Bone marrow samples were originated from the First Hospital of Chongqing Medical University (Chongqing, China; collection lasted from 1/1/2024 to 1/6/2024). Volunteers (of either sex) without hematologic malignant diseases who had only symptoms of iron-deficiency anemia (IDA) were eligible for collection. The bone marrow samples were then categorized by age, designating volunteers aged ≤25 years as the young group (Y group), those aged 45-55 years as the middle-aged group (M group) and those aged ≥65 years as the elderly group (O group). Sex and age distribution of patients is included in Table I. The present study was approved (approval no. 2023092) by The Ethics Committee of Chongqing Medical University (Chongqing, China). Written informed consent was obtained from all participants prior to publication of the present study.

The posterior superior iliac spine was selected to be the site for bone marrow puncture. Specifically, 0.2 ml bone marrow fluid was extracted for the preparation of bone marrow smears, followed by Wright's staining (cat. no. S0217; BIOSS) to observe the morphological characteristics of bone marrow cells. A total of 3 ml bone marrow was collected from each volunteer for bone marrow culture.

Qualified bone marrow smears were prepared using Wright's staining technique. In total, 0.5-0.8 ml Wright's Giemsa A solution was added to the smear, ensuring that the staining solution covers the entire specimen for 1 min. Subsequently, Wright's Giemsa B solution was added to the A solution, with a volume that is 2-3X that of the A solution to facilitate thorough mixing, followed by a staining period of 3-10 min. After washing with water and allowing to dry, the proportions and counts of various hematopoietic cells in the bone marrow were observed under a light microscope (Olympus CKX41; Olympus Corporation) to assess whether the bone marrow sample exhibits signs of hematological disorders or malignant lesions.

The supernatant from the bone marrow samples and the co-culture system were isolated before the capture antibody solution was added, which took place in a 96-well plate (cat. no. FCP962; Beyotime Institute of Biotechnology) and was incubated overnight at 4°C. Subsequently, the plate was washed five times with a washing solution to ensure drying. Plasma was then diluted with the sample diluent to concentrations of 1:5, 1:10, 1:50, 1:100 and 1:2,000 to ascertain the optimal concentration. Finally, the concentrations of stem cell factor (SCF; cat. no. QZ-10611), granulocyte macrophage (GM) colony stimulating factor (CSF; cat. no. QZ-10908), IL-3 (cat. no. QZ-10502) and erythropoietin (EPO; cat. no. QZ-10444; all from Quanzhou Jiubang Biotechnology Co., Ltd.) were measured through antigen-antibody interactions, before the absorbance (450 nm) of the cytokines was quantified using a microplate reader (Rayto RT-6100; Rayto Life and Analytical Sciences Co., Ltd.) in conjunction with a standard curve.

Human bone marrow mononuclear cells (MNCs) were isolated through red blood cell lysis (cat. no. BL503A; Biosharp Life Sciences), lymphocyte separation (cat. no. LTS1077; Tianjin Haoyang Biological Products Technology Co., Ltd.) and density gradient centrifugation (room temperature; 5 min; 16,020 × g). The isolated MNCs were then counted and evenly distributed into 60-mm culture dishes for in vitro culture. During the initial stages of cell culture, the medium DMEM F12 (Gibco; Thermo Fisher Scientific, Inc.) supplemented 10% fetal bovine serum (FBS; cat. no. FSP500; Shanghai ExCell Biology, Inc.) was partially replaced every 3 days until adherent-dependent cells emerged. Subsequently, a complete medium exchange and cell passaging procedures were performed to separate and purify the adherent-dependent cells. Furthermore, during the early phases of the in vitro culture of bone marrow MNCs, the supernatant cells were collected for additional in vitro culture. The supernatant cells were treated with a magnetic bead sorting kits (cat. nos. 130-092-263 and 130-100-453; Miltenyi Biotec GmbH), and these cells were successively treated with CD38- and CD34-labeled magnetic beads, respectively; and CD34-positive and CD38-negative HSCs were collected for co-culture operation under the action of a magnetic field by the negative and positive sorting methods, respectively.

After counting (1×10⁶), adherent-dependent cells and suspended cells were allocated to Eppendorf tubes separately and treated with 5% BSA (cat. no. SW3015; Beijing Solarbio Science & Technology Co., Ltd.) blocking buffer under light conditions for 30 min. The adherent cells were stained with specific surface marker antibodies (dilution ratio 1:100) at 4°C for 30 min in the dark, including CD73 (cat. no. 561258), CD90 (cat. no. 555595), CD105 (cat. no. 561443), CD14 (cat. no. 557153), CD34 (cat. no. 555821), CD45 (cat. no. 555482) and human leukocyte antigen (HLA-DR; cat. no. 555811), whilst the suspended cells were stained with the appropriate antibodies against CD34 (cat. no. 555821) and CD38 (cat. no. 567147) (all from BD Biosciences) before identification using flow cytometry. Following washing with PBS (cat. no. G4250-500ML; Wuhan Servicebio Technology Co., Ltd.), the isolated and purified adherent cell pellet was collected, resuspended in 100 μl PBS and mixed gently with 900 μl of pre-chilled 75% ethanol to ensure complete homogenization for subsequent cell cycle assessment. After counting, these cells were also used for apoptosis measurement and subsequently resuspended in 500 μl PBS. Finally, the results of the cell cycle and apoptosis analyses were obtained through flow cytometry (BD FACSAria™ Fusion; BD FACSDiva™ Software; BD Biosciences).

The present study used multiple co-culture systems. The first system utilized bone marrow plasma from groups Y and O to intervene in the proliferation of bone marrow MSCs across the respective groups. The second system leveraged the suspended proliferative characteristics of HSCs and the adherent proliferative characteristics of bone marrow MSCs. In this system, Y group HSCs were selected as the experimental subjects to investigate their hematopoietic activity, with bone marrow MSCs from each group serving as intervention factors to establish an in vitro hematopoietic co-culture system. The third system involved co-culturing the bone marrow MSCs from group Y with the differentiated adherent bone marrow cells from group O. Subsequently, the pre-treated hematopoietic cells were collected for in vitro hematopoiesis assays and assessment of colony forming unit (CFU)-erythroid (E), burst forming unit (BFU)-E, CFU-GM and CFU-granulocytes/erythroids/macrophages/megakaryocytes (GEMM) formation.

Y group hematopoietic stem progenitor cells were collected from the hematopoietic co-culture system, added to MethoCult^® medium (cat. no. 04034; STEMCELL Technologies, Inc.), shaken well and then plated into culture dishes for incubation at 37°C with 5% CO₂ for 14 days. The types and counts of colonies were assessed using an inverted microscope (Olympus CKX41; Olympus Corporation). (A single CFU-E is formed by 8-200 erythrocytes; a single BFU-E is formed by >200 erythrocytes; a single CFU-GM is formed by at least 20 granulocytes and macrophages; a single CFU-GEMM is formed by >500 cells containing erythrocytes, granulocytes and macrophages).

Following intervention with bone marrow plasma, equal amounts of Y group bone marrow MSCs were seeded into culture dishes to restore normal proliferation. An equal volume of pre-warmed EdU (cat. no. C0078S; Beyotime Institute of Biotechnology) working solution (final concentration of EdU at 10 μM) and culture medium was then added to the dishes, followed by an additional 5-h incubation at 37°C with 5% CO₂. After EdU labeling was completed, cells were treated with fixation and permeabilization solutions for 15 min, incubated with Click reaction solution in the dark at room temperature for 30 min and then stained for nuclei (cat. no. C1005; Beyotime Institute of Biotechnology). Finally, proliferation was analyzed using fluorescence detection.

After intervention with bone marrow plasma, equal amounts of M group bone marrow MSCs were seeded into culture dishes to restore normal proliferation. The cells were fixed with 4% paraformaldehyde (room temperature; 15 min) and permeabilized with Triton X-100 (room temperature; 15 min). Subsequently, they were incubated with antibodies for senescence-associated proteins p53 (cat. no. BF8013; Affinity Biosciences) (dilution ratio 1:500) and p21 (cat. no. AF6290; Affinity Biosciences) (dilution ratio 1:500) at room temperature for 1 h, followed by blocking (cat. no. P0260; Beyotime Institute of Biotechnology) (room temperature; 10 min) and incubation with secondary antibodies (DyLight 488, Goat Anti-Mouse IgG; cat. no. A23210; and Dylight 549, Goat Anti-Rabbit IgG; cat. no. A23320; both from Abbkine Scientific Co., Ltd.) (dilution ratio 1:100) at room temperature for 1 h in the dark to form antigen-antibody complexes. Fluorescence was then observed and quantified using a confocal microscope (AX/AX R with NSPARC; Nikon Corporation).

Equal amounts of O group bone marrow MSCs, following intervention with bone marrow plasma, were seeded into culture dishes to restore normal proliferation. Upon reaching 70% confluence, the culture medium was removed, before 2 ml OriCell^® Human BMMSC Osteogenic Differentiation Medium (cat. no. HUXMA-90021; Cyagen Biosciences, Inc.) was added into each dish, with the medium replaced every 3 days. After 2-4 weeks of induction, alkaline phosphatase staining (cat. no. G1481; Beijing Solarbio Science & Technology Co., Ltd.) was performed to assess osteogenic differentiation. Similarly, when the bone marrow MSCs attained 70% confluence, the upper culture medium was removed and 2 ml OriCell^® Human BMMSC Adipogenic Differentiation Medium A (cat. no. HUXMA-90031; Cyagen Biosciences, Inc.) was added to each dish. Following 3 days of induction, the medium was aspirated and 2 ml OriCell^® Human BMMSC Adipogenic Differentiation Medium B (cat. no. HUXMA-90031; Cyagen Biosciences, Inc.) was introduced. After 1 day, medium B was discarded and medium A was reinstated. The two media were thereafter alternated in this manner whilst regularly monitoring cell status. Finally, Oil Red O staining (cat. no. G1262; Beijing Solarbio Science & Technology Co., Ltd.) was used to evaluate the adipogenic differentiation of the cells.

Bone marrow MSCs were seeded into 96-well plates, with 100 μl (containing 2,000 cells) added to each well. Following this, 10 μl CCK-8 (cat. no. C0038; Beyotime Institute of Biotechology) solution was added to each well and the cells were incubated in a culture incubator (cat. no. BB150-2TCS-L; Thermo Fisher Scientific, Inc.) at 37°C with 5% CO₂ for 6 h. Absorbance was measured at 450 nm and cell viability was analyzed based on the results.

Bone marrow MSCs were digested with trypsin (cat. no. 25200072; Gibco; Thermo Fisher Scientific, Inc.), counted and the cell pellet (≥10⁶) was collected. Next, 3% paraformaldehyde fixative was gently added along the wall of the tube for at least 2 h at room temperature. Cell precipitates were stained (room temperature; 3 h) with heavy metal salts and embedded (room temperature; 1 h) in epoxy resin. After making ultrathin sections, (70-100 nm) they were used for observation. The prepared samples were subsequently sent to the Electron Microscopy Laboratory of Chongqing Medical University for observation.

Bone marrow MSCs were cultured to 70% confluency and then the upper layer of medium was removed. Following a wash with PBS, β-galactosidase fixation solution was added and the cells were fixed at room temperature for 15 min. After fixation, PBS was used to wash away the fixative and the cells were incubated overnight at 37°C with β-galactosidase staining solution (cat. no. C0602; Beyotime Institute of Biotechnology). For suspended hematopoietic stem progenitor cells, these cells were collected by centrifugation and stained using the same method. The stained cells were then placed on a glass slide for observation under an optical microscope.

The isolated and purified Y group and O group bone marrow MSCs were sent to the Beijing Genomics Institute for RNA-Seq (BGI Genomics Co., Ltd.). The sequences and expression information of the transcripts from both groups were analyzed (https://biosys.bgi.com/#/report/login) using Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Ontology (GO) enrichment methods. The sequencing results were processed using the STRING database (https://cn.string-db.org/) and Cytoscape software (Cytoscape 3.7.2; https://cytoscape.org/download.html) to construct a protein-protein interaction (PPI) analysis of differentially expressed genes.

Under enzyme-free conditions, RNA was extracted from human bone marrow MSCs using an RNA extraction kit (cat. no. R0018M; Beyotime Institute of Biotechnology). Primer sequences for the genes to be validated are included in Table SI. Reverse transcription was performed using RT kit (cat. no. RR037A; Takara Bio, Inc.) according to the manufacturer's protocol, followed by qPCR using SYBR Green qPCR Master Mix (cat. no. 639676; Takara Bio, Inc.) on the circulator real-time detection system (Bio-Rad). Actin was used as the reference gene. (Thermal cycle conditions: the first step of pre-denaturation 95°C for 30 sec, the second step 95°C for 10 sec and 60°C for 30 sec, cycle 40 times). The quantification method was the ΔΔCq method (28).

All experiments were performed in triplicate at a minimum. Statistical analysis was conducted using GraphPad Prism 9.5 (Dotmatics). Measurement data were expressed as the mean ± SD. Comparisons between two groups of data that met the criteria of normal distribution and homogeneity of variance were performed using an unpaired t-test. For comparisons among multiple groups, one-way analysis of variance was employed, followed by Tukey's post hoc test. For comparisons of data between groups at different time points, repeated measures ANOVA was utilized, with Bonferroni correction for post hoc analysis. Pearson correlation analysis was used to assess the relationships among various indicators. P<0.05 was considered to indicate a statistically significant difference.

For the present study examining the causes of diminished bone marrow function in the elderly, 184 anemia samples were collected using random sampling methodology (Table SII). Bone marrow smears indicated that the proportions of myeloid, erythroid and lymphoid lineages, in addition to the counts of megakaryocytes, were normal across all samples, suggesting the absence of hematological malignancies (Fig. 1A-C). Statistical analysis revealed that the O group comprised the highest proportion of samples, suggesting a significant degradation of hematopoietic function in the elderly. Furthermore, there were no sex differences among the groups, indicating that the decline in hematopoietic function did not vary between elderly male and female population (Table I).

The lipid composition in bone marrow samples exhibited differential levels based on age, with the O group showing the highest lipid content (Fig. 1D and F). MNCs were obtained from the bone marrow samples using density gradient centrifugation, where the data indicated a gradual decline in the number of MNCs as age increased (Fig. 1E and G). Assessments of hematopoietic-related factors in bone marrow plasma showed reduced levels of SCF, GM-CSF, IL-3 and EPO in the O group (Fig. 1H). In vitro cultured, magnetically-sorted bone marrow hematopoietic stem progenitor cells (Fig. 1I and J) were subsequently analyzed for the expression of surface markers using flow cytometry, revealing that CD34 was positive and CD38 was negative, consistent with the definition of bone marrow HSCs (Fig. 2A). Statistical analysis indicated that the content of HSCs sourced from the O group was the lowest, whilst that from the Y group was the highest (Fig. 1K). In the early stages of the in vitro culture of bone marrow MNCs (Fig. 1L), the upper culture medium of the O group exhibited the highest lipid production (Fig. 1M), although this phenomenon was not observed in subsequent culture and passaging stages. On day 6 of in vitro culture, bone marrow MSCs displayed trends of adherent aggregative proliferation under microscopy. Flow cytometry for the expression of surface markers demonstrated that CD73, CD90 and CD105 were positive, whilst CD14, CD34, CD45 and HLA-DR2 were negative, aligning with the definition of bone marrow MSCs (Fig. 2B). Compared with the other groups, the O group exhibited the smallest clustered area of bone marrow MSCs and the fewest number of cells forming clusters (Fig. 1N and O). These results suggested that the decline in bone marrow function among the elderly population is attributed to a reduction in the number of stem cells and a deterioration of their functional capacity, which is closely associated with the bone marrow microenvironment. However, the causal relationship remains to be validated.

To further investigate the effects of alterations in the bone marrow microenvironment on bone marrow MSCs, the in vitro proliferation of MSCs derived from bone marrow plasma samples of the Y and O groups across age gradients was next assessed (Fig. 3A). Control group results (no intervention) exhibited distinctively dense contact phenomena in MSCs during extensive proliferation (Fig. 3B). Microscopic observations revealed that MSCs in the experimental Y group exhibited a regular and organized growth pattern, forming membrane ridges at points of contact (Fig. 3C). Under identical culture conditions, MSCs in the experimental O group displayed fused growth characteristics, with unclear cell boundaries and varied morphologies (Fig. 3D). EdU assays demonstrated a reduced proliferative capacity in Y group MSCs in the O group compared with that in the control group (Fig. 3E and F). Immunofluorescence results indicated that Y group bone marrow plasma could decrease the expression of senescence-associated proteins p53 and p21 in M group MSCs, whereas M group MSCs exhibited increased levels of p53 and p21 following intervention with O group bone marrow plasma (Fig. 3G and H). Regarding the multidirectional differentiation capacity of bone marrow MSCs, a complementary trend toward osteogenic and adipogenic differentiation was observed (29). Differentiation results suggested that, compared with those in the control group, Y group bone marrow plasma enhanced the osteogenic potential of O group MSCs whilst diminishing their adipogenic capacity (Fig. 3I and J). During the in vitro culture phase, Y group bone marrow plasma significantly increased the viability of MSCs across all groups, whilst O group plasma led to decreased cell viability in all groups (Fig. 3K). Electron microscopy results revealed that MSCs from the experimental Y group contained abundant intracellular materials and active organelles, most notably the endoplasmic reticulum. By contrast, MSCs from the experimental O group exhibited irregular membrane folds, enlarged cell bodies, increased translucent intracellular materials and significant cell differentiation (Fig. 3L). These findings suggested that bone marrow MSCs are influenced by their growth environment, where modulating such bone marrow microenvironment will likely effectively alter the stemness of these cells (30,31).

To investigate the direct relationship between age-related changes in bone marrow stem cells and hematopoiesis in the elderly population, bone marrow MSCs from various age groups were co-cultured with HSCs from the Y group using in vitro modeling (Fig. 4A). The MSCs co-cultured for this experiment exhibited robust proliferative capabilities in vitro (Fig. 4B). Statistical analysis indicated that during the early stages of proliferation in vitro, Y group MSCs displayed the highest proliferation capacity, whilst there was no statistically significant difference in cell numbers among the groups as culture time extended (Fig. 4E). Cell viability assessments revealed that primary cultured MSCs maintained robust cell viability at all stages (Fig. 4F). The proliferation of stem cells is closely associated with cell contact, where prolonged in vitro culture time may reduce the differences in stemness among the stem cells (32-35). Microscopic observations indicated that MSCs predominantly exhibited a spindle shape, where after extensive proliferation, a fusion growth phenomenon was observed, particularly pronounced in the O group (Fig. 4C). β-galactosidase staining for cellular aging indicated a significant presence of senescence-positive MSCs at areas of severe cell fusion growth (Fig. 4D and G). The levels of hematopoietic-related factors in different co-culture system supernatants were next measured and it was found that O group MSCs, as intervention factors, exhibited the lowest expression of SCF, GM-CSF and IL-3 when co-cultured with Y group HSCs (Fig. 4H). Compared with that in other systems, co-culturing with O group MSCs resulted in the significant aging of Y group HSCs (Fig. 4I), as evidenced by the increase in the proportion of senescence-positive HSCs (Fig. 4J). In human bone marrow colony formation assays, CFU-E and BFU-E colonies formed earliest, where statistical analysis of colony numbers indicated age-related differences among the experimental groups, with the O group exhibiting the poorest colony formation capability. In terms of CFU-GM and CFU-GEMM formation, the number of colonies and the density of cells within colonies differed significantly. Y group HSCs co-cultured with Y group MSCs formed a greater number of colonies with higher cell density, whilst Y group HSCs co-cultured with O group MSCs produced the fewest colonies with the lowest cell density (Fig. 4K). These results suggested that the state of MSCs coexisting with HSCs in the bone marrow will likely affect the hematopoietic capacity of HSCs (Fig. 5).

Bone marrow MSCs serve an important role in bone marrow function. To investigate the growth characteristics of MSCs, their morphological features were analyzed at various stages of in vitro culture. During the early primary culture of Y group MSCs, cells exhibited a spindle shape, characterized by relatively large nucleoplasm and a small number of protruding branches at the ends. As age increased or as the in vitro culture progressed, the intracellular content of MSCs became more abundant, where the nucleoplasm decreased and the spindle characteristics gradually diminished, resulting in a transformation to polygonal or irregular shapes until differentiation into terminal cells. This morphology was commonly observed in the long-term in vitro culture of O group MSCs (Fig. 6A and C). Culturing results indicated that the various morphologies of MSCs coexisted in the different age groups, each with specific proportions and patterns. Electron microscopy observations revealed that MSCs primarily existed in the following two states: A quiescent phase and an active phase. Compared with quiescent MSCs, those in the active phase exhibited increased organelle function, particularly in the endoplasmic reticulum, along with an increase in vesicular structures (Fig. 6B). Cell cycle analysis showed that MSCs predominantly remained in the G₁ phase during in vitro culture, with age-related differences based on the culture stage. In the primary culture (P0 stage) of Y group MSCs, the proportion of cells in the G₁ phase was the highest, whilst in the passaged culture (Pn, n≥3) of O group MSCs, the proportion of cells in the G₁ phase was the lowest (Fig. 7A and C). Apoptosis results indicated that as age increased and culture time extended, the proportion of apoptotic MSCs was also increased, with a particularly rapid rise in late-stage apoptosis (Fig. 7B and D). These findings suggested that the stemness of MSCs evolves in relation to differences in origin and culture stage, highlighting the importance of the in vitro culture phase as a critical factor for intervening in their stemness.

In the investigation of the key factors influencing the differences in stemness among bone marrow MSCs, early in the primary culture of the O group, certain bone marrow materials were found to exhibit substratum adherence, resembling the behavior of MSCs, primarily consisting of morphologically diverse bone marrow stromal cells, alongside the generation of highly refractive bone marrow terminal substances. Under these conditions, MSCs demonstrated reduced proliferation (Fig. 8A and B). Passage of these bone marrow materials showed no significant changes (Fig. 8C and D). The MSCs were then isolated from these materials for passage culture, where the results indicated that MSCs could effectively proliferate when the influence of the bone marrow materials was removed (Fig. 8E). When the efficiently proliferating Y group MSCs were co-cultured with these bone marrow materials, the morphology of the Y group MSCs transitioned from distinctly spindle-shaped to irregular polygonal forms, characterized by fusion growth (Fig. 8F). These results suggested that changes in the stemness of MSCs are influenced by their growth microenvironment.

To analyze this variability in MSC proliferation, Y group and O group MSCs, which were isolated and purified, were observed using both optical and electron microscopy. Y group MSCs primarily communicated through three types of contact: 'Synapse-to-synapse', 'synapse-to-nuclear zone' and 'nuclear zone-to-nuclear zone'. By contrast, O group MSCs exhibited more complex modes of contact due to increased synaptic branching and morphological diversity, with these three forms of cell contact being more pronounced among O group MSCs (Fig. 8G). Electron microscopy results indicated that Y group MSCs had smooth plasma membranes, uniform cell shapes and equal cell body sizes, whilst O group MSCs displayed irregular membrane protrusions, diverse cell morphologies and increased accumulation of translucent intracellular materials and glycogen (Fig. 8H).

In the transcriptomic analysis, KEGG enrichment analysis of differential genes between Y group and O group MSCs indicated significant differences in hematopoiesis and cell adhesion. Cluster heat maps revealed distinctions in genes related to hematopoiesis and cell adhesion between the two groups (Fig. 8I and J). GO_cellular component enrichment results suggested that the differences between Y group and O group MSCs are reflected in spatial changes of the plasma membrane and cell membrane transport systems (Fig. 9A). GO_molecular function enrichment results highlighted that the differences in membrane components predominantly manifested in the expression of cell recognition-related receptors and signaling molecules on membranes (Fig. 9B). GO_biological process enrichment results indicated that these differences are closely associated cell proliferation, adhesion, immune recognition and regulation of bone marrow MSCs (Fig. 9C). PPI analysis displayed the associations of the top 53 differential genes with the greatest variation (Fig. 9D). Finally, RT-qPCR validation results demonstrated substantial differences in cell growth and communication between Y group and O group MSCs (Fig. 9E), with differential validation also observed in genes associated with signaling recognition and immune response (Fig. 9F). These findings suggested that bone marrow MSCs exhibit varying functional stress responses to changes in the bone marrow microenvironment, ultimately adapting their stemness to meet the organism's demands.