Mechanistic study of ANXA3-mediated endoplasmic reticulum stress promoting M1 macrophage polarization in pulmonary arterial hypertension based on bioinformatics and nine machine learning algorithms

Background: Endoplasmic reticulum stress (ERS) plays a critical role in inflammation-related disorders, but its immunoregulatory mechanisms in pulmonary arterial hypertension (PAH) remain unclear. This study aimed to identify key ERS-associated genes and clarify their roles in immune modulation during PAH progression.

Methods: Transcriptomic data from three Gene Expression Omnibus (GEO) datasets comprising peripheral blood mononuclear cells (PBMCs) of PAH patients (n = 100) and healthy controls (n = 61) were integrated to screen differentially expressed ERS-related genes. Nine machine learning algorithms were applied, and four candidate genes were identified and validated by quantitative Real-Time PCR (qRT-PCR) in PBMCs from 20 idiopathic PAH (IPAH) patients and in two PAH rat models: monocrotaline (MCT)-induced and SU5416/hypoxia (SuHx)-induced. Immune infiltration analysis, consensus clustering, single-cell RNA Sequencing (scRNA-seq), and immunofluorescence (IF) were used to determine immune-specific localization. Functional experiments were then performed to investigate the role of the key gene in ERS regulation, immune responses, macrophage function, and macrophage-pulmonary artery smooth muscle cell (PASMC) interactions.

Results: Fifty differentially expressed ERS-related genes were identified. Least absolute shrinkage and selection operator (LASSO) regression achieved the best diagnostic performance and identified four key candidates: BCL2 Like 1 (BCL2L1), Tumor Necrosis Factor Receptor Superfamily Member 17 (TNFRSF17), DnaJ Heat Shock Protein Family Member B1 (DNAJB1), and Annexin A3 (AnxA3). qRT-PCR confirmed altered expression of these genes in PBMCs and lung tissues from both PAH models. AnxA3 was specifically overexpressed in PAH-associated macrophages and co-localized with Cluster of Differentiation 68-positive (CD68⁺) macrophages in SuHx rat lungs. Functionally, AnxA3 silencing improved right ventricular function and reduced pulmonary vascular remodeling in vivo, while in vitro knockdown suppressed lipopolysaccharide (LPS)-induced M1 polarization, decreased secretion of proinflammatory cytokines including interleukin-1β (IL-1β), interleukin-6 (IL-6), and interleukin-18 (IL-18), and inhibited ERS markers such as glucose-regulated protein 78 (GRP78), C/EBP homologous protein (CHOP), and activating transcription factor 4 (ATF4). These inhibitory effects were reversed by the ERS agonist thapsigargin (TG) but enhanced by the ERS inhibitor 4-phenylbutyric acid (4-PBA). In addition, AnxA3 knockdown alleviated macrophage-induced PASMC proliferation and migration triggered by inflammatory stimuli.

Conclusion: This study demonstrates that AnxA3 regulates ERS to drive M1 macrophage polarization and inflammation, which subsequently promotes PASMC function and promotes PAH progression. AnxA3 may serve as a novel immunoinflammatory target and potential therapeutic candidate.

ANXA3; Endoplasmic reticulum stress; Inflammatory response; Machine learning; Macrophage polarization; Pulmonary arterial hypertension.