Real-ambient PM_2.5 exposure disrupts hematopoietic homeostasis via HIF-1α-driven myeloid skewing and promotes organ inflammation

Abstract

Environmental pollutants like PM_2.5 threaten hematopoietic homeostasis, yet how real-world exposure disrupts blood cell production, especially locally in the lung and systemically in the bone marrow (BM), remains poorly understood. Previous studies often used artificial particles or lacked mechanistic insights into systemic effects. Hypoxia-inducible factor-1alpha (HIF-1α) is essential for hematopoietic stem cell (HSC) maintenance. Herein, we utilized a real-ambient PM_2.5 exposure system and conducted a detailed characterization of hematopoietic and downstream immune cell populations in mice with myeloid lineage-specific knockout of HIF-1α (mHIF-1α^−/−) and their wild-type littermate controls. Our findings demonstrate that real-ambient PM_2.5 exposure induces a HIF-1α-dependent myeloid-biased hematopoiesis within both the lung and BM. This bias results in an accumulation of mature myeloid cells, particularly neutrophils and macrophages, in peripheral organs such as the liver and spleen. Critically, this cellular redistribution precipitates inflammatory injury in a HIF-1α-dependent manner. These results provide novel insights into how environmental contaminants, exemplified by PM_2.5, perturb hematopoiesis, highlighting the critical role of HIF-1α in mediating lineage-specific hematopoietic responses and subsequent inflammatory sequelae.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12989-025-00654-5.

Introduction

Hematopoiesis, a dynamic process responsible for continuous formation of all blood cells, is vital for maintaining whole-body homeostasis by generating around 1 × 10¹³ cells into the circulation daily [1]. Under steady-state conditions, hematopoietic stem cells (HSCs) maintain a finely-tuned equilibrium characterized by self-renewal and multipotent differentiation potential [2]. However, this equilibrium is vulnerable to disruption by both intrinsic factors and extrinsic stimuli. Excessive environmental stress, such as pathogens [3], toxins [4], and pollutants [5], can adversely affect hematopoietic processes [6], leading to compromised cell formation and subsequent health hazards. Aberrant hematopoiesis is closely linked to a large number of adverse health outcomes, including atherosclerosis [7], acute kidney injury [8], liver fibrosis [9], and malignancy [10]. In these diseases, demand-adapted hematopoiesis tends to establish a new equilibrium, often resulting in a hyper-responsive state that triggers chronic inflammation and inflammatory diseases [11, 12]. Given the critical role of steady-state hematopoiesis in ensuring healthy status, it is imperative to define the in vivo characteristics of hematopoietic responses and the resultant adverse outcomes following environmental stress under real-life scenarios.

As a ubiquitous environmental stress, ambient fine particulate matter (PM_2.5) is regarded as one of the prominent risk factors for entire hematopoietic hierarchy, from HSCs to hematopoietic progenitors [13, 14], and has been implicated in inflammatory diseases [15]. The bone marrow (BM) is a primary hematopoietic organ, playing a crucial role in producing a diverse array of immune cells, encompassing T cells, B cells, neutrophils (NEs), dendritic cells (DCs), and myeloid-derived cells (immature myeloid cells and macrophages) [16]. Notably, the lung, a chief interface for inhaled contaminants [17], has been recognized as an extramedullary hematopoietic site with the potential for hematopoietic reconstitution [18]. Megakaryocyte progenitors (MkPs) resident in the lung are responsible for about 50% of the total platelet production in the body [18], suggesting a significant regulatory function in the pathogenesis of lung disorders. Our previous study demonstrated that acute PM_2.5 exposure via intratracheal instillation for three days was sufficient to promote myeloid-biased hematopoiesis in both the lung and BM, leading to systemic inflammation [19]. Nonetheless, the intratracheal instillation method fails to reflect real-life exposure scenarios, warranting further investigation.

Regarding the potential mechanisms of particle exposure on myeloid hematopoiesis, we found that nuclear factor erythroid 2-related factor 2 (NRF2) plays a pivotal role in this process [19]. However, due to its broad impacts across entire hematopoietic hierarchy, NRF2 cannot be definitively identified as the sole mechanism underlying PM_2.5-induced myeloid-biased hematopoiesis. Therefore, further research is needed to elucidate the potential mechanisms of myeloid hematopoietic imbalance stimulated by environmental pollutants. Of note, hypoxia-inducible factor-1alpha (HIF-1α), a heterodimeric transcription factor associated with cellular adaptive responses to hypoxia, is essential for maintaining HSCs quiescent within the hypoxic endosteal niche [20] and mobilizing myeloid progenitors (MPs) via granulocyte colony-stimulating factor (G-CSF) [21]. HIF-1α also plays a crucial role in mature myeloid cell-mediated inflammation through the regulation of glycolytic capacity [22]. The knockout of HIF-1α in HSCs has been reported to decrease infiltration of myeloid cells and improve cardiac function following myocardial infarction (MI) [23, 24]. However, this approach did not exclude the influence of non-myeloid hematopoietic lineages during disease progression. This highlights the need for a detailed investigation into the role of myeloid lineage-specific HIF-1α in PM_2.5-induced hematopoiesis, with the aim of identifying precise molecular mechanisms.

In the present study, we attempted to identify in vivo characteristics of hematopoiesis, elucidate the potential molecular mechanisms, and reveal the resultant adverse outcomes induced by real-ambient PM_2.5 exposure. For these aims, we employed a real-ambient PM_2.5 exposure system with applications of myeloid lineage-specific HIF-1α^−/− (mHIF-1α^−/−) mice and their wild-type (WT) littermates for in vivo experimentation. Our results reveal that real-ambient PM_2.5 exposure facilitates pulmonary and BM hematopoiesis towards the myeloid axis via a HIF-1α-dependent manner. This myeloid-biased hematopoiesis induces an expansion of mature myeloid cells, especially the NEs and macrophages (Macros), in the liver and spleen, ultimately contributing to the formation of inflammatory lesions and exacerbating hepatic and splenic injury. Above all, our findings unveil lineage-specific hematopoietic toxicity of real-ambient PM_2.5 and offer novel mechanistic insights into how environmental contaminants perturb hematopoiesis.

Methods

Animal experiment

Five-week-old C57BL/6J mice with myeloid lineage-specific knockout of HIF-1α (mHIF-1α^−/−) and their WT littermate controls were purchased from GemPharmatech Co. Ltd., (Nanjing, China). Myeloid lineage-specific HIF-1α^−/−mice and their WT littermates were generated through the mating of LysM-Cre mice and HIF-1α^flox/flox mice, and distinguished via PCR-based genotyping. With ad-libitum food and water, mice were housed in a pathogen-free room with 24 ± 2 °C and 50 ± 5% humidity under 12 h/12 h of light/dark cycle. All animal procedures were approved by the Ethical Animal Experiment Committee of Qingdao University, Hebei Medical University, and the Research Center for Eco-Environmental Sciences, Chinese Academy of Science.

After one week of acclimatization, mHIF-1α^−/− and WT mice were randomly divided into two experimental groups (n = 12 for each group, 6 mice/cage, 48 mice in total) and housed in the filtered air (FA) control and real-ambient PM_2.5 exposure chambers for 8 weeks, respectively. As described in our previous study for the real-ambient PM_2.5 exposure system [25], the air in the FA chambers was filtered by three layers of high-efficient particulate air (HEPA) filters, shaping a PM_2.5-free environment. However, HEPA filters were not applied in the PM_2.5 exposure chambers, which was highly consistent with that of the ambient outdoor air. The mean concentration of PM_2.5 in the chambers was monitored using Aerosol Detector DUSTTRAKTM II (TSI Incorporated, USA). At the predetermined endpoint of 8-week exposure, mice were anesthetized to collect the organs, such as the lung, BM, peripheral blood, liver, and spleen, for subsequent experiments.

Single-cell suspension Preparation

Lung, BM, liver, and spleen were collected to prepare single-cell suspensions according to a modified published protocol. Lung tissue was dissociated using the tissue dissociation kit (Miltenyi Biotec, Germany) and then filtered by 70-µm cell strainers (Biosharp, China). Followed by the centrifugation (400 × g, 10 min, 4℃), red blood cells (RBC) were removed using RBC cracking liquid (Gibco, USA). The single-cell suspension of BM was obtained by flushing the mouse tibia and femur with 2 mL of pre-chilled DMEM (HyClone, USA), filtering through 40-µm cell strainers (Falcon, USA), and eliminating RBC with red cell lysis buffer (Solarbio, China). For the single-cell suspensions of liver and spleen, tissues were submitted to sample pretreatment series, including tissue dissociation, cell filtration, centrifugation (400 × g, 10 min, 4℃), and RBC lysis. Following cell counting with the TC20™ Automated Cell Counter (Bio-Rad, USA), single-cell suspension was submitted for the subsequent flow cytometry assay.

Flow cytometry

Single cells (5 × 10⁶/tube) were resuspended in 50 µL of staining buffer (PBS containing 1% fetal bovine serum (FBS); Gibco, USA) and incubated with fluorescent antibody cocktails. For the evaluation of hematopoietic process, single-cell suspensions of lung and BM were stained with antibodies of a hematopoietic panel for 25 min in the dark at room temperature, followed by the addition of 7-amino-actinomycin D (7-AAD) for an additional 5 min to distinction the live and dead cells within the samples. For the assessment of immune response, single-cell suspensions from BM, liver, and spleen were subjected to a 30-min incubation with the fluorochrome-conjugated antibodies specific to the immune panel. The antibody information and multiple-staining protocols are provided in Supplemental Table S1-S2. After antibody staining, single cells were re-suspended with 500 µL of cell buffer and then analyzed using a flow cytometer (Beckman Coulter, USA). Data analysis was performed using FlowJo Software (version 10.0.7).

Histopathological observation

Mouse liver and spleen tissues were gently extracted, washed with PBS, and then immediately immobilized in 4% paraformaldehyde (PFA, Servicebio, China) for 24 h. The PFA-fixed liver and spleen tissues were then submitted to sample pretreatment series, including gradient dehydrating, paraffin embedding, sectioning (around 5 µm), and staining with hematoxylin and eosin (H&E, Servicebio, China). The histopathological alterations of liver and spleen tissues were observed and photographed using a standard light microscope (Olympus, Japan), and quantified in a blinded manner using Image J (version 1.80).

Quantification of serum cytokines and chemokines

Peripheral blood samples were rested for 2 h at room temperature and then centrifuged at 1500 × g for 10 min to obtain the serum samples, which were stored at -80 °C until analysis. The concentrations of cytokines and four chemokines were measured using the Luminex Bead Immunoassay System (LXSAMSM-17, Univ-bio, China) following the manufacturer’s instructions. In this study, we analyzed ten cytokines and chemokines in the serum, including granulocyte-macrophage colony-stimulating factor (GM-CSF), interleukin-3 (IL-3), IL-4, IL-6, IL-10, tumor necrosis factor-α (TNF-α), C-C motif chemokine ligand 2 (CCL2), CCL3, CCL5, and CCL7. Briefly, 25 µL of standards and serum samples from each group were added into 96-well plates and incubated with a cocktail of respective antibody-coated beads targeting the above ten cytokines and chemokines for 120 min at room temperature. Following 5 times of washing with the buffer, the plate was incubated with 50 µL of biotinylated antibodies for 60 min and then incubated with streptavidin phycoerythrin fluorescent conjugate for another 30 min at room temperature. The fluorescent signal intensity of each bead in the samples was measured by the Luminex 200™ System (Univ-bio, China). Finally, the levels of cytokines and chemokines were expressed in pg/ml, according to the standard curve regression by xPONENT^® Software (version 3.1).

Statistical analysis

The quantitative data were expressed as mean ± standard error of the mean (SEM) with a minimum of six analytical duplicates per group. Differences were deemed statistically significant when the p-value was ≤ 0.05. Statistical analyses were performed using GraphPad Prism (version 8.0.1) and SPSS (version 21.0). Normality and homogeneity of variance were verified for all datasets. For comparisons across the four experimental groups, a two-way analysis of variance (ANOVA) was conducted with “genotype” and “exposure” as fixed factors. When a significant main or interaction effect was detected (p < 0.05), Tukey’s post-hoc test was used for pairwise comparisons.

Results and discussion

PM_2.5-stimulated HIF-1α-dependent myeloid-biased hematopoiesis in the lung

To elucidate the effects of PM_2.5 on hematopoiesis in vivo, mHIF-1α^−/−mice and their WT littermates were exposed to real-ambient PM_2.5 for 8 weeks (Fig. 1A). During the exposure period, the mean PM_2.5 concentration in the exposure chamber was 65.9 ± 36.8 µg/m³ (Fig. 1B), which was equivalent to 60–70% of ambient levels [25]. In contrast, PM_2.5 concentration in the FA chamber was virtually negligible (Fig. 1B). The particle size distribution in the exposure chamber, as reported in our earlier study [25], ranged from 0.5 to 1.5 µm by particle number and 0.5 to 10 µm by mass concentration (Supplemental Fig. 1). Notably, the concentrations of key toxic compositions, including benzo[a]pyrene (BaP), polychlorinated dibenzofurans (PCDFs), polychlorinated dibenzo-p-dioxins (PCDDs), chromium (Cr), and arsenic (As), exceeded the daily limits of Air Quality Standards of China [26]. Our real-ambient exposure system is situated in Shijiazhuang, Hebei Province, China, where PM emissions are predominantly derived from combustion and industrial activities. In this region, sulfur contents [27], toxic elements (e.g., As, Sb, Cr, Co, and Cd) [28], and organic components [29] were present at higher levels than those from other sources, such as dust and vehicle exhaust. Following 8 weeks of PM_2.5 exposure, no significant differences were observed in body weight or organ coefficients of mice (Supplemental Fig. 2).

Fig. 1.

Effects of real-ambient PM_2.5 on the hematopoietic cells in mice lungs. (A) Experimental design for real-ambient PM_2.5 exposure. (B) The mean daily concentrations of PM_2.5 in the FA and PM chambers over the duration of 8-week exposure. (C) Representative flow cytometry analysis of hematopoietic cells in the lungs. Statistics of the cell counts of (D) HSPCs (LT-HSCs, ST-HSCs, MMP2, and MMP3-4), (E) myeloid progenitors (CMPs, GMPs, and MEPs), and (F) lymphoid progenitors (CLPs). (G) Schematic diagram illustrating the impact of 8-week PM_2.5 exposure on pulmonary hematopoietic process in WT and mHIF-1α^−/− mice. Data are expressed as mean ± SEM. Statistical analysis was performed using a two-way ANOVA. ^*p < 0.05, ^**p < 0.01, compared with corresponding control group; ^#p < 0.01, ^##p < 0.01 compared with PM_2.5-exposed WT mice

We then determined hematopoietic responses in the lung, the chief interface of inhaled exposure [17]. With the Sca-1 and CD117 as markers, hematopoietic stem and progenitor cells (HSPCs), MPs, and Sca-1^lowCD117^low cells were isolated from Lin⁻CD45⁺ cells (Fig. 1C). Using CD48 and CD150 antibodies, we defined long-term hematopoietic stem cells (LT-HSCs), short-term hematopoietic stem cells (ST-HSCs), and two types of multipotent progenitors (i.e., MPP2 and MPP3-4) from HSPCs (Fig. 1C). Based on the expression levels of CD34 and CD16/32, three types of MPs containing common myeloid progenitors (CMPs), granulocyte-macrophage progenitors (GMPs), and megakaryocyte-erythroid progenitors (MEPs) were gated from the MP cells (Fig. 1C). From the Sca-1^lowCD117^low cells, common lymphoid progenitors (CLPs) were identified by CD127 antibody (Fig. 1C).

In the lung, PM_2.5-exposed WT mice exhibited a compensatory increase in the numbers of LT-HSCs (2.1-fold, p < 0.05), ST-HSCs (2.0-fold, p < 0.01), and MPP3-4 (1.5-fold, p < 0.01) relative to the control, but displayed a marked reduction in LT-HSCs of mHIF-1α^−/− mice (0.2-fold, p < 0.05, Fig. 1D). This decline reflects a failure to maintain the LT-HSC pool in the absence of HIF-1α, highlighting its role as a key regulator of stem cell resilience under inflammatory stress [20, 30]. Compared with the PM_2.5-exposed WT mice, PM_2.5-exposed mHIF-1α^−/− mice showed a decrease in the number of LT- HSCs (0.2-fold, p < 0.01), ST-HSCs (0.5-fold, p < 0.01), and MPP3-4 (0.7-fold, p < 0.05, Fig. 1D), suggesting HIF-1α acted as a crucial role in PM_2.5-promoted pulmonary hematopoiesis. Myeloid progenitors, especially the GMPs, were increased to 1.7-fold (p < 0.01) in WT mice but not in mHIF-1α^−/− mice upon PM_2.5 exposure (Fig. 1E). In contrast, lymphoid progenitors, i.e., CLPs, were decreased in WT mice but not in mHIF-1α^−/− mice upon PM_2.5 exposure (Fig. 1F). Real-ambient PM_2.5 exposure promoted in situ pulmonary hematopoiesis towards myeloid lineage, however, this myeloid-biased hematopoiesis was attenuated in the absence of HIF-1α, as observed in mHIF-1α^−/− mice (Fig. 1G).

Our study uncovers that real-ambient PM_2.5 exposure facilitates myeloid-biased hematopoiesis in the lung via a HIF-1α-dependent manner. Relative to prior exposure methods, such as intratracheal instillation and intranasal instillation, our unique real-ambient PM_2.5 exposure system was capable of precisely simulating and evaluating in vivo biological responses [25, 31]. A previous finding has demonstrated the inhibition of myeloid-biased hematopoiesis upon combined exposure to formaldehyde and PM_2.5 [32], likely due to the unique real-ambient PM_2.5 exposure in our study and the additional gaseous formaldehyde exposure in the prior investigation. Notably, inhalation exposure to PM_2.5 can upregulate HIF-1α expression in both in situ lung (e.g., bronchial epithelial cells [33] and circulatory system (e.g., aortic endothelial cells [34] and myocardial cells [35]. Due to its indispensable role in determining cell fate [36], HIF-1α has currently attracted considerable attention, which is regarded as a crucial regulator for the self-renewal and differentiation of HSCs [37], especially for the myeloid differentiation. The mHIF-1α^−/−mice model utilized in our study demonstrated the essential role of myeloid HIF-1α in PM_2.5-stimulated myeloid-biased hematopoiesis.

PM_2.5-stimulated HIF-1α-dependent myeloid-biased hematopoiesis in the BM

Our investigation next shifted to examine hematopoiesis within the distal BM. In a similar manner to pulmonary hematopoiesis, we identified eight subsets of hematopoietic cells in the BM (Fig. 2A). Compared with the control, PM_2.5-exposed WT mice exhibited an increase in the numbers of MPP3-4 (2.0-fold, p < 0.01, Fig. 2B) and GMPs (1.3-fold, p < 0.05, Fig. 2C), while restored in the absence of myeloid HIF-1α (Fig. 2B, C), indicating the crucial role of HIF-1α in myeloid-biased hematopoiesis in the BM. Consistent with these results, the levels of granulocyte-macrophage colony-stimulating factor (GM-CSF), a well-established regulator in myeloid hematopoietic cell fate [38], displayed a HIF-1α-dependent elevation in the circulation upon PM_2.5 exposure (Supplemental Fig. 3). Similar to the observation in pulmonary CLPs, a decrease was detected in BM CLPs in the WT mice after PM_2.5 exposure (0.6-fold, p < 0.05, Fig. 2D). As the schematic diagram shown in Fig. 2E, we found that real-ambient PM_2.5 exposure facilitated myeloid-biased hematopoiesis in the BM via the regulation of myeloid HIF-1α.

Fig. 2.

Effects of real-ambient PM_2.5 on the hematopoietic cells in mice BM. (A) Representative flow cytometry analysis of hematopoietic cells in the BM. Statistics of the cell counts of (B) HSPCs (LT-HSCs, ST-HSCs, MMP2, and MMP3-4), (C) myeloid progenitors (CMPs, GMPs, and MEPs), and (D) lymphoid progenitors (CLPs). (E) Schematic diagram illustrating the impact of 8-week PM_2.5 exposure on BM hematopoietic process in WT and mHIF-1α^−/− mice. Quantitative analysis of HSPCs derived from (F) lungs and (G) BM after PM_2.5 exposure. (H) Summary diagram comparing the hematopoietic differentiation in the lungs and BM. Data are expressed as mean ± SEM. Statistical analysis was performed using a two-way ANOVA. ^*p < 0.05, ^**p < 0.01, compared with corresponding control group; ^#p < 0.01, ^##p < 0.01 compared with PM_2.5-exposed WT mice

To further elucidate the relative contributions of pulmonary and BM hematopoiesis to subsequent immune dysregulation and resultant adverse outcomes, we assessed the alterations of HSPCs, a pivotal determinant of hematopoiesis (i.e., generating all blood cells) [39, 40], in these two sites. Compared with the control, PM_2.5-exposed WT mice displayed an increase of HSPCs in the lungs (1.6-fold, p < 0.01, Fig. 2F) and BM (1.7-fold, p < 0.01, Fig. 2G), with no alteration in the absence of myeloid HIF-1α. As illustrated in the comprehensive analysis of key hematopoietic cells, we distinguished the myeloid and lymphoid progenitors derived from HSPCs and color gradients indicate cell quantities (Fig. 2H). The colors red, yellow, green, and white are indicative of four distinct magnitudes, with red denoting the highest and white the lowest. Notably, the abundance of hematopoietic cells, especially myeloid progenitor cells, in the BM exceeded that in the lungs by a significant margin (Fig. 2H), highlighting a dominant role of BM hematopoiesis to immune cell replenishment. Consistent with these phenotypic observations, colony-forming unit (CFU) assays of HSCs in BM showed an increase in colony-forming units-granulocyte macrophage (CFU-GM) output following PM_2.5 exposure (Supplemental Fig. 4), functionally affirming a shift toward myeloid-biased hematopoiesis.

Although pulmonary hematopoiesis is sensitive to inhaled PM_2.5 exposure, our results demonstrate that long-term exposure mainly affects hematopoiesis in the distal BM, with most HSCs remaining in this site, except for a minor fraction mobilizing under certain conditions [41]. As a critical repository for immune cells, the BM is indispensable for host defense, where regulatory signals orchestrate and facilitate the production of the billions of blood cells essential for maintaining physiological homeostasis [42]. The data obtained herein provides, for the first time, direct evidence of in vivo characteristics of hematopoietic differentiation upon long-term real-ambient PM_2.5 exposure, highlighting the impact of HIF-1α in the myeloid-biased hematopoiesis. Chronic inflammatory signaling mediated by HIF-1α induces lasting epigenetic and functional changes in hematopoietic cells [43]. Therefore, we posit that the myeloid-biased hematopoiesis following 8-week PM_2.5 exposure represents a stable hematopoietic reprogramming. Supporting this, prior studies demonstrated that inflammatory stress like PM_2.5 caused irreversible functional exhaustion in HSCs, with effects persisting long after insult removal [32, 44].

PM_2.5-induced mature myeloid cell expansion via a HIF-1α-dependent manner

To assess the contributions of PM_2.5-stimulated BM hematopoiesis to downstream immune cell generation (Fig. 3A), we initially identified six types of mature immune cell populations in the BM (Fig. 3B). In detail, we used specific markers CD4 and CD8 to gate CD4⁺ T and CD8⁺ T cells from CD45⁺ cells (Fig. 3B). Using specific CD19 and Ly6G, we defined B cells and NEs (Fig. 3B). With the F4/80 and CD11c as markers, we identified DCs, monocytes (Monos), and Macros (Fig. 3B). Compared with the control, PM_2.5-exposed WT mice had a higher abundance of Macros (1.3-fold, p < 0.05), while exhibited a decrease in CD4⁺ T cells (0.8-fold, p < 0.05, Fig. 3C), suggesting the expansion of mature myeloid cells caused by PM_2.5. The myeloid HIF-1α deficiency attenuated the influences of PM_2.5 on the increased Macros and decreased CD4⁺ T cells (Fig. 3C), indicating myeloid HIF-1α was involved in the PM_2.5-induced mature myeloid cell expansion in the BM. Likewise, PM_2.5 exposure caused a HIF-1α-dependent mature myeloid cell expansion in the circulation (Supplemental Fig. 5).

Fig. 3.

Alterations of mature immune cells in mice BM, liver, and spleen stimulated by real-ambient PM_2.5. (A) Schematic diagram for hematopoietic differentiation and immune cell production. (B) Representative flow cytometric dot plots and (C) quantitative analysis of CD4⁺ T cells, CD8⁺ T cells, B cells, NEs, Macros, Monos, and DCs in the BM. (D) Representative and (E) quantitative flow cytometric analysis of Macros, Monos, and DCs derived from liver tissues. (F) Fluorescence-activated cell sorting (FACS) plots for the identification of NE. (G) Cell numbers of NE in the liver. (H) Representative flow cytometry of Macros, Monos, and DCs in the spleen and (I) corresponding quantitative analysis. (J) FACS plots of NE as quantified in (K) with cell counts in the spleen. (L) Heatmap for the alterations of mature immune cells in the BM, liver, and spleen tissues of WT and mHIF-1α^−/− mice. Data are expressed as mean ± SEM. Statistical analysis was performed using a two-way ANOVA. ^*p < 0.05, ^**p < 0.01, compared with corresponding control group; ^#p < 0.01, ^##p < 0.01 compared with PM_2.5-exposed WT mice

We then investigated the in vivo characteristics of immune cell composition in the liver and spleen, two vital peripheral immune organs that are susceptible to PM_2.5 exposure [45–47]. In a similar manner to the BM, we characterized six types of immune cells in these tissues (Supplemental Fig. 6). As depicted in Fig. 3D, a significant increase of Macros (1.8-fold, p < 0.05) was observed in the PM_2.5-exposed WT mice, however, the deficiency of myeloid HIF-1α attenuated the elevation of Macros (Fig. 3E). Meanwhile, the number of NEs was increased to 2.5-fold (p < 0.01) in WT mice but not in mHIF-1α^−/− mice upon PM_2.5 exposure (Fig. 3F, G), indicative of a HIF-1α-dependent mature myeloid cells expansion in the liver. In the spleen, PM_2.5-exposed WT mice exhibited an increase in the number of Macros (1.6-fold, p < 0.05) as compared to the control, while decreased Macros numbers in mHIF-1α^−/− mice (0.8-fold, p < 0.05, Fig. 3H, I). As illustrated in Fig. 3J, NE counts also displayed a HIF-1α-dependent expansion in the spleen upon PM_2.5 exposure (Fig. 3K), further confirming the role of HIF-1α in mature myeloid cell expansion. As expected, in both WT and mHIF-1α^−/− mice, lymphoid cells, including CD4⁺ T, CD8⁺ T, and B cells, remained unchanged in the liver and spleen after PM_2.5 exposure (Supplemental Fig. 7). A heatmap analysis showed that PM_2.5 exposure led to a substantial increase in mature myeloid cells in BM, liver, and spleen, notably characterized by the elevation of Macros and NEs, however, this increase was attenuated in the absence of myeloid HIF-1α (Fig. 3L). Therefore, PM_2.5 exposure induced a substantial increase in mature myeloid cells in a HIF-1α-dependent manner.

Our findings clarify that real-ambient PM_2.5 exposure causes a HIF-1α-dependent mature myeloid cell expansion in the BM. Undoubtedly, these mature immune cells migrate to target tissues of infection or injury to fulfill their functions in surveillance and immune responses [48]. In this study, we have observed that real-ambient PM_2.5 exposure results in a significant increase of mature myeloid cells within the liver and spleen, a phenomenon modulated by the myeloid HIF-1α. This phenomenon was partially supported by an in vivo tracking study, which has shown that fine and ultrafine particles are capable of penetrating the lung barrier and reaching other organs, such as the liver and spleen [49]. Accordingly, our study underscores the systemic impact of environmental PM_2.5 on hematopoietic and immune compartments beyond the primary site of exposure.

PM_2.5-stressed myeloid-biased hematopoiesis facilitated the expansion of mature myeloid cells in a HIF-1α-dependent manner

To clarify the role of myeloid HIF-1α in PM_2.5-stressed myeloid-biased hematopoiesis and mature myeloid cell expansion, we then analyzed the hematopoietic and immune cells with specific alterations in WT or mHIF-1α^−/− mice (Fig. 4). Under the PM_2.5-induced hematopoietic stress, HIF-1α emerged as a pivotal regulator in myeloid-biased hematopoiesis, orchestrating both the expansion of myeloid progenitors (i.e., GMPs) and the attenuation of lymphoid progenitors (i.e., CLPs) (Fig. 4A, B and Supplemental Fig. 8A). The primary hematopoietic site, i.e., BM, is located in the hypoxic bone cavities due to its anatomical structure and substantial oxygen consumption of hematopoiesis [50]. As a hypoxia-responsive sensor protein, myeloid HIF-1α certainly plays pivotal roles in HSCs and other cells residing in the hypoxic BM microenvironment [51].

Fig. 4.

HIF-1α-dependent expansion of myeloid lineage in hematopoietic and immune cells upon real-ambient PM_2.5 exposure. PM_2.5 caused HIF-1α-dependent lineage development in the lung and BM, as well as the resultant immune expansion in the PB, BM, and peripheral immune organs (i.e., liver and spleen). Two Venn diagrams showed the different hematopoietic cells between WT and mHIF-1α^−/− mice in the lung and BM, and four Venn diagram showed the different mature immune cells between WT and mHIF-1α^−/− mice in the PB, BM, liver, and spleen

During the hematopoietic process, GMPs can differentiate into mature myeloid cells (e.g., Macros, Monos, and NEs), and CLPs possess the capacity to differentiate into lymphoid cells (e.g., T cells and B cells) [52]. In the BM, HIF-1α played a mediating role in the expansion of Monos and Macros, concurring with the increased number of Monos in the PB (Fig. 4C, D and Supplemental Fig. 8B). Under the regulation of myeloid HIF-1α, mature myeloid cell expansion ultimately influenced the amplification of mature effector immune cells, especially for the Macros and NEs, in the liver and spleen (Fig. 4E, F and Supplemental Fig. 8B). This observation is consistent with a prior study, which elucidated that emergency hematopoiesis contributed to replenishing innate immune cells to ensure adequate production of granulocytes and Monos during systemic infections [53]. Therefore, HIF-1α emerged as a critical role in PM_2.5-induced myeloid-biased hematopoiesis characterized by the differentiation into GMPs, which prompted an augmentation in myeloid effector cells (i.e., Macros and NEs) within peripheral immune organs (i.e., liver and spleen), thereby culminating in an elevated risk of inflammatory injury.

PM_2.5-triggered hepatic and Splenic inflammatory injury mediated by HIF-1α-dependent mature myeloid cell expansion

To determine whether PM_2.5-induced mature myeloid cell expansion caused liver and spleen injury in a HIF-1α-dependent manner, we initially conducted a histopathological analysis in mice. As depicted in representative H&E staining of liver tissues, PM_2.5 exposure led to the infiltration of inflammatory cells (red arrows in Fig. 5A), but no obvious alteration was observed in the absence of HIF-1α. Meanwhile, PM_2.5-exposed mouse liver exhibited an increase in the number of inflammatory lesions (5.0-fold, p < 0.01) and the thickness of hepatic sinusoidal (1.4-fold, p < 0.01), indicative of a hepatocyte remodeling as a consequence of liver injury [54], however, this histopathological alteration was mitigated in mHIF-1α^−/− mice (Fig. 5B). In the spleen, PM_2.5-exposed WT mice displayed the enlargement of splenic red pulp (red stars in Fig. 5C), accompanied by the atrophy of white pulp (Fig. 5C, D), suggesting the mature myeloid cell expansion within this region. Meanwhile, the thickness of splenic capsule was reduced after PM_2.5 exposure (0.5-fold, p < 0.01, Fig. 5C, D), hinting at the possibility of splenic enlargement and chronic inflammation. However, there was no notable alteration in the spleen tissues of PM_2.5-exposed mHIF-1α^−/− mice.

Fig. 5.

Inflammatory injury of liver and spleen induced by real-ambient PM_2.5 exposure. (A) Representative images of H&E staining and (B) quantitative assessment of the hepatic sinusoidal and inflammatory foci in the liver. Black arrows indicate the infiltration of inflammatory cells. Scale bar: 200 µm. (C) Representative H&E staining of mouse spleen upon PM_2.5 exposure. The white pulp (WP) was circled by the yellow dashed line and the red pulp (RP) was marked by red stars. Scale bar: 200 µm. (D) Quantification of the percentage of RP/WP ratio and the thicknesses of splenic capsules. (E) Statistics of the levels of relevant cytokines in the PB. (F) Quantitative analysis of CCL2 and IL-4 in the PB. Data are expressed as mean ± SEM. Statistical analysis was performed using a two-way ANOVA. ^**p < 0.01, compared with corresponding control group; ^##p < 0.01 compared with PM_2.5-exposed WT mice

We then examined the levels of crucial cytokines related to mature myeloid cell recruitment (i.e., CCL2, CCL3, CCL5, CCL7) and tissue damage (i.e., TNF-α, IL-3, IL-4, IL-6, IL-10). In the plasma, PM_2.5-exposed WT mice displayed higher levels of CCL2 and IL-4, as well as lower levels of CCL3 and CCL7 relative to the control, whereas only CCL5 concentrations exhibited a decrease in mHIF-1α^−/− mice (Fig. 5E). These results hinted at the potential role of CCL2 and IL-4 in PM_2.5-induced mature myeloid cell recruitment and tissue damage. Compared with the control, PM_2.5 exposure elevated the concentrations of CCL2 and IL-4 to 1.2-fold (p < 0.01) and 1.9-fold (p < 0.01) in WT mice, respectively (Fig. 5F), which was restored to normal levels in mHIF-1α^−/− mice (Fig. 5F). CCL2, the primary ligand for CCR2, exhibited an elevated presence in the liver of individuals with non-alcoholic steatohepatitis (NASH) and murine models with steatohepatitis and fibrosis [55]. The CCL2/CCR2 signaling axis orchestrates the recruitment of Macros into target tissues, making a significant contribution to the progression of tissue injury [55]. Moreover, large amounts of IL-4 can lead to massive accumulation of Macros within the inflammatory tissue and aggravate tissue injury [56].

Our results indicate that real-ambient PM_2.5-triggered hepatic and splenic inflammatory injury is associated with HIF-1α-dependent mature myeloid cell expansion. Emerging epidemiological and experimental evidence have elucidated the multi-tissue injury induced by PM_2.5, including the lung [57], heart [58], liver [59], and spleen [60]. A prior study utilizing the real-ambient system demonstrated that 16 weeks of PM_2.5 exposure caused lung injury with characteristics of inflammation and collagen deposition, which may be attributed to increased abundance of NEs and monocyte-derived cells within lung tissues [61]. Our previous research also revealed that chronic exposure to real-ambient PM_2.5 resulted in obvious myocardial hypoxia injury, accompanied by an upregulation of HIF-1α protein in the myocardium of mice [35]. Our findings on liver injury related to the mature myeloid cell expansion are supported by considerable evidences, for instance, liver-resident and recruited Macros have been reported to be a major contributing factor to the progression of nonalcoholic fatty liver disease (NAFLD) [62]. A prior study demonstrated that NEs transformed into an inflammatory phenotype under consistent inflammatory conditions, further exacerbating the development and progression of liver damage [63]. Furthermore, mature myeloid cells also exhibited a close relevance with spleen injury. Splenic Macros, alongside their fibroblasts, established robust and potential cellular circuits, maintaining homeostasis and splenic functionality through mutually beneficial interactions [64]. Notably, a previous study revealed that PM_2.5-triggered irreversible multi-organ injury due to sustained pro-inflammatory responses [25], implying that the hepatic and splenic inflammatory injury observed in our study may similarly persist. Altogether, the present study uncovers that real-ambient PM_2.5 exposure stimulates pulmonary and BM hematopoiesis towards myeloid axis via a HIF-1α-dependent manner. The HIF-1α-dependent myeloid-biased hematopoiesis facilitates the expansion of mature myeloid cells, especially the NEs and Macros, in the liver and spleen, further resulting in the formation of inflammatory lesions and even the aggravation of hepatic and splenic injury (Supplemental Fig. 9). Considering the widespread population impacted by airborne PM_2.5 contaminations, more extensive studies are still urgently required to elucidate the intricate relationship between HIF-1α and PM_2.5-induced inflammatory injury, with the aim of identifying the potential therapeutic targets for prevention and mitigation of PM_2.5-associated inflammatory diseases.

Conclusion

Utilizing a real-ambient PM_2.5 exposure system, we employed mHIF-1α^−/− mice alongside their wild-type littermates to investigate in vivo hematopoietic responses to PM_2.5 and critically evaluate the role of HIF-1α. Our study provides in vivo evidence demonstrating that chronic PM_2.5 exposure promotes HIF-1α-dependent myeloid-biased hematopoiesis, a phenomenon observed in local lung and systemic BM. This shift drives myeloid cell expansion and culminates in inflammatory injury. These findings present novel mechanistic insights, establishing a direct link between ambient PM_2.5 exposure and hematopoietic toxicity via HIF-1α signaling. Critically, these data underscore considerable environmental health implications and reinforce the need for air quality regulations and public health interventions aimed at mitigating the adverse effects of PM_2.5. However, there are, as well as, inherent limitations in our study. First, while we address PM_2.5-induced inflammatory injury in immune-responsive organs like the liver and spleen, further work is required to elucidate its effects on other crucial organs. Second, our experimental design prioritized the chronic impacts of PM_2.5 exposure, ignoring the early and sensitive responses of lung-resident hematopoiesis. Finally, although our data suggest a shift toward myeloid-biased hematopoiesis upon PM_2.5 exposure, state-of-the-art methods—such as transplantation or radiation experiments—will be essential to functionally confirm this process.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

H.Y.Y. and Y.D.C.: synthesized performed the experiments and prepared the manuscript. Y.Y.W., M.L., G.T., R.Z.: conducted the correlation analysis and analyzed the data and made the graphs. X.T.J. and G.B.Q.: reviewed and edited the manuscript. Y.X.Z. and G.B.J.: supervised this research project. The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

Funding

This work was supported by grants from the Research Funds of Joint Research Center for Occupational Medicine and Health of IHM (OMH-2024-011), the National Natural Science Foundation of China (82473674, 82241086, U24A20772), and the Qingdao Natural Science Foundation (24-4-4-zrjj-157-jch).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

The study protocol was proven by the Institutional Animal Care and Use Committee of Qingdao University, Hebei Medical University, and the Research Center for Eco-Environmental Sciences, Chinese Academy of Science.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.