Abstract

   Pulmonary hypertension (PH) is a severe disease characterized by
   pulmonary vascular remodeling in which various immune cells play a
   critical role in vascular remodeling, although the details are still
   vague. Furthermore, current clinical treatments primarily focus on
   pulmonary vasodilation, but do not fundamentally address vascular
   remodeling itself. Here, first significant changes in neutrophils
   during the development of PH are demonstrated and show that neutrophil
   depletion can effectively attenuate disease progression. Moreover, the
   data show that neutrophil‐derived S100A9 is the key mediator to promote
   vascular remodeling, while both knockout and inhibition of S100A9 can
   prevent PH. In a co‐culture system of neutrophils and endothelial cells
   (ECs), hypoxic stimulation leads to increased S100A9 secretion by
   neutrophils, which activates the RAGE/PI3K/AKT pathway and causes
   dysfunction of ECs. These findings suggest that neutrophil‐derived
   S100A9 mediated neutrophil‐EC crosstalk plays an important role in
   pulmonary vascular remodeling, providing a promising strategy for
   treatment of PH.

   Keywords: endothelial dysfunction, neutrophils, pulmonary vascular
   remodeling, S100A9
     __________________________________________________________________

   This study investigates the critical mechanisms underlying pulmonary
   hypertension progression, with a focus on the novel role of
   neutrophil‐derived S100A9 in endothelial dysfunction‐mediated pulmonary
   vascular remodeling. By integrating human lung samples, multi‐omics
   analyses, animal models, and mechanistic in vitro studies, it is
   revealed that how neutrophil–endothelial interactions mediated through
   the S100A9‐RAGE‐PI3K‐AKT signaling pathway contribute to endothelial
   dysfunction and vascular remodeling in PH.

   graphic file with name ADVS-12-e04397-g001.jpg

1. Introduction

   Pulmonary hypertension (PH) is a heterogenous clinical disease
   characterized foremost by an abnormal increase in pulmonary artery
   pressure,^[ [50]^1 ^] universally defined by a mean pulmonary artery
   pressure > 20 mmHg.^[ [51]^2 ^] Across all its forms, pulmonary
   hypertension is associated with adverse vascular remodeling with
   obstruction and stiffening of the pulmonary vasculature.^[ [52]^3 ^]
   Without proactive management vascular remodeling leads to hypertrophy
   and ultimately failure of the right ventricle.^[ [53]^4 ^]

   The pathological mechanism underlying pulmonary vascular remodeling is
   intricate and remains incompletely understood. A key characteristic of
   vascular remodeling is the abnormal proliferation of structural cells
   and the accumulation of immune cells within and around the vascular
   wall,^[ [54]^5 , [55]^6 ^] in which the endothelium, as the innermost
   layer of the vascular wall, is at the forefront of sensing and
   responding to hemodynamic and hypoxic stimuli within blood vessels.^[
   [56]^7 ^] Therefore, endothelial dysfunction is a common initiating
   event in many diseases characterized by vascular remodeling.

   In recent years, immune and inflammatory responses in PH have gained
   significant attention because numerous studies have suggested that
   there is a close relationship between inflammation in the pathogenesis
   of PH.^[ [57]^8 , [58]^9 , [59]^10 ^] When immune cells are activated,
   they release large amounts of inflammatory factors, inducing
   endothelial cell proliferation and dysfunction, ultimately leading to
   vascular remodeling and right ventricular (RV) dysfunction.^[ [60]^5 ^]
   These findings suggest that implementing immunotherapy during PH
   progression may have a beneficial effect on the course of the disease.
   Therefore, it is urgent to clarify the inflammatory mechanisms during
   pulmonary vascular remodeling.

   Our group have previously shown that various immune cells and
   inflammatory factors, including IL‐17, IL‐33, Th17 cells, and
   macrophages, are involved in endothelial dysfunction and vascular
   remodeling of PH.^[ [61]^11 , [62]^12 , [63]^13 ^] Furthermore, our
   analysis of transcriptomic and proteomic data showed that the common
   differentially expressed genes and proteins were primarily focused on
   inflammation‐ and immunity‐related molecules, in which S100A9 is one of
   the most significantly expressed molecules among them. S100A9, also
   known as calgranulin B or myeloid‐related protein14 (MRP‐14), is an
   endogenous damage‐associated molecular pattern molecule of the S100
   calcium‐binding protein family.^[ [64]^14 ^] This molecule is mainly
   released by neutrophils,^[ [65]^15 , [66]^16 ^] and primarily functions
   as an inflammatory mediator, contributing to sustained inflammation and
   endothelial injury in septic or aged mice.^[ [67]^17 , [68]^18 ,
   [69]^19 ^] Notably, S100A9 has been identified as a key regulator of
   inflammation, playing a critical role in the development and
   progression of various cardiovascular diseases—including
   atherosclerosis, myocardial infarction, aortic aneurysm, and peripheral
   arterial diseases, through modulating inflammatory responses,
   endothelial function, cellular proliferation, autophagy, apoptosis, and
   cell death.^[ [70]^14 ^] Up to date, however, the role of S100A9 in
   endothelial dysfunction‐mediated pulmonary vascular remodeling is still
   unclear.

   In this context, we aimed to explore whether neutrophil‐derived S100A9
   contribute to endothelial dysfunction in the development of PH and its
   potential mechanism, which hopefully provide novel therapeutic targets
   to mitigate pulmonary vascular remodeling and improve outcomes in
   patients suffering from PH.

2. Results

2.1. Increased Neutrophil Populations and Lung Infiltration in SuHx‐PH Mice

   To explore neutrophil dynamics in the context of PH, we constructed a
   SuHx‐induced PH mouse model and conducted scRNA‐seq on lung tissue.
   After sequencing, all cells were annotated based on their corresponding
   marker genes (Table [71]S2, Supporting Information), identifying 18
   distinct cell clusters (Figure [72]1A). To assess cell population
   changes across experimental groups, we first performed clustering and
   cell type annotation based on the scRNA‐seq data (Figure [73]1B). A
   significant increase in the proportion of neutrophils was observed in
   the SuHx‐induced PH group compared to the control group. We further
   conducted differential abundance (DA) analysis using the miloR
   framework, as shown in Figure [74]1C. The left panel displays uniform
   manifold approximation and projection (UMAP) projections of Control
   group (blue) and SuHx‐induced PH group (pink) cells, while the right
   panel shows DA results highlighting differential cell abundance across
   conditions. Figure [75]1D illustrates a honeycomb plot based on log2
   fold change, which further confirms a substantial enrichment of
   neutrophil populations in the SuHx‐induced PH group, suggesting a
   potential involvement of these cells in disease progression. To further
   validate the increased neutrophil infiltration observed in the lungs of
   SuHx‐induced PH mice, immunohistochemical staining for neutrophil
   elastase (NE) was conducted. The results demonstrated a significant
   elevation in NE‐positive cells within the pulmonary tissues of
   SuHx‐induced PH mice, indicating enhanced neutrophilic activity
   (Figure [76]1E). Complementary flow cytometry analysis of CD11b^+Ly6G^+
   neutrophil populations confirmed a marked increase in neutrophil
   abundance in the SuHx‐induced PH group relative to controls
   (Figure [77]1F).

Figure 1.

   Figure 1
   [78]Open in a new tab

   Increased neutrophil populations and lung infiltration in SuHx‐PH mice.
   A) UMAP visualization showed the presence of eighteen distinct lung
   cell types, highlighting the diversity of cell populations within the
   tissue. B) The percentage of neutrophils among total cells of mice with
   or without SuHx‐induced PH. The distribution of S100A9^+ cells on the
   UMAP plot. C) UMAP showing cell neighborhood distributions from both
   groups (left), and differential abundance (DA) analysis using miloR,
   with neighborhoods colored by log fold change (SuHx‐PH vs.Control,
   right). Significant DA regions were identified by GLMM (FDR < 0.1). D)
   Violin plots show the distribution of log fold change (logFC) values
   for each annotated cell type. E) IHC for NE in lung sections of mice
   with or without SuHx‐induced PH (n = 5 per group). F) Representative
   flow cytometric images of CD11b^+ Ly6G^+ neutrophils among live cells
   in lung tissues of mice with or without SuHx‐induced PH. G) Flow
   cytometric analyses of lung tissues of mice (n = 5‐6 per group). H)
   Quantity analysis of NE‐positive cells. *p < 0.05, **p < 0.01. Unpaired
   t‐test was used.

   Taken together, these results suggest a potential role of neutrophils
   in the pathogenesis of PH.

2.2. Neutrophil Depletion Attenuates Symptoms of Mice with SuHx‐Induced PH

   Because studies derived from ours and others have shown that there are
   increases in neutrophil numbers which might associate with disease
   progression in SuHx‐induced PH, we further investigated the role of
   neutrophil in the pulmonary vascular remodeling. To achieve this, an
   antibody against Ly6G was employed. It has been known that Ly6G is a
   glycoprotein specifically expressed on the surface of neutrophils, and
   antibodies targeting Ly6G results in immune‐mediated neutrophil
   depletion.^[ [79]^20 , [80]^21 ^] To clarify role of neutrophils in
   SuHx‐induced PH pathogenesis, we depleted neutrophils in mice by
   administering intraperitoneal injections of an anti‐Ly6G antibody (200
   µg on alternate days) and established a SuHx‐induced PH model in both
   anti‐Ly6G‐treated and isotype control‐treated mice (Figure [81]2A). The
   results showed that there were significantly reduced percentages of
   neutrophils in the lung tissues and blood of anti‐Ly6G‐treated mice
   compared to those of isotype control‐treated mice (Figure [82]2B,E;
   Figure [83]S2A,B, Supporting Information). After neutrophil depletion,
   a population of CD11b^+ and Ly6G^low cells still remained, which is
   possibly due to the stimulation of neutrophil precursors in the bone
   marrow, leading to an increase in newly generated neutrophils.^[
   [84]^21 ^] However, these cells are not capable of performing their
   full functional roles. Western blot confirmed that S100A9 expression
   was also markedly reduced in the lung tissues of the anti‐Ly6G‐treated
   group (Figure [85]2C). Additionally, anti‐Ly6G‐meidiated neutrophil
   depletion resulted in lower levels of RVSP, right ventricular
   hypertrophy index (RVHI), and RV weight radio (right ventricular/body
   weight (RV/BW) of mice with SuHx‐induced PH (Figure [86]2D). Since
   vascular remodeling in PH primarily occurs in the distal pulmonary
   arterioles, we calculated percent medial thickness in intrapulmonary
   arterioles with a diameter of <100 µm. Consistent with these results,
   we observed the significantly reduced vascular remodeling <100 µm
   (Figure [87]2F,G), and cardiomyocyte hypertrophy (Figure [88]2F,H) in
   neutrophil‐depleted SuHx‐induced PH mice. Pulmonary vascular remodeling
   leads to increase pulmonary arterial pressure, RV hypertrophy and RV
   dysfunction, which is a hallmark of end‐stage PH^3. Furthermore, RVWT
   was significantly reduced after anti‐Ly6G treatment, while analysis of
   RV function, including PAAT and TAPSE, showed that there was a marked
   improvement (Figure [89]2I,J).

Figure 2.

   Figure 2
   [90]Open in a new tab

   Neutrophil deficiency attenuates PH in mice. A‐J: Anti‐Ly6G antibody
   and isotype (i.p.) were administered to C57BL/6 mice under control and
   SuHx‐induced conditions. A) Schematic diagram for establishing
   experimental murine model. B) Representative flow cytometric images of
   CD11b^+ Ly6G^+ neutrophils among CD45^+ cells in lung tissues (left
   panel) and blood (right panel) of experimental mice. C) Representative
   Western blot of S100A9 in the lungs of experimental mice. D) RVSP, RVHI
   and RV/BW of mice (n = 4‐8 per group). E) Flow cytometric analyses of
   lung tissues (left panel) and blood (right panel) of mice (n = 3‐5 per
   group). F) Representative images of H&E and IF staining for α‐SMA/CD31
   in lung sections of experimental mice, as well as representative images
   of WGA staining showing cardiomyocyte hypertrophy in the RV. G)
   Quantification of wall thickness of the pulmonary vasculature for
   vessels of 0–25 µm, 26–50 µm, 51–75 µm and 76–100 µm in diameter (n = 8
   per group). H) Quantification of WGA staining (n = 5 per group). I)
   Echocardiography showing the representative figures of PAAT, RVWT and
   TAPSE. J) Statistical analysis of PAAT, RVWT and TAPSE (n = 3‐5 per
   group). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Two‐way
   ANOVA was performed, followed by Bonferroni correction for six pairwise
   comparisons among the four experimental groups.

2.3. S100A9 Expression is Upregulated in the Lungs and Predominantly
Originates from Neutrophils during the Development of PH

   To further investigate the specific molecules and underlying mechanisms
   by which neutrophils exert their effects in PH, we conducted
   integrative multi‐omics analyses. Transcriptomic and proteomic profiles
   were obtained from the lung tissues of SuHx‐induced PH mice.
   Transcriptome sequencing identified 656 differentially expressed genes
   (DEGs), while proteomics analysis revealed 323 differentially expressed
   proteins (DEPs, Figure [91]3A). Cross‐analysis showed 37 overlapping
   differentially expressed genes and proteins (Figure [92]3A), among
   which S100A9, a protein involved in inflammatory events,^[ [93]^22 ^]
   was significantly upregulated (Figure [94]3B,C). It has been shown that
   S100A9 is predominantly derived from monocytes and neutrophils in vivo,
   accounting for ≈40% of neutrophil cytosolic proteins, and functions as
   a key alarmin‐like molecule.^[ [95]^23 ^] We further validated the
   significant upregulation of S100A9 in scRNA‐seq data from the same
   mouse model, confirming its elevated expression (Figure [96]3D). As
   shown in Figure [97]3E, immunoreactivity for S100A9 was significantly
   higher in lung sections of COPD patients with PH than those of controls
   and COPD patients without PH. Correspondingly, data of immunostaining
   and Western blot showed that the protein expression of S100A9
   significantly increased in the lung tissues of SuHx‐induced PH mice
   (Figure [98]3F,G), chronic hypoxia‐induced PH mice and rats with
   MCT‐induced PAH (Figure [99]S3A,B, Supporting Information) compared to
   those of control animals. Additionally, ELISA analysis showed that
   serum levels of S100A9 were significantly elevated in patients with the
   COPD‐PH compared to those of controls (Figure [100]3H). We further
   analyzed the expression profile of S100A9 using scRNA‐seq data,
   revealing that S100A9 was predominantly expressed in neutrophils
   (Figure [101]3I), with significantly higher expression levels observed
   in neutrophils of SuHx‐induced PH mice (Figure [102]3J). This finding
   was further supported by immunofluorescence staining in which the
   number of neutrophils was markedly increased in the lung tissues of
   both COPD and COPD‐PH patients compared with those of controls, and the
   most S100A9 immunoreactivities were co‐localized with neutrophils
   (Figure [103]3K). Similar patterns were observed in the lung tissues of
   SuHx‐PH mice, where an increased number of S100A9‐positive neutrophils
   were found surrounding pulmonary endothelial cells (Figure [104]3L).

Figure 3.

   Figure 3
   [105]Open in a new tab

   S100A9 is upregulated in the lungs during the pathogenesis of PH. A)
   Venn diagram showing the overlap between DEGs from transcriptomic
   analysis and DEPs from proteomic analysis. B) Heatmap of the 37
   differentially genes identified in the SuHx‐induce PH model (|log[2]FC|
   ≥ 1.5, FDR < 0.05). C) Heatmap of the 37 differentially expressed
   proteins identified in the SuHx‐induce PH model (|log[2]FC| ≥ 1.0, FDR
   < 0.05). D) UMAP plot visualizing the expression of S100A9 across all
   cells. Color intensity indicates the normalized expression level of the
   gene. E) IHC and quantity analysis of S100A9 in sections of human lung
   tissues (3 groups, n = 3 per group). F) IHC and quantity analysis of
   S100A9 in sections of the lung tissues of SuHx‐induced PH mice (n = 8
   per group). G) Western blot analysis of S100A9 in the lung tissues of
   the experimental mice (fold change of SuHx versus Control group, n = 6
   per group). H) ELISA analysis of serum levels of S100A9 in patients
   with COPD‐PH and control groups (n = 12‐15 per group). I) Dot plot
   showing the expression level of S100A9 across different cell types. J)
   Boxplot showing the expression level of S100A9 in neutrophils across
   different groups. K) Representative images of IF staining for MPO (red)
   and S100A9 (green) in sections of the lung tissues of healthy controls,
   COPD patients without PH and COPD‐PH patients. Nuclei were
   counterstained with DAPI (blue). L) Representative images of IF
   staining for CD31 (white), Ly6G (red) and S100A9 (green) in sections of
   the lung tissues of mice with or without SuHx‐induced PH. Nuclei were
   counterstained with DAPI (blue). Data are presented as mean ± SD
   deviation, normalized to β‐actin relative gene expression was
   calculated using the 2^(‐ΔΔCt) method. *p < 0.05, **p < 0.01, ***p <
   0.001, ****p < 0.0001. One‐way ANOVA was performed, followed by
   Bonferroni post hoc correction for three pairwise comparisons among the
   three experimental groups (E). Unpaired t‐test was used for comparisons
   in panels F‐H.

2.4. S100A9 Deficiency Attenuates Symptoms of Mice with SuHx‐Induced PH

   To investigate whether change in S100A9 expression influences the
   phenotype in PH development, S100A9 knockout (S100A9^−/− ) mice were
   used to establish a SuHx‐induced PH model (Figure [106]4A). Western
   blot showed that there was the absence of S100A9 protein in the lung
   tissues of the S100A9^−/− group (Figure [107]4B). Data also showed that
   the absence of S100A9 significantly attenuated the increase in RVSP,
   reduced the extent of RVHI and RV/BW (Figure [108]4C) in mice with
   SuHx‐induced PH. The results showed that treatment with SuHx induced
   significant remodeling of these arterioles in WT mice, while S100A9^−/−
   mice exhibited reduced remodeling (Figure [109]4D,E). WGA staining
   (Figure [110]4D,F) and echocardiography in mice revealed that
   S100A9^−/− significantly mitigated SuHx‐induced RVWT and cardiomyocyte
   hypertrophy, and improved RV function, including PAAT and TAPSE
   (Figure [111]4G,H).

Figure 4.

   Figure 4
   [112]Open in a new tab

   S100A9 deficiency attenuates symptoms of mice with SuHx‐induced PH. A)
   Schematic diagram of establishment of experimental murine model. B‐H:
   WT or S100A9^−/− mice were under control and SuHx‐induced conditions.
   B) Representative Western blot of S100A9 in the lungs of experimental
   mice. C) RVSP, RVHI and RV/BW of mice (n = 4‐8 per group). D)
   Representative images of haematoxylin and eosin (H&E) and α‐SMA (smooth
   muscle α‐actin)/CD31 IF staining in lung sections of the experimental
   mice, as well as representative images of WGA staining showing
   cardiomyocyte hypertrophy in the RV. E) Quantification of wall
   thickness of the pulmonary vasculature of the experimental mice (n =
   5‐8 per group). F) Quantification of WGA staining (n = 8‐9 per group).
   G) Echocardiography showing representative images of PAAT, RVWT and
   TAPSE. H) Statistical analysis of PAAT, RVWT and TAPSE (n = 4‐8 per
   group). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Two‐way
   ANOVA was performed, followed by Bonferroni correction for six pairwise
   comparisons among the four experimental groups.

2.5. The S100A9 Inhibitor Paquinimod Attenuates Symptoms of Mice with
SuHx‐Induced PH

   To further clarify the role of S100A9 in mice with SuHx‐induced PH, an
   S100A9 inhibitor Paquinimod (ABR‐25757) was used to examine its effects
   on disease progression. ABR‐25757, a small‐molecule drug, has primarily
   been studied for its potential immunomodulatory effects, one of which
   is S100A9 inhibition effect.^[ [113]^24 ^] The inhibitor ABR‐25757 was
   first given three days before establishing the SuHx‐induced PH model
   and continuatively given daily via intraperitoneal injection (10
   mg·kg^−1/per dose) throughout the experiment^[ [114]^25 , [115]^26 ^]
   (Figure [116]5A). Results showed that administration of ABR‐25757
   effectively reduced SuHx‐induced increases in RVSP, RVHI and RV/BW
   (Figure [117]5B) compared to those of controls. Immunostaining
   indicated that vascular thickening was less pronounced in sections of
   the lung tissues (Figure [118]5C), while cardiomyocytes hypertrophy of
   SuHx‐induced PH mice treated with ABR‐25757 was also improved
   (Figure [119]5D,E). Additionally, ABR‐25757 treatment also
   significantly reduced RV wall thickness, with marked improvement in RV
   function parameters, including PAAT and TAPSE (Figure [120]5F,G).

Figure 5.

   Figure 5
   [121]Open in a new tab

   The S100A9 inhibitor Paquinimod attenuates symptoms of mice with
   SuHx‐induced PH. A) Schematic diagram of the establishment of
   experimental murine model. B‐H: ABR‐25757(i.p.) was administered to
   C57BL/6 mice under control and SuHx‐induced conditions. B) RVSP, RVHI
   and RV/BW of mice (n = 6‐7 per group). C) Representative images of H&E
   and α‐SMA/CD31 IF staining in lung sections of the experimental mice,
   as well as representative images of WGA staining showing cardiomyocyte
   hypertrophy in the RV. D) Quantification of wall thickness of the
   pulmonary vasculature (n = 8 per group). E) Quantification of WGA
   staining (n = 7‐9 per group). F) Echocardiography showing the
   representative images of PAAT, RVWT and TAPSE. G) Statistical analysis
   of PAAT, RVWT and TAPSE, per group (n = 4‐7 per group). *p < 0.05, **p
   < 0.01, ***p < 0.001, ****p < 0.0001. Two‐way ANOVA was performed,
   followed by Bonferroni correction for six pairwise comparisons among
   the four experimental groups.

2.6. Hypoxia‐Induced Neutrophil‐Endothelial Interaction Promotes Endothelial
Dysfunction via S100A9‐Mediated Activation of the RAGE‐PI3K‐AKT Pathway

   As we demonstrated that neutrophils and the secreted S100A9 play a
   critical role in PH progression, while neutrophils primarily localized
   around CD31^+ ECs in lung tissue. To further investigate which
   structural cells interacted more closely with neutrophils in the lungs,
   we used the ‘CellChat’ R package to analyze differential interactions
   number across all cell types. The data showed that from an interaction
   strength perspective, neutrophils seemed to have potential effects on
   ECs, and the number of interactions between neutrophils and ECs also
   significantly increased (Figure [122]6A; Figure [123]S4A, Supporting
   Information). S100A9 primarily signals through its receptors RAGE and
   TLR4. To investigate the expression patterns of these receptors, we
   first analyzed scRNA‐seq data from the lung tissues. TLR4 expression
   was undetectable at the transcriptomic level. Furthermore,
   protein‐level analysis following exogenous stimulation with S100A9
   showed no significant induction of TLR4 expression in endothelial cells
   (Figure [124]S4B, Supporting Information). In contrast, RAGE (encoded
   by the Ager gene) was found to be more highly expressed in endothelial
   cells than that of smooth muscle cells (Figure [125]S4C, Supporting
   Information). Notably, Ager expression was significantly upregulated in
   endothelial cells of the SuHx‐induced PH mice compared to those of the
   control mice (Figure [126]S4D, Supporting Information). These findings
   suggest that RAGE, rather than TLR4, may serve as the predominant
   receptor mediating S100A9 signaling in endothelial cells during the PH
   pathogenesis.

Figure 6.

   Figure 6
   [127]Open in a new tab

   Hypoxia‐induced neutrophil‐endothelial interaction promotes endothelial
   dysfunction via S100A9‐mediated activation of the RAGE‐PI3K‐AKT
   pathway. A) Number and strength of cellular interactions were analyzed
   using ‘CellChat’, with results shown in a circle plot. B)
   Representative images of the neutrophil adhesive assay. The red arrows
   point at the PKH26‐labelled neutrophil (red). C) Quantification of the
   neutrophil adhesive assay (n = 4 per group). D) Representative images
   of adhesion and angiogenesis of hPAESs co‐cultured with neutrophil
   under hypoxic or normoxic conditions, in the presence or the absence of
   ABR‐25757. E) Quantification of adhesive hPAECs (n = 4 per group). F)
   Quantification of angiogenesis of hPAECs (n = 3 per group). G) KEGG
   pathway enrichment analysis of mouse lung endothelial cells. H)
   Representative Western blot of hPAECs co‐cultured with neutrophils
   under normoxic or hypoxic conditions, in the presence or the absence of
   ABR‐25757, showing protein expression of the RAGE/PI3K/AKT pathways in
   hPAECs (n = 6 per group). I) Representative images and quantification
   of adhesion and angiogenesis of hPAECs evaluated after stimulation with
   rhS100A9 or treated with ABR‐25757 (n = 3‐5 per group). J)
   Representative Western blot and quantification of hPAECs induced by
   rhS00A9 or treated with ABR‐25757, focusing on the RAGE/PI3K/AKT
   pathways (n = 6 per group). ^ns p > 0.05, *p < 0.05, **p < 0.01, ***p <
   0.001, ****p < 0.0001. One‐way ANOVA was performed, followed by
   Bonferroni post hoc correction for three pairwise comparisons among the
   three groups (C, I, and J) and the six pairwise comparisons among the
   four experimental groups (E, F, and H).

   Using a neutrophil adhesion assay, we found that hypoxic conditions
   increased neutrophil adhesion to ECs, which was reduced by S100A9
   inhibitor ABR‐25757 (Figure [128]6B,C). To test the hypothesis that
   hypoxia enhances neutrophil‐endothelial cell interactions and affects
   endothelial function, we co‐cultured neutrophils isolated from healthy
   human peripheral blood with hPAECs. The results showed a marked
   increase in endothelial adhesion and angiogenic capacity when
   co‐cultured with neutrophils in a transwell system under either hypoxic
   condition (Figure [129]6D–F) or stimulated with supernatants collected
   from hypoxia‐exposed neutrophils (Figure [130]S4F, Supporting
   Information), all of which, again, were attenuated by ABR‐25757.

   Next, to further investigate the mechanisms underlying S100A9‐mediated
   endothelial dysfunction, we extracted endothelial cell populations from
   the scRNA‐seq dataset and conducted KEGG and GSEA pathway enrichment
   analyses. The results revealed significant alterations in the PI3K‐AKT
   and AGE‐RAGE signaling pathways (Figure [131]6G; Figure [132]S4E,
   Supporting Information), in which the RAGE‐PI3K‐AKT pathway was
   activated in ECs, while ABR‐25757 significantly attenuated these
   effects (Figure [133]6H; Figure [134]S4G, Supporting Information).

   CCK8 assay showed that treated hPAECs with recombinant human S100A9
   protein (rhS100A9, 10 µg mL^−1) significantly promoted cell
   proliferation (Figure [135]S4H, Supporting Information) Following this
   assay, the concentration of 10 µg mL^−1 was used in subsequent
   experiments, showing that rhS100A9 significantly enhanced endothelial
   adhesion and angiogenesis (Figure [136]6I), along with activation of
   the corresponding signaling pathways. Again, these effects were
   reversed by adding the inhibitor (Figure [137]6J).

   To further confirm the critical role of the RAGE–PI3K–AKT pathways in
   mediating endothelial dysfunction under high S100A9 conditions, we
   replaced the S100A9 inhibitor ABR‐25757 with a PI3K inhibitor
   (LY294002) and repeated the aforementioned endothelial functional
   assays. The results demonstrated that LY294002 also effectively
   ameliorated endothelial dysfunction induced by elevated S100A9 levels
   or neutrophil exposure (Figure [138]S5A–C, Supporting Information).

   Taken together, these data suggest that hypoxia enhances
   neutrophil‐endothelial cell interactions, leading to endothelial
   dysfunction and subsequent vascular remodeling, in which
   neutrophil‐derived S100A9 is likely involved though activating the
   RAGE‐PI3K‐AKT signaling pathways of ECs.

2.7. S100A9‐RAGE‐PI3K‐AKT Pathways Participate in the Process of SuHx‐Induced
PH

   We next explored whether the expression level or presence of S100A9
   affects the overall expression of the RAGE‐PI3K‐AKT pathways in vivo.
   We analyzed relevant protein levels in the lung tissues of mice with
   SuHx‐induced PH. To achieve these, the transcriptomic, proteomic, and
   single‐cell transcriptomic analyses were performed. KEGG and GSEA
   enrichment results showed that there were significant changes in
   PI3K‐AKT‐related pathways in mice with SuHx‐induced PH compared to
   those of normal controls across all three omics datasets (Figure
   [139]7A–C).

Figure 7.

   Figure 7
   [140]Open in a new tab

   S100A9‐RAGE‐PI3K‐AKT pathway in murine model of SuHx‐induced PH. A)
   KEGG and GSEA pathway enrichment analysis of mouse lung tissue in
   transcriptome. B) KEGG and GSEA pathway enrichment analysis of mouse
   lung tissue in proteomics. C) KEGG and GSEA pathway enrichment analysis
   of mouse lung tissue in scRNA‐seq. D) Representative Western blot and
   quantification of protein expression in lung tissues of S100A9^− ^/ ^−
   mice, focusing on the RAGE/PI3K/AKT pathway (n = 3 per group). E)
   Representative Western blot and quantification of protein expression in
   lung tissues of mice treated with ABR‐25757, focusing on the
   RAGE/PI3K/AKT pathway (n = 3‐6 per group). F) Representative Western
   blot and quantification of protein expression in lung tissues of mice
   treated with anti‐Ly6G antibody, focusing on the RAGE/PI3K/AKT pathway
   (n = 3 per group). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.
   Two‐way ANOVA was performed, followed by Bonferroni correction for six
   pairwise comparisons among the four experimental groups.

   Next, we analyzed protein levels of the RAGE‐PI3K‐AKT pathways in three
   SuHx‐induced PH animal models, including S100A9 ^−/− mice, mice treated
   with ABR‐25757, and those of neutrophil depletion with anti‐Ly6G
   antibody, using Western blot (Figure [141]7D–F). The results indicated
   that either S100A9 knockout, reduced expression, or inhibition of its
   binding to the RAGE receptor all significantly suppressed SuHx‐induced
   activation of the RAGE‐PI3K‐AKT pathways. These findings suggest that
   S100A9 plays a crucial role in activating such pathway in SuHx‐induced
   PH.

3. Discussion

   To the best of our knowledge, there are lack of studies regarding the
   role of neutrophils‐derived S100A9 in the development of SuHx‐induced
   PH and the potential underlying mechanisms. We first demonstrated that
   the expression of S100A9 increased during the development of
   SuHx‐induced PH, which was primarily from the increased neutrophils.
   Notably, both the depletion of neutrophils and the absence of S100A9
   significantly attenuated symptoms of mice with SuHx‐induced PH.
   Furthermore, our results showed that neutrophils exposed to hypoxia
   impaired endothelial function, an effect that was abolished by
   inhibiting S100A9. Data of our mechanistic experiments indicated that
   S100A9 mediated endothelial dysfunction by activating the RAGE/PI3K/AKT
   pathways, and its upregulation enhanced neutrophil adhesion to ECs,
   thereby promoting the progression of SuHx‐induced PH. Therefore, S100A9
   likely represents a promising anti‐inflammation therapeutic target for
   reversing pulmonary vascular remodeling (Figure [142]8 ).

Figure 8.

   Figure 8
   [143]Open in a new tab

   In the pulmonary vasculature, hypoxia not only increases numbers of
   neutrophils but also enhances the ability of these cells to adhere to
   ECs. These neutrophils release S100A9, which binds to RAGE receptors on
   ECs, and activates the RAGE‐PI3K‐AKT signaling pathway. This process
   further induces endothelial cell proliferation and vascular remodeling,
   contributing to the progression of PH. Thus, it is reasonably to
   speculate for that targeting S100A9 might be beneficial for patients
   with PH.

   Currently, most clinical treatments for PH are vasoconstriction
   inhibitors, targeting the classical pathways of PH pathogenesis, such
   as endothelin‐1 (ET‐1),^[ [144]^27 ^] prostacyclin,^[ [145]^28 ^] and
   nitric oxide pathways.^[ [146]^29 ^] However, these approaches mainly
   focus to inhibit pulmonary vasoconstriction and reduce pulmonary
   vascular resistance but less fundamentally to address vascular
   remodeling. As the role of inflammation in pulmonary vascular
   dysfunction has become increasingly evident, immunotherapy has shown
   great promise for PH treatment.^[ [147]^5 ^] Over recent years,
   multiple strategies targeting immune cells, inflammatory cytokines, and
   related signaling pathways have been developed to effectively treat
   PH.^[ [148]^30 , [149]^31 , [150]^32 ^]

   As known, the altered inflammatory environment surrounding pulmonary
   blood vessels mainly results from immune cell infiltration and the
   production and recruitment of inflammatory factors. Our previous and
   ongoing studies have also confirmed that some cytokines, such as
   IL‐17^[ [151]^11 ^] and IL‐33,^[ [152]^13 ^] as well as immune cells,
   including T lymphocytes^[ [153]^12 ^] and macrophages, are potential
   immune factors contributing to pulmonary vascular remodeling in PH.
   Furthermore, numerous studies have demonstrated elevated levels of
   circulating immune cells in PH patients and animal models, with a
   significant increase in neutrophil counts,^[ [154]^33 , [155]^34 ^]
   however, the specific role of neutrophils in PH has yet to be fully
   elucidated. First, we observed an increased number of neutrophils in
   the peripheral blood and the lung tissues of mice with SuHx‐induced PH,
   as well as these cells aggregated around the blood vessels within the
   lung tissues of patients with PH. These observations are consistent
   with clinical findings showing an inverse relationship between
   neutrophil ratio and PH patient prognosis.^[ [156]^35 ^]

   Recent findings have revealed that ECs participate in a variety of
   biological processes. As the outermost layer of the vascular intima,
   ECs are the first to respond to stimuli and subsequently transmit
   activation signals to the outer layers, including smooth muscle cells,
   collectively mediating the formation of pulmonary vascular
   remodeling.^[ [157]^7 ^] Second, in vitro experiments also confirmed
   that, following hypoxia, neutrophils exhibited functional impairment,
   which in turn induces endothelial cell dysfunction. Therefore, it is
   evident that endothelial dysfunction may serve as one of the critical
   initiating factors in pulmonary vascular remodeling. These suggest that
   neutrophils likely play an important role in the progression of PH,
   either through promoting an inflammatory phenotype or directly
   contributing to vascular impairment, or both.

   Thirdly, we found that there was the increase of S100A9 in the lung
   tissue of COPD‐PH patients. Such increased S100A9 was also confirmed in
   the lung tissues of numerous established PH models, including the
   MCT‐induced model and the chronic hypoxia and hypoxia combined with
   Sugen‐induced animal models. Furthermore, neutrophils are the major
   cellular source for S100A9 in clinical samples and our SuHx‐induced
   models. In recent years, the potential role of S100A9 in PH has
   received increasing attention. Taz et al. reported, through comparative
   genomic analyses, that SARS‐CoV‐2 infection upregulates the expression
   of the inflammation‐related protein S100A9, thereby exacerbating the
   inflammatory response and vascular injury, ultimately worsening the
   condition of PH patients. These findings suggest that S100A9 may serve
   as a potential therapeutic target in PH, and that inhibiting its
   expression or activity may alleviate SARS‐CoV‐2 infection‐associated PH
   deterioration.^[ [158]^36 ^] However, the precise molecular mechanisms
   of S100A9 in PH, as well as its predominant cellular source, remain
   incompletely understood. Our data suggest that neutrophil‐derived
   S100A9 might be a critical factor in the pathogenesis of PH, at least
   in mice with the SuHx‐induced PH. At the molecular level, S100A8 and
   S100A9, as key members of the S100 protein family, are typically
   co‐expressed in neutrophils and monocytes in the form of a heterodimer
   (S100A8/A9, also known as calprotectin).^[ [159]^37 ^] As
   damage‐associated molecular patterns (DAMPs), they are capable of
   amplifying immune responses.^[ [160]^15 , [161]^22 ^] For instance, Wu
   et al. demonstrated that S100A8/A9 released from CD11b⁺Gr1⁺ neutrophils
   activate cardiac fibroblasts, thereby promoting angiotensin II‐induced
   cardiac inflammation and injury.^[ [162]^38 ^] In our experiments, we
   also found that exogenous administration of the S100A8/A9 heterodimer
   similarly enhances endothelial cell functions, including proliferation,
   angiogenesis, and cell adhesion, with S100A9 likely playing a key role
   in these processes (Figure [163]S6A,B, Supporting Information).
   Obviously, understanding the similarities and differences in the
   mechanisms of action between S100A9 alone and the S100A8/A9
   heterodimer, as well as their interrelationship, will be one of the key
   focuses of any future relevant studies. Although both S100A8 and S100A9
   are involved in the regulation of inflammation, their temporal roles
   and functional intensities might differ in the phases of inflammation,
   i.e., S100A8 primarily acts in the early phase of inflammation, whereas
   S100A9 is more responsible for sustaining and amplifying the
   inflammatory response.^[ [164]^39 ^]

   Finally, we demonstrated that targeting S100A9 or neutrophils either
   reversed S100A9‐induced changes in ECs in vitro or in vivo attenuated
   symptoms of mice with PH. While the S100A8/A9 complex is known to be an
   effective ligand for receptors such as RAGE and TLR4, the monomeric
   form of S100A9 has a significantly higher affinity for these receptors
   compared to the heterodimer.^[ [165]^40 ^] It has been reported that
   through interactions with RAGE or TLR4, S100A9 induces neutrophil
   migration and adhesion to ECs,^[ [166]^41 ^] as well as promotes the
   formation of a large number of neutrophil extracellular traps (NETs).^[
   [167]^42 ^] On the other hand, the S100A9‐RAGE axis also induces
   endothelial dysfunction via F‐actin, ZO‐1, and occluding.^[ [168]^43 ^]
   Significantly, the PI3K‐AKT signaling pathway plays a crucial role in
   regulating these pathological processes mediated by S100A9. Tang et al.
   demonstrated that the deletion of AKT1 conferred protection against the
   development and progression of chronic hypoxic pulmonary hypertension,
   and played a dominant role in pulmonary vascular remodeling.^[ [169]^44
   ^] Specifically, AKT1 might be involved in glycolytic shift and
   metabolic reprogramming, as well as endothelial dysfunction mediated by
   various factors including PHD2 and AIP1.^[ [170]^45 , [171]^46 ^] In
   the present study, we mainly focused on the activation of the PI3K/AKT
   pathway by S100A9 via its receptor RAGE, and its role in modulating
   endothelial cell function. However, PI3K/AKT may not be the only
   downstream signaling pathways involved in S100A9‐RAGE‐mediated
   endothelial dysfunction. Previous studies have shown that RAGE, as a
   multi‐ligand receptor, can interact with various signaling molecules
   and activate multiple downstream pathways, including MAPK, NF‐κB, and
   JAK/STAT, which are implicated in inflammation, apoptosis, and barrier
   regulation.^[ [172]^47 , [173]^48 ^] Therefore, it is reasonable to
   speculate that S100A9‐RAGE also acts through these non‐PI3K/AKT
   pathways to form a broader signaling network that co‐regulates
   endothelial cell proliferation, migration, and adhesion. In future
   studies, it is worth to further explore whether S100A9‐RAGE interacts
   with other signaling molecules using molecular intervention approaches,
   such as specific inhibitors or RAGE mutant constructs, to
   comprehensively elucidate the mechanisms underlying endothelial
   dysfunction. Consistent with these, our data showed that
   neutrophil‐derived S100A9 causes endothelial dysfunction through
   RAGE/PI3K/AKT pathways. Notably, both co‐culture of ECs with
   neutrophils and/or with exogenous rhS100A9 activated RAGE expression.
   However, exogenous rhS100A9 did not significantly activate TLR4 (Figure
   [174]S4, Supporting Information), another receptor of S100A9,
   suggesting that RAGE may play a more dominant role. Moreover, these
   pathological changes were alleviated following treatment with the
   S100A9 inhibitor Paquinimod.

   Taken together, S100A9 appears to be a crucial inflammatory molecule
   during the process of pulmonary vascular remodeling, suggesting that
   S100A9 likely a potential molecule for therapeutic target in PH. In the
   development of drugs targeting S100A9, Paquinimod, which blocked its
   interaction with RAGE and TLR4, has already reached Phase II clinical
   trials for the treatment of systemic sclerosis, demonstrating promising
   efficacy and safety.^[ [175]^24 ^] Our study further sheds likely light
   on the development of novel treatment strategy for PH. However, the
   present study also has certain limitations. One notable limitation is
   the lack of sex‐specific analyses, which is particularly important
   given that estrogen‐mediated autophagy attenuation may contribute to
   improved survival and better right ventricular function in female PH
   patients.^[ [176]^49 ^] Second, while we have demonstrated that
   neutrophils mediate endothelial dysfunction through the secretion of
   S100A9, whether this process further drives abnormal proliferation and
   dysfunction of smooth muscle cells remains unclear. Finally, although
   both the S100A9 knockout model and the pharmacological inhibition model
   affect S100A9 expression and function at a systemic level, however, due
   to technical limitations and the intrinsic characteristics of
   neutrophils, we were unable to further investigate the specific effects
   of targeting S100A9 in neutrophils. Investigating this mechanism will
   be a key focus of our future research.

   In conclusion, our data show that targeting the
   neutrophil‐S100A9‐RAGE‐endothelial cell axis could represent a novel
   therapeutic approach for managing PH, potentially improving outcomes
   for patients suffering from the disease. In addition, considering the
   crucial role of neutrophils in pulmonary vascular remodeling, our
   findings also highlight the importance of controlling and preventing
   infections as a key strategy in clinical practice to hinder the
   progression of PH.

4. Experimental Section

Ethical Statements

   The present study was approved by the Ethics Committee of Capital
   Medical University and conducted in accordance with the principles of
   the Declaration of Helsinki. Informed consent was obtained from all
   patients or their relatives involved in lung tissue collection.

   All animal experiments were approved by the Animal Review Committee of
   Capital Medical University (Approval Numbers: AEEI‐2020‐100,
   AEEI‐2021‐156) and conducted in strict accordance with the Regulations
   for the Administration of Affairs Concerning Experimental Animals. All
   procedures also conformed to the guidelines of Directive 2010/63/EU of
   the European Parliament on the protection of animals used for
   scientific purposes. The study adhered to the 3R principles
   (Replacement, Reduction, and Refinement) to ensure the ethical and
   scientific integrity of animal research.

Human Lung Samples

   Lung tissues were, either via lung transplantation or lobectomy,
   collected from donors, patients with chronic obstructive pulmonary
   disease (COPD) only, and those with COPD‐PH, fixed in 10% formaldehyde,
   and then embedded in paraffin. The patients’ clinical data can be seen
   in our previous published paper^[ [177]^50 ^] and Table [178]S1
   (Supporting Information).

Animal Models

   C57BL/6 mice and Sprague Dawley rats were purchased from Beijing Vital
   River Laboratory Animal Technology Co., Ltd. Global S100A9 knockout
   (KO) mice (S100A9^−/− mice, C57BL/6 background,
   KOCMP‐20202‐S100A9‐B6N‐VA) were kindly provided by Professor Huihua Li
   from Beijing Chao‐Yang hospital.^[ [179]^51 , [180]^52 ^] Animals were
   randomly assigned to various treatment groups for each experiment. Mice
   and rats were maintained on a 12 h light‐dark cycle with a regular
   unrestricted diet. To establish a Sugen‐hypoxia‐induced PH mouse model
   (SuHx), 8‐10‐week‐old male mice received weekly subcutaneous injections
   of SU5416 (20 mg·kg^−1/per does; Selleck, catalog no. S2845, TX) for 4
   weeks and were housed in an airtight plexiglass chamber with 10%
   O[2].^[ [181]^53 ^] To establish a chronic‐hypoxia‐induced PH mouse
   model, 8‐10‐week‐old male mice were housed in an airtight plexiglass
   chamber with 10% O[2] for 4 weeks. To generate monocrotaline
   (MCT)‐induced PH, male Sprague Dawley rats (≈200–250 g) received a
   single intraperitoneal injection of MCT (60 mg·kg^−1/per dose; MCE,
   catalog no. HY‐N075, NJ) followed by a 4‐week observation period. For
   the murine model with neutrophil depletion, isotype IgG (Selleck,
   catalog no. A2123, TX) or anti‐Ly6G‐antibody (Selleck, A2158, TX, 200
   µg per dose) was injected intraperitoneally 24 h prior to the assay,
   followed by injections alternate days. In the Paquinimod group, the
   Paquinimod (Selleck, catalog no. S9963, TX) was given 2 days prior to
   the start of the assay (10 mg·kg^−1/per dose), and the next consecutive
   4 weeks, while their control group was given equal amount of 1 × PBS
   injection.

   All animals were anesthetized with isoflurane (3% for induction, 1.5%
   for maintenance), administered via inhalation using a calibrated
   vaporizer. For terminal procedures, an overdose of sodium pentobarbital
   (150 mg kg^−1) was administered by intraperitoneal injection. All
   efforts were made to minimize the suffering of the animals.

ScRNA‐Seq and Analysis

   To obtain single‐cell RNA‐seq (scRNA‐seq) data, lung tissues of the
   experimental mice were harvested and analyzed using 10x Genomics
   Chromium single‐cell 3’v2 and a V(D)J kit (10×Genomics) according to
   the manufacturer's instructions. Libraries were sequenced at 150
   paired‐end cycles using the HiSeq3000 system (Illumina). The
   differentially expressed genes (DEGs) of each cluster were identified
   using “FindAllMarkers” function, and the clusters were annotated based
   on classic marker genes. Kyoto Encyclopaedia of Genes and Genomes
   (KEGG) analysis was performed by “clusterProfiler” R package.
   “CellChat” package was used to evaluate the interactions between cells.
   Detailed information regarding the R packages and analytical procedures
   is provided in the Table [182]S3 (Supporting Information).

   Cell–cell communication was inferred using the R package CellChat
   (v1.6.1). The Seurat object containing the annotated scRNA‐seq data was
   converted to a CellChat object. Genes were filtered to retain only
   those expressed in more than 10% of cells within each cell group. The
   built‐in database CellChatDB.mouse was used to identify known
   ligand‐receptor pairs. Overexpressed genes and interactions were
   identified using default thresholds, followed by the computation of
   communication probabilities based on a mass action model. Cell‐cell
   communication networks were visualized using heatmaps, circle plots.

Transcriptomic and Proteomic Data Analysis and Pathway Enrichment

   Lung tissues were collected from the experimental mice, total RNAs were
   extracted for transcriptomic analysis, while proteins were isolated
   separately for proteomic analysis. Transcriptomic analysis was
   performed using RNA sequencing, while proteomic analysis was conducted
   through mass spectrometry for protein identification and
   quantification.

Hemodynamic and Ventricular Weight Measurements

   The right ventricular systolic pressure (RVSP) was measured through
   closed‐chest insertion of the experimental mice. Briefly, after
   anesthesia, a 22‐gauge needle connected to a pressure transducer
   interfaced with a PowerLab system (AD Instruments, Australia) was
   inserted into the RVs of anesthetized animals through a xiphocostal
   angle approach, and the waveform was used to confirm the position of
   the needle. Prior to RVSP measurement, the RV wall thickness (RVWT),
   tricuspid annular plane systolic excursion (TAPSE), and pulmonary
   artery acceleration time (PAAT) were measured using a Vevo 2100 Imaging
   system (FUJIFILM VisualSonics, lnc., SC). The results were calculated
   using Visual Sonics Vevo 2100 analysis software (v.1.6) with a cardiac
   measurement package, based on the average of three cardiac cycles.

Histological Analyses

   Animals were anesthetized with sodium pentobarbital, sacrificed after
   hemodynamic measurements, and perfused with chilled saline. Right lung
   and heart tissues were fixed with 10% formaldehyde at 25 cm H[2]O
   pressure and processed for histological and immunostaining analyses.
   For H&E staining, 4‐µm lung sections were analyzed for medial thickness
   (MT%) and medial area of pulmonary vessels (≤100 µm external diameter)
   using a Nikon microscope and image analysis software. For IHC staining,
   slides underwent antigen retrieval (citric acid buffer, pH 6.0),
   blocking, and incubation with primary antibodies against neutrophil
   elastase (ab68672, 1:2500, Abcam) and S100A9 (#73425, 1:200, CST),
   followed by HRP‐conjugated secondary antibodies and DAB development.
   For IF staining, slides were incubated with primary antibodies against
   CD31 (ab182981, 1:200, Abcam), Ly6G (ab25377, 1:200, Abcam), MPO
   (ab20257, 1:200, Abcam), α‐SMA (#BM0002, 1:200, Boster), and S100A9
   (#73425, 1:200, CST), followed by Alexa Fluor‐conjugated secondary
   antibodies and DAPI nuclear staining. For WGA staining, slides were
   treated with 5 mM WGA ([183]W11262, Invitrogen) and counterstained with
   DAPI. Quantification of positive staining was performed on nine
   randomly selected images per mouse using Image‐J, with positive cell
   counts normalized to field area (mm^2).

Flow Cytometry

   Neutrophil from blood and lung tissues of mice were analyzed by flow
   cytometry. In brief, single‐cell suspensions of mouse lung tissue were
   prepared by repeated rounds of enzymatic digestion and trituration with
   collagenase type I (Roche, Switzerland), neutral protease and DNase I
   (Sigma, catalog no. 10104159001, CA). Primary antibodies of
   APC‐Cy^TM7‐CD45 (BD, catalog no. 557659, NJ), PE‐Ly6G (Invitrogen,
   catalog no. 12‐5931‐82, CA) and FITC‐CD11b (Invitrogen, catalog no.
   11‐0112‐82, CA) were used at a 1:100 dilution. The results were
   collected by BD FACSymphony A5, and analyzed with FlowJo 10.8.1
   software. More than 10,000 events were collected. Live cells were
   selected in FSC_SSC dot plot. The single cells were chosen in
   FSC‐A_FSC‐H dot plot. Events in the single‐cell gate were analyzed.
   After sorting the live cells and single cells, at least 5,000 events
   were ready for further analysis.

Western Blot Analyses

   Protein extracts were prepared by lung tissue in ice‐cold RIPA buffer
   supplemented with protease and phosphatase inhibitors (Roche, catalog
   no. 04693116001 and 04906837001, Switzerland), and protein
   concentration was determined using the BCA protein assay kit (Thermo
   Fisher Scientific, catalog no. A55860, CA). Equal amounts of protein 20
   µg were resolved by SDS‐PAGE and transferred to NC membranes
   (Millipore, catalog no. HATF00010, MA). Membranes were blocked with a
   rapid blocking buffer (Epizyme Biotech, catalog no. PS108P, China) for
   10–15 min at room temperature and then incubated with primary antibody
   RAGE (Abcam, catalog no. ab216329, 1:1,000 dilution, CA), S100A9 (CST,
   catalog no. 73425, 1:1,000 dilution, CA), PI3K (CST, catalog no. #4249,
   1:1,000 dilution, CA), Phospho‐PI3K (CST, catalog no. #4228, 1:500
   dilution), Akt (CST, catalog no. #9272, 1:1,000 dilution), and
   Phospho‐Akt (CST, catalog no. #9271, 1:500 dilution), membranes were
   incubated with fluorophore‐conjugated secondary antibodies (LI‐COR,
   catalog no. 926–32211 and 926–68020, I:5,000 dilution, CA) for 1 h at
   room temperature. Protein bands were detected using an Odyssey infrared
   imaging system (LI‐COR Biosciences, CA). Band intensities were
   quantified using Image Studio software (LI‐COR Biosciences), and
   relative protein levels were normalized to the loading control.

ELISA Detection of Serum S100A9

   Concentrations of human serum S100A9 were measured using an ELISA kit
   (Biolegend, catalog no. 445307, CA) according to the manufacturer's
   instructions.

Cell Culture and Preparation

   Human pulmonary artery endothelial cells (hPAECs) from normal human
   subjects (ScienCell Research Laboratories, catalog no. 3100, CA) were
   cultured in culture medium (EC Medium, ECM, ScienCell Research
   Laboratories, catalog no. 1001, CA), supplemented with
   Penicillin/Streptomycin (P/S), growth supplements and 5% (v/v) fetal
   bovine serum (FBS) and maintained at 37°C in a humidified normoxia
   condition (21% O[2], 5% CO[2], 74% N[2]) or hypoxia condition (3% O[2],
   5% CO[2], 92% N[2]). Cells were passaged (passages 3–5) after reaching
   80–90% confluence, detached with 0.05% trypsin, 0.04% EDTA
   (Sigma‐Aldrich, catalog no. EDS‐100G, Germany) in phosphate‐buffered
   saline (PBS).

   Human neutrophils were collected from fresh peripheral blood of healthy
   volunteers using EDTA or heparin as an anticoagulant to prevent
   clotting, and separated using Histopaque‐1119 (Sigma‐Aldrich, catalog
   no. 11191, Germany) and Histopaque‐1077 (Sigma‐Aldrich, catalog no.
   10771, Germany). After separation, flow cytometry is used for
   validation.

Tubule Formation and Adhesion Assay of hPAECs

   To detect the tube formation ability of hPAECs, a tube formation assay
   was performed. BD Matrigel (BD Biosciences, catalog no. 356234, NJ) was
   distributed in a 24‐well plate (200 µL/well) on ice and allowed to
   solidify at 37 °C for at least 30 min. After the Matrigel solidified,
   hPAECs (1 × 10^5 per well) were gently added to each of the triplicate
   wells. Tube formations were quantitatively measured by calculating the
   number of tube‐like structures using NIS‐Elements Basic
   Research‐Microscope Imaging Software (Nikon, Japan). Tracks of hPAECs
   organized into networks of cellular cords were counted in 6 random
   fields. Additionally, for the cell adhesion assay, hPAECs were
   trypsinized and seeded on top of fibronectin (30 µL of 20 µg mL^−1;
   Millipore, catalog no. FC010, MA)‐coated 96‐well plates (1×10^4 per
   well in 5% FBS and growth factor–free ECM) and incubated for 30 min at
   37 °C. After washing with PBS, the adhered cells were photographed (4
   pictures per well, 5 wells per group), and the data were analyzed using
   ImageJ software.

Neutrophil Adhesion Assay

   Neutrophils were labeled with PKH26 according to the manufacturer's
   instructions (Sigma‐Aldrich, PCLDIL, Germany). After that, hPAECs were
   then co‐cultured with PKH26‐labeled neutrophils at a concentration of 2
   × 10^5 mL^−1 for 1 h. The culture medium and unattached cells were
   removed by washing twice with PBS. Slides were mounted and images were
   captured to record the PKH26‐labeled neutrophils.^[ [184]^54 ^]

Statistical Analyses

   All statistical analyses were performed using GraphPad Prism version
   10.2.3. Data are presented as means ± standard deviation (SD). For
   comparisons between two groups, an unpaired t‐test was applied. For
   comparisons among three, four or six groups, one‐way analysis of
   variance (ANOVA) was performed, followed by Bonferroni post hoc
   correction applied to three, six or fifteen pairwise comparisons. For
   comparisons among four experimental groups, two‐way ANOVA was
   performed, followed by Bonferroni post hoc correction applied to six
   pairwise comparisons. A P value less than 0.05 was considered
   statistically significant. Randomization and blinded analyses were
   applied whenever possible.

Data Availability

   Sequencing data generated in the present study have been deposited at
   the Genome Sequence Archive ([185]https://ngdc.cncb.ac.cn/gsa) and are
   publicly available at the date of publication. The accession nos. of
   scRNA‐seq, proteomic and RNA‐seq data are CRA020769, OMIX008029 and
   CRA020767, respectively.

Conflict of Interest

   The authors declare no conflict of interest.

Author Contributions

   Y.G. performed investigation, methodology, visualization, and
   wrote‐original draft. Z.G. performed formal analysis, investigation,
   and methodology. R.Z. performed software, formal analysis, data
   curation, and visualization. H.L. performed software, validation, and
   visualization. M.Z. performed formal analysis, and data curation. L.L.,
   Z.A., and Y.S. performed software. Y.C. performed supervision. Y.J.
   performed conceptualization. L.W. performed conceptualization, wrote,
   reviewed, and edited. Y.S. performed conceptualization, supervision,
   project administration, funding acquisition, writote, reviewed, and
   edited. J.L. performed conceptualization, supervision, project
   administration, funding acquisition. W.W. performed conceptualization,
   supervision, project administration, and funding acquisition.

Supporting information

   Supporting Information
   [186]ADVS-12-e04397-s001.docx^ (52.7MB, docx)

Acknowledgements