medium was refreshed every 3 days throughout the differentiation process.

3T3-L1 cells were cultured either alone or in co-culture with hucMSCs. Co-culture was conducted using a 35 mm Transwell chamber system equipped with 0.4 μm pore polyester membrane inserts (Corning). HucMSCs (1 × 10⁵ cells/well) were seeded in the upper insert, while 3T3-L1 cells (5 × 10⁵ cells/well) were placed in the lower chamber. Co-cultures were maintained in DMEM (Gibco) containing 10% FBS (Gibco) or adipogenic induction differentiation medium.

Fluorescence staining, oil red O staining, and determination of triglyceride (TG)

3T3-L1 cells were fixed with 4% paraformaldehyde (Solarbio). Fluorescence staining was performed with Immunostaining Permeabilization Buffer with Saponin (Beyotime) for permeabilization, followed by blocking with QuickBlock Immune Staining Blocking Solution (Beyotime). Cell membranes were stained with DIO (Beyotime) and nuclei were counterstained with DAPI (Beyotime). Fluorescence microscopy (Olympus) was performed for observation and image acquisition. Following permeabilization of cells with 60% isopropanol, Oil Red O staining was conducted using Oil Red O working solution, followed by rinsing with running water and microscopic examination.

TGs were quantified using the Triglyceride Assay Kit (Abbkine CheKine). Following cell lysis, TGs were hydrolyzed, subjected to glycerol periodate oxidation, and condensed with acetylacetone, followed by absorbance measurement at 420 nm. TG concentrations were calculated based on a standard curve. Concurrently, protein concentrations in the cell lysates were determined using the Bradford method. The relative expression of TGs was represented as the ratio of TG concentration to protein concentration.

RNA extraction and quantitative reverse transcription polymerase chain reaction (RT-qPCR) assay

The experiment was performed as previously described [29]. In brief, total RNA was extracted from cells using TRIzol (TaKaRa) and reverse transcribed into cDNA using the PrimeScript RT reagent Kit (TaKaRa). Quantitative real-time PCR was then carried out with the Universal Blue qPCR SYBR Green Master Mix (Yeasen), and gene expression was assessed using the 7500 Real-Time PCR System (Applied Biosystems) according to the manufacturer's instructions. The relative mRNA expression levels of target genes in the samples were calculated as 2^−ΔΔCt, with GAPDH serving as the internal control. Primer sequences are listed in Table S2.

Western blot (WB) analysis

The experiment was performed as previously described [29]. Briefly, cells were lysed thoroughly in RIPA Lysis Buffer (Servicebio) on ice, followed by centrifugation at 13,000 rpm for 15 min at 4 °C. The supernatant containing protein solution was collected and its concentration was determined. Proteins were mixed with loading buffer (Wanleibio), heat-denatured, separated by Sodium Dodecyl Sulfate–Polyacrylamide Gel Electrophoresis (SDS-PAGE), and transferred onto a 0.22 μm polyvinylidene fluoride membrane (Bio-Rad). After blocking with 5% skim milk powder (BD Difco) at room temperature for 2 h, the membrane was incubated overnight at 4 °C with the following primary antibodies. Details of antibodies are listed in Table S3. Subsequently, the membrane was incubated with the corresponding HRP-conjugated (Proteintech) secondary antibodies for 1.5 h, and chemiluminescence detection was performed using the ChemiDoc MP Imaging System (Bio-Rad).

RNA sequencing

Transcriptome libraries were constructed using the Universal V6 RNA-seq Library Prep Kit (VAHTS), followed by sequencing on the Novaseq 6000 (Illumina). Differential gene expression analysis was conducted using DESeq2 software, with genes considered differentially expressed if they exhibited an adjusted q-value < 0.05 and a fold change of either > 2 or < 0.5. The identified differentially expressed genes (DEGs) were further subjected to clustering analysis and visualization using R 3.2.0 software. The transcriptome sequencing and analysis were conducted by OE Biotech Co., Ltd (Shanghai, China).

Isolation and co-culture of mouse bone marrow cells

Under sterile conditions, BM cells were flushed from femurs and tibias using a syringe and subsequently filtered through a 75 μm mesh. Red blood cells were removed using Mouse Bone Marrow Lymphocyte Separation Solution (Solarbio). The isolated cells were cultured in Iscove's Modified Dulbecco's Medium (IMDM, Gibco) supplemented with 10% FBS (Gibco), 100 U/mL recombinant murine interleukin-3 (IL-3, PeproTech), 100 U/mL recombinant murine interleukin-6 (IL-6, PeproTech), and 100 U/mL recombinant murine stem cell factor (SCF, PeproTech). Cultures were maintained in a 37 °C incubator with a 5% CO₂ atmosphere.

Mouse BM cells were cultured either individually or co-cultured with 3T3-L1 cells using a 35 mm Transwell chamber system featuring 0.4 μm pore polycarbonate membrane inserts (Corning). In this setup, mouse bone marrow cells (2.5 × 10⁵ cells/well) were seeded in the upper insert, while 3T3-L1 cells (5 × 10⁵ cells/well) were placed in the lower chamber. Co-cultures were maintained in IMDM (Gibco) medium containing 10% FBS (Gibco) and the aforementioned cytokines. Cells in the chemotherapy intervention group were treated with CTX (320 μM, Sigma-Aldrich) and BUS (80 μM, Sigma-Aldrich) for 24 h. Following this, the chemotherapy drugs were removed from the cells, and subsequent cultivation was conducted as previously specified.

Flow cytometry

Apoptosis, primitive cell status, cellular differentiation, and reactive oxygen species (ROS) levels were assessed using the Navios Flow Cytometer (Beckman Coulter) and analyzed with Kaluza software. Briefly, 3T3-L1 cells or mouse bone marrow cells were collected in PBS. Cell apoptosis was detected using the Annexin V-FITC/PI Apoptosis Detection Kit (Yeasen). Mouse antibodies were used to assess BM primitive cells and BM cell differentiation status. Details of antibodies are listed in Table S4. ROS levels were quantified using the Reactive Oxygen Species Assay Kit (Beyotime). For cell cycle analysis, 70% ethanol-fixed 3T3-L1 cells were stained using the Cell Cycle and Apoptosis Analysis Kit (Yeasen). Data were acquired using the Cytomics FC 500 flow cytometer (Beckman Coulter) and analyzed with ModFit LT software.

Statistical analysis

Statistical analyses were performed using SPSS 25.0 software and GraphPad Prism 8.0.1. All data are presented as means ± standard deviation (SD). Independent sample t test, one-way ANOVA, and LSD-t test were used to assess the statistical significance of differences between groups, with p < 0.05 considered statistically significant.

Results

HucMSCs ameliorate hematopoiesis and reduce BMAT in chemically induced BMF mice

First, we evaluated the therapeutic efficacy of hucMSCs in a chemically induced BMF mouse model (Fig. 1a). The MSCs used were isolated from human umbilical cords. Their characteristic features are presented in Fig. S1. These mice were then administered two rounds of treatment with hucMSCs within 14 days after the BMF model was established. This treatment resulted in a significant improvement in the bone marrow Lin-Sca-1 + c-Kit + cells (LSK cells) as well as peripheral WBC, RBC, HGB, and PLT levels in the BMF mice (Fig. 1b-c). Histopathology of the bone marrow revealed a significant decrease in the number and density of nucleated cells, accompanied by an increase in lipid vacuolation in BMF mice. The hucMSC treatment resulted in an augmented number of nucleated cells and reduction in lipid vacuolation within the BM (Fig. 1d). Additionally, hucMSCs significantly downregulated the mRNA (Fig. 1e) and protein (Fig. 1f) expression levels of Peroxisome proliferator-activated receptor gamma (PPARG), CCAAT/enhancer-binding protein alpha (CEBPA), and Fatty acid-binding protein 4 (FABP4) in the BM of the BMF mice. Collectively, these data indicate that hucMSCs can effectively restore hematopoietic function in BMF mice while concurrently suppressing BMAT expansion.

Fig. 1.

HucMSCs can mitigate hematopoietic dysfunction and suppress BMAT expansion in chemotherapy drug-induced BMF mice. a Schematic representation of BMF mice model establishment and treatment with hucMSCs. b Bone marrow LSK cells analysis on Day 23. c Blood counts of WBC, RBC, HGB, and PLT on Day 23. d Femur of each mouse was sectioned and stained with hematoxylin and eosin. e–f Relative mRNA and protein expression of PPARG, CEBPA, and FABP4 in bone marrow of each group of mice. NC, normal control; BMF, BMF mice without treatment of hucMSCs; BMF-MSC, BMF mice with treatment of hucMSCs. Data represent mean ± SD (n = 6 mice per group); *p < 0.05, **p < 0.01, and ***p < 0.001 indicate significant differences (one-way ANOVA, LSD-t test)

HucMSCs enhance the activity of 3T3-L1 preadipocytes and disrupt functions of mature adipocytes

Next, we conducted a more detailed investigation at the cellular level to elucidate the observed therapeutic effects. We used 3T3-L1 cells as a model of preadipocytes and investigated the effect of hucMSCs on these cells and their differentiated adipocytes under indirect co-culture conditions. Results showed that hucMSCs decreased the proportion of 3T3-L1 cells in the G0/G1 phase and increased the proportion in the S phase, indicating that hucMSCs could modulate the cell cycle of 3T3-L1 cells (Fig. 2a). Additionally, hucMSCs suppressed apoptosis of 3T3-L1 cells (Fig. 2b). Subsequently, 3T3-L1 cells were differentiated into adipocytes for 2 weeks until they displayed prominent lipid droplets, after which they were co-cultured with hucMSCs for 48 h. Results revealed that hucMSCs promoted apoptosis (Fig. 2c), concurrently impairing adipocyte function through interference with the expression of functionally pertinent mRNAs and proteins (Fig. 2d-e), as evidenced by reduced TG levels (Fig. 2f). These data suggest that hucMSCs enhance preadipocyte activity and disrupt mature adipocyte function.

Fig. 2.

HucMSCs can enhance activity of 3T3-L1 preadipocytes and disrupt functions of mature adipocytes. a–b Cell cycle distribution and apoptosis profile of 3T3-L1 cells after co-culture with hucMSCs for 48 h. c Apoptosis profile of mature adipocytes after co-culture with hucMSCs for 48 h. d–e Relative mRNA and protein expression of function-related genes in mature adipocytes after co-culture with hucMSCs for 48 h. f TG relative quantitation of mature adipocytes after co-culture with hucMSCs for 48 h. 3T3-L1, 3T3-L1 cells; 3T3-L1-MSC, 3T3-L1 cells co-cultured with hucMSCs; MAD, mature adipocytes; MAD-MSC, mature adipocytes co-cultured with hucMSCs. Data represent mean ± SD (n = 3 independent experiments); *p < 0.05, **p < 0.01, and ***p < 0.001 indicate significant differences (independent t test)

HucMSCs inhibit differentiation of 3T3-L1 preadipocytes

Subsequently, we sought to determine whether hucMSCs could also modulate preadipocyte differentiation. To this end, we subjected 3T3-L1 cells to adipogenic differentiation using a conventional cocktail protocol. Membrane fluorescence staining performed during the initial 24 h of preadipocyte differentiation demonstrated that under the influence of inducers, the morphology of 3T3-L1 cells gradually transformed from spindle-shaped to round or oval; in contrast, during co-culture with hucMSCs, the majority of 3T3-L1 cells maintained a spindle-like shape akin to the state before differentiation (Fig. 3a). Furthermore, Oil Red O staining (Fig. 3b) and TG quantification (Fig. 3c) over the first 72 h of 3T3-L1 differentiation both demonstrated an inhibitory effect of hucMSCs on lipid synthesis. To gain deeper insight into the impact of hucMSCs on preadipocyte differentiation, we assessed the expression of genes related to adipogenesis, lipid synthesis, and transport. While 3T3-L1 cells co-cultured with hucMSCs exhibited an increasing trend in mRNA and protein levels of PPARG, CEBPA, CCAAT/enhancer-binding protein beta (CEBPB), FABP4, Fatty acid synthase (FASN), and Acetyl-CoA carboxylase 1 (ACC1) during the initial 72 h of differentiation, these levels were significantly lower than those in 3T3-L1 cells undergoing differentiation at the same time points (Fig. 3d–e). These observations suggest that hucMSCs disrupt preadipocyte differentiation at the transcriptional level.

Fig. 3.

HucMSCs can inhibit differentiation of 3T3-L1 preadipocytes. a Fluorescence staining of 3T3-L1 cells at 0, 12, and 24 h during differentiation, co-cultured with hucMSCs. Green fluorescence represents the cell membrane labeled with DIO and blue fluorescence represents the nucleus labeled with DAPI. b Oil Red O staining of 3T3-L1 cells at 0, 24, 48, and 72 h during differentiation, co-cultured with hucMSCs. Red areas represent the TGs. c TG relative content in each group of 3T3-L1 cells at 0, 24, 48, and 72 h during differentiation. d Western blot analysis of differentiation- and function-related proteins in each group of 3T3-L1 cells undergoing differentiation for 72 h. e Relative mRNA expression in each group of genes related to differentiation and lipid related function, at 0, 12, 24, 48, and 72 h during differentiation. NC, 3T3-L1 cells; AD, induced differentiated 3T3-L1 cells; AD-MSC, induced differentiated 3T3-L1 cells co-cultured with hucMSCs. Data represent mean ± SD (n = 3 independent experiments); *p < 0.05, **p < 0.01, and ***p < 0.001 indicate significant differences (one-way ANOVA, LSD-t test)

HucMSCs might inhibit differentiation of 3T3-L1 preadipocytes through interference with the JAK2/STAT3 pathway

To gain a deeper understanding of the mechanisms underlying hucMSC-mediated inhibition of preadipocyte differentiation, we used RNA sequencing to assess gene expression in 3T3-L1 cells following 72 h of differentiation. Transcriptomic analysis revealed that hucMSC intervention led to significant differential expression of 305 genes between the groups, with 207 genes upregulated and 98 genes downregulated (Fig. 4a–b). DEGs were predominantly enriched in transcriptional regulation, cell signaling, cell proliferation, and immune response pathways (Fig. 4d). Pathway analysis showed the enrichment of DEGs in TNF, JAK/STAT, IL-17, PI3K/Akt, Toll-like receptor, and Apelin signaling pathways (Fig. 4c). Based on these findings, we further validated the protein level of the JAK2/STAT3 pathway, which is closely associated with adipogenesis. The results confirmed that hucMSCs indeed suppress the activation of the JAK2/STAT3 pathway during 3T3-L1 cell differentiation (Fig. 4e).

Fig. 4.

RNA-Seq of 3T3-L1 cells differentiated under conditions with or without co-culture with hucMSCs. a–b Heatmap and volcano plots of differentially expressed genes between 3T3-L1 cells differentiated for 72 h with and without co-culturing with hucMSCs (n = 3 independent experiments per condition). Log fold change (FC) > 2 or < 0.5; p < 0.05. c–d Gene ontology (GO) enrichment analysis and Kyoto Encyclopedia of Genes and Genome (KEGG) enrichment analysis of DEGs (|log2FC|> 1) represented in 3T3-L1 cells after 72 h of differentiation, co-cultured with hucMSCs versus without co-culturing with hucMSCs; p < 0.05. e Western blot of JAK2/STAT3 pathway key proteins in each group of 3T3-L1 cells differentiated for 72 h (n = 3 independent experiments). AD, induced differentiated 3T3-L1 cells; AD-MSC, induced differentiated 3T3-L1 cells co-cultured with hucMSCs. Data represent mean ± SD (n = 3 independent experiments); *p < 0.05, **p < 0.01, and ***p < 0.001 indicate significant differences (one-way ANOVA, LSD-t test)

HucMSCs attenuate adverse effects of adipocytes on hematopoietic cells

We further explored the impact of the aforementioned effects on mouse BM hematopoietic cells in vitro. 3T3-L1 cells were differentiated into mature adipocytes with prominent lipid droplets following a 2-week induction period, after which they were co-cultured indirectly with hucMSCs for 72 h. Subsequently, the hucMSCs and culture medium were removed, and a co-culture was established with mouse BM cells for an additional 48 h. Our results demonstrate that hucMSCs effectively suppress the apoptotic induction of BM cells by adipocytes (Fig. 5a). Adipocytes resulted in a slight augmentation of BM cells, an effect that was mitigated by the intervention with hucMSCs (Fig. 5b). Assessment of primitive cells within the mouse BM cell population indicates that hucMSCs partially counteract the decline in LSK cells caused by adipocytes (Fig. 5c). Furthermore, adipocytes elicit an elevation in the proportions of Mac-1 + Gr-1 + , CD41 + , and CD3 + cells, while concurrently decreasing the ratio of Ter-119 + and B220 + cells. Notably, under the influence of hucMSCs, these aberrant differentiation tendencies are significantly ameliorated (Fig. 5d-f). These findings suggest that adipocytes exert deleterious effects on hematopoiesis, with hucMSCs exerting a protective role against these effects.

Fig. 5.

HucMSCs attenuate adverse effects of adipocytes on mouse hematopoietic cells. a–c Apoptosis profile, cell count, and ratio of primitive cells in each group of mouse bone marrow cells after co-culture with adipocytes for 48 h. d–f Mac-1 + Gr-1 + cell, Ter-119 + cell, CD41 + cell, B220 + cell, and CD3 + cell ratio in each group of mouse bone marrow cells after co-culture with adipocytes for 48 h. BMNC, mouse BM cells; BMAD, mouse BM cells co-cultured with adipocytes; BMAD-MSC, mouse BM cells co-cultured with adipocytes influenced by hucMSCs. Data represent mean ± SD (n = 3 independent experiments); *p < 0.05, **p < 0.01, and ***p < 0.001 indicate significant differences (one-way ANOVA, LSD-t test)

HucMSCs attenuate adverse effects of adipocytes on hematopoietic cells affected by chemotherapeutic agents

Then, we proceeded to examine whether hucMSCs could mitigate the effects of adipocytes on mouse BM hematopoietic cells affected by chemotherapeutic agents. Adipocytes derived from 3T3-L1 cells were indirectly pre-cultured with hucMSCs for 72 h. Subsequently, the hucMSCs and culture medium were removed, and mouse BM cells treated with CTX and BUS for 24 h were co-cultured with the adipocytes for an additional 48 h. BM cells exhibited a significant reduction in viability under the influence of chemotherapeutic agents. BM cells, after chemotherapy drug treatment, show a pattern of changes similar to those seen in normal BM cells when exposed to adipocyte influence. Furthermore, pre-treating adipocytes with hucMSCs also provided a protective effect (Fig. 6a–f).

Fig. 6.

HucMSCs attenuate adverse effects of adipocytes on hematopoietic cells affected by chemotherapeutic agents. a–c Apoptosis profile, cell count, and ratio of primitive cells in each group of mouse bone marrow cells that had been affected by chemotherapeutic agents after co-culture with mature adipocytes for 48 h. d–f Mac-1 + Gr-1 + cell, Ter-119 + cell, CD41 + cell, B220 + cell, and CD3 + cell ratio in each group of mouse bone marrow that had been affected by chemotherapeutic agents cells after co-culture with mature adipocytes for 48 h. BMNC, mouse BM cells; BMAD, mouse BM cells co-cultured with adipocytes; BMAD-MSC, mouse BM cells co-cultured with adipocytes influenced by hucMSCs. Data represent mean ± SD (n = 3 independent experiments); *p < 0.05, **p < 0.01, and ***p < 0.001 indicate significant differences (one-way ANOVA, LSD-t test)

HucMSCs ameliorate oxidative stress induced in BM cells by adipocytes

Considering the potential link between BMAT and oxidative stress, we proceeded to assess ROS levels in mouse BM cells within the aforementioned adipocytes co-culture system. Our results show that adipocytes induce high levels of ROS accumulation in mouse BM cells, and the induction of ROS accumulation is attenuated significantly under hucMSC intervention (Fig. 7a). To further probe the underlying mechanisms, we investigated the antioxidant system within BM cells. We found that adipocytes cause damage to the antioxidant system in mouse BM cells, while hucMSC intervention confers protection to this system (Fig. 7b–c). In summary, hucMSCs afford protection to mouse BM cells against oxidative stress damage induced by adipocytes.

Fig. 7.

The impact of HucMSCs on the antioxidant system in bone marrow cells induced by adipocytes. a ROS levels in each group of mouse bone marrow cells after co-culture with mature adipocytes for 48 h. b–c Relative mRNA and protein expression of antioxidant-related genes in each group of mouse bone marrow cells after co-culture with mature adipocytes for 48 h. BMNC, mouse BM cells; BMMAD, mouse BM cells co-cultured with mature adipocytes; BMMAD-MSC, mouse BM cells co-cultured with mature adipocytes that had been pre-cultured with hucMSCs for 72 h. Data represent mean ± SD (n = 3 independent experiments); *p < 0.05, **p < 0.01, and ***p < 0.001 indicate significant differences (one-way ANOVA, LSD-t test)

Discussion

In the treatment of malignancies, chemotherapy persists as a conventional yet efficacious approach. However, one of the primary adverse effects associated with chemotherapy is impairment of hematopoietic function [30]. Dysfunction in the hematopoietic system poses significant challenges for patients, affecting not only the course of treatment and prognosis but also entailing substantial risks. In such instances, colony-stimulating factors (CSFs) are commonly used in clinical intervention to accelerate hematopoietic recovery; nonetheless, a subset of patients exhibit limited responsiveness to these approaches [31]. Chemotherapeutic agents, while inflicting damage to hematopoietic cells, concurrently deteriorate the BMM [4]. This may represent a pivotal reason for the reduced responsiveness of the hematopoietic system to CSFs following chemotherapy. Consequently, there is a pressing need to holistically reconsider this hematopoietic impairment and develop more comprehensive and efficacious treatment strategies based on emerging insights. MSCs, as vital constituents of the BMM, have attained substantial recognition for their contributions to hematopoiesis. Although exhibiting lesser sensitivity to chemotherapeutic agents compared to HSCs, MSCs nonetheless display tendencies toward senescence and adipogenic transformation under the influence of these drugs, severely compromising their normal functioning [32]. Therefore, the use of exogenous healthy MSCs constitutes a hopeful strategy in the treatment of hematopoietic impairment. In alignment with the therapeutic effects of hucMSCs observed in hematopoietic failure stemming from different etiologies, our results demonstrate that hucMSCs similarly promote hematopoietic recovery in mice subjected to chemotherapy-induced hematopoietic impairment. Although a substantial portion of intravenously administered hucMSCs is primarily concentrated in the lungs of mice with rarely homing to the bone marrow, several studies have confirmed their significantly therapeutic efficacy in the treatment of hematological disorders nonetheless [33, 34]. A wide proportion of this positive influence is likely exerted through cytokine secretion and exosome-mediated pathways acting upon the BMM via systemic circulation.

As a pivotal constituent of the BM microenvironment, BMAT assumes a multifaceted but contested role in hematopoiesis, conferring it with enhanced research significance. Some researchers are exploring the potential benefits of modulating BMAT to enhance hematopoiesis [18]. Regardless of the inciting cause, BMF cases often present with the pathological feature of BM adiposity [8], which could serve as another crucial factor contributing to the decline in hematopoietic function. BM adiposity may therefore represent a potential new therapeutic target in the treatment of hematopoietic impairment. In our study, alongside promoting hematopoietic recovery in mice with chemotherapy-induced hematopoietic impairment, hucMSCs were also observed to suppress the expansion of BMAT. This inhibitory effect on BMAT expansion may contribute to facilitating hematopoietic recovery. HucMSCs have the potential to differentiate into adipocytes. However, specific protective mechanisms usually keep them in an undifferentiated state [35]. These protective mechanisms might grant MSCs the ability to modulate adipose tissue, including BMAT. Our data support this hypothesis, indicating that under in vitro conditions, hucMSCs exert significant impacts on both preadipocytes and mature adipocytes. To explore the regulatory influence of hucMSCs on preadipocyte differentiation, we performed high-throughput RNA sequencing (RNA-seq) and comprehensive bioinformatic analyses. The data revealed that DEGs were significantly enriched in transcriptional regulation, cell signaling, and cell proliferation. Among the notably suppressed signaling pathways, downregulation of the TNF pathway is not considered to interfere with preadipocyte differentiation [36], whereas suppression of the JAK/STAT pathway may potentially play a role in this process. During the preadipocyte differentiation process, there is a marked upregulation of several proteins within the STAT family, which is essential for the expression of key transcriptional regulators involved in adipogenesis [37]. Among these, the JAK2/STAT3 pathway is considered to be one of the earliest activated pathways in the adipogenic cascade, additionally modulating the activity of C/EBPβ to regulate subsequent stages of differentiation [38]. Targeted inhibition of either JAK2 or STAT3 effectively inhibits preadipocyte maturation [39]. This mechanism may play a crucial role in inhibiting the excessive expansion of BMAT and reducing the production of pro-inflammatory adipokines, especially leptin, thereby providing protection for hematopoiesis. Leptin, primarily secreted by adipocytes, can be feedback-regulated via the JAK2/STAT3 pathway [40]. While it acts on the central nervous system to modulate energy metabolism, it is considered a potent pro-inflammatory factor peripherally and can induce oxidative stress injury [41]. Studies have shown that leptin not only significantly increases T-cell proliferation and activation in AA [42] but is also closely associated with the development of chronic GVHD following allogeneic hematopoietic stem cell transplantation [43].

Scholars refer to the detrimental effects of excessive lipid deposition in non-adipose tissues as lipotoxicity, which commonly presents as adipocyte-induced local inflammation and supplementary damage due to elevated free fatty acid concentrations [44]. An enlarged BMAT not only weakens its brown adipocyte characteristics and becomes more akin to peripheral white adipose tissue [45], but also releases excessive pro-inflammatory adipokines, instigating chronic inflammation within the BM milieu [16], thereby triggering oxidative stress responses in HSCs. Notably, when cells are exposed to an overabundance of saturated long-chain fatty acids, their fatty acid metabolism veers toward harmful pathways, resulting in the production of noxious compounds such as ceramides and cholesterol esters [46] that enhance ROS formation [47]. As widely acknowledged, ROS, being inherent byproducts of metabolic processes, can lead to mitochondrial and nuclear DNA damage upon excessive intracellular buildup, ultimately precipitating apoptotic cell death. Elevated ROS levels, specifically in HSCs, can disrupt their quiescent state [48], giving rise to a senescent phenotype typified by increased myeloid differentiation and loss of cellular polarity [49]. Our in vitro experimentation revealed that regardless of whether they had undergone chemotherapeutic drug-induced injury or not, mouse BM cells co-cultured with adipocytes consistently demonstrated increased apoptosis, decreased stem cell proportions, and abnormal myeloid differentiation phenotypic traits that strikingly resemble ROS-mediated hematopoietic cell injury. Although a subtle increase in BM cell counts is noted under the impact of adipocytes, this observation is likely attributable to ROS mediating the disruption of the quiescent state of long-term hematopoietic stem cells (LT-HSCs) and precipitating their aberrant differentiation [48]. The escalated ROS levels within BM cells serve as compelling evidence supporting the assertion that adipocytes exert detrimental effects on hematopoiesis by provoking oxidative stress responses. Curiously, while there is a significant elevation of ROS levels in BM cells under the influence of adipocytes, the Nrf2-regulated antioxidant machinery demonstrates a reduced level of activity relative to normative conditions. Hence, it is reasonable to infer that the increase in ROS levels may partly be attributed to adipocyte-induced impairment of the antioxidant system in BM cells. Research has indicated that Kelch-like ECH-associated protein 1 (KEAP1) functions as a principal modulator of the Nrf2 antioxidant pathway, particularly affecting Nrf2 activity [50]. Elevated concentrations of free fatty acids might detrimentally affect the KEAP1/Nrf2 system, compromising its function [51]. When hucMSCs interfere with adipocytes, BM cells demonstrate restoration of antioxidant capacity accompanied by a reduction in ROS levels and, to a certain degree, amelioration of the impairment of HSCs and their anomalous differentiation patterns. This finding suggests that hucMSCs, in addition to their direct antioxidative effects, can also improve hematopoiesis through adipocyte-related indirect antioxidative mechanisms.

Conclusion

In summary, this study represents the first to posit that hucMSCs may enhance hematopoietic function via modulating BMAT. Through in vivo experiments, we observe an inhibitory effect of hucMSCs on BMAT excessive expansion, while in vitro studies delineate their regulatory function on both preadipocytes and mature adipocytes. Concurrently, these regulatory effects were substantiated to confer protective benefits on the hematopoiesis. These discoveries present novel theoretical underpinnings for the application of hucMSCs in chemotherapy-induced hematopoietic impairment therapy and offer fresh insights for the investigation of other hematological disorders related to BMAT.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contributions

Jingyi Feng, Huanying Ren, Sicheng Bian and Hongwei Wang conceptualized and designed the research. Jingyi Feng, Miao Zhang, Yan Ren, Zhuanghui Hao, Jing Xu, Yaofang Zhang, Jianmei Chang and Jiangxia Cui performed the experiments and analyzed the data. Jingyi Feng and Huanying Ren wrote the manuscript. Shuo Li and Muteb Muyey Daniel wrote part of the experimental procedures and results sections of the manuscript. Zhifang Xu, Yanhong Tan and Xiuhua Chen reviewed and edited the manuscript. Hongwei Wang and Fanggang Ren were responsible for acquisition of funding, reviewing, and editing the manuscript.

Funding

This work was supported by the Science and Technology Department of Shanxi Province (Grant No. YDZJSX2021B009); Health Commission of Shanxi Province (Grant No. 2021XM07).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Conflict of interests

The authors declare no competing interests.

Ethical approval

The study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of the Second Hospital of Shanxi Medical University (No. DW2023003, 20 February 2023).

Consent to participate

Not applicable.

Consent to publish

Not applicable.