GBP5 Aggravates Acute Lung Injury Via NLRP3 Inflammasome Activation While Inducing a HIF-1α–CD73–adenosine Feedback Loop

Abstract

Guanylate-binding protein 5 (GBP5) is an interferon-inducible GTPase that promotes NLRP3 inflammasome activation. However, its role in acute lung injury (ALI) and its relationship with compensatory anti-inflammatory pathways remain poorly defined. Methods: We employed an LPS-induced ALI mouse model with AAV-mediated intratracheal GBP5 knockdown, transcriptomic profiling (RNA-seq), and in vitro studies in THP-1 macrophages. GBP5 overexpression and CD73 overexpression were used to dissect the GBP5–HIF-1α–CD73–adenosine axis. RNA-seq analysis of LPS-induced lung injury revealed upregulation of GBP family members (particularly GBP5) and multiple adenosine metabolism genes. AAV-mediated GBP5 knockdown in male C57BL/6 mice (6–8 weeks, n = 10/group) attenuated LPS-induced lung injury, reduced NLRP3 inflammasome activation (NLRP3, ASC, cleaved caspase-1), and decreased inflammatory cytokine levels (IL-1β, IL-6, TNF-α). In parallel, GBP5 knockdown suppressed CD73 and ADORA2A expression, whereas GBP5 overexpression in THP-1 macrophages enhanced CD73 expression via HIF-1α-dependent transcriptional activation confirmed by dual-luciferase reporter assays. CD73 overexpression in turn elevated cAMP/p-CREB signaling and suppressed NLRP3 inflammasome activity. GBP5 exacerbates ALI through NLRP3 inflammasome activation while simultaneously driving a compensatory HIF-1α–CD73–adenosine–cAMP–CREB feedback loop that restrains excessive inflammation. Targeting this dual pathway may offer novel therapeutic strategies for ALI and ARDS.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10753-026-02528-0.

Introduction

Acute lung injury (ALI) and its severe form, acute respiratory distress syndrome (ARDS), are characterized by overwhelming pulmonary inflammation, increased vascular permeability, and diffuse alveolar damage, that can progress rapidly to life-threatening respiratory failure [1, 2]. These syndromes most commonly arise from severe infection, trauma, or sepsis and are associated with high mortality; the COVID-19 pandemic has further compounded their global burden [3]. Despite considerable advances in understanding ALI/ARDS pathophysiology, effective targeted therapies remain elusive, largely because multiple inflammatory signaling networks act in concert to drive injury.

Guanylate-binding protein 5 (GBP5), an interferon-inducible GTPase, has recently been identified as a critical regulator of NLRP3 activation. GBP5 has been shown to facilitate inflammasome assembly [4], and upon activation, the NOD-like receptor family pyrin domain-containing 3 (NLRP3) inflammasome drives the maturation and release of pro-inflammatory cytokines IL-1β and IL-18, contributing to cytokine storm and tissue injury [5].GBP5 plays important regulatory roles in liver injury, skin inflammation, and phlebitis [6–8]. Our previous research has also found that GBP5 can exacerbate pulmonary fibrosis [9]. Nevertheless, how GBP5 interacts with endogenous counter-regulatory mechanisms in the setting of acute lung injury has not been investigated.

In parallel, adenosine is a potent immunomodulatory nucleoside generated from AMP by CD73 (ecto-5’-nucleotidase, NT5E). CD73-derived extracellular adenosine exerts its anti-inflammatory effects through engagement of the adenosine receptors, activating cAMP-dependent signaling cascades [10–12]. Acting through A2A and A2B receptors, extracellular adenosine dampens innate immune activation by elevating intracellular cAMP and suppressing pro-inflammatory cytokine production in macrophages and other immune cells, thereby promoting the resolution of acute inflammation and restoring tissue homeostasis [13, 14]. Notably, CD73 is highly responsive to inflammatory stimuli, yet the molecular mechanisms regulating its expression in macrophages during ALI are not fully elucidated.

We postulated that GBP5 serves a dual role in ALI by not only enhancing NLRP3 inflammasome activity but also transcriptionally upregulating CD73 expression via inflammatory transcription factors such as HIF-1α. In this model, GBP5 exacerbates inflammation through inflammasome activation but also triggers a compensatory increase in CD73-adenosine signaling, forming a negative feedback loop. Understanding this duality may reveal novel regulatory nodes in lung inflammation and identify therapeutic targets that balance immune activation and resolution.

To investigate this hypothesis, we utilized a combination of in vivo and in vitro approaches, including an LPS-induced ALI mouse model, transcriptomic profiling, adeno-associated virus(AAV)-mediated gene knockdown, and a suite of molecular and immunological assays in THP-1 macrophages. Our study provides mechanistic insights into the GBP5–HIF-1α–CD73–adenosine–cAMP–CREB axis and its role in modulating pulmonary immune responses.

Materials and Methods

Animals and Experimental Design

Male C57BL/6 mice (6–8 weeks old, 18–23 g) were purchased from Experimental Animal Center of Zhengzhou University. Mice were housed in a specific pathogen-free facility with a 12-h light/dark cycle and ad libitum access to food and water. All procedures were approved by the Ethics Committee of the Experimental Animals of Zhengzhou University (ZZU-LAC20230526[01]), and conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. For transcriptomic analysis, mice were randomly divided into two groups (n = 6 per group) and challenged with intraperitoneal LPS (10 mg/kg, Escherichia coli O55:B5; Servicebio, Wuhan, China) or an equivalent volume of saline, as described previously [15]. For the AAV knockdown study, recombinant AAV vectors (AAV-shGBP5 and AAV-Ctrl; 5 × 10¹¹ vg/mL, 50 µL) were obtained from HanHeng Biotechnology (China). Mice were randomly allocated to four groups (n = 10 per group): AAV-Ctrl + saline, AAV-Ctrl + LPS, AAV-shGBP5 + saline, and AAV-shGBP5 + LPS. Under pentobarbital anesthesia (50 mg/kg, i.p.), mice were administered AAV vectors intratracheally using a mouse intubation kit (Ruiwode Inc., Shenzhen, China). After a 3-week transduction period, LPS (10 mg/kg, i.p.) or saline was then administered, and mice were euthanized by pentobarbital overdose 8 h later for tissue and bronchoalveolar lavage fluid (BALF) [9] collection.

Lung Histology and Injury Score

Lung tissues were fixed in 4% paraformaldehyde overnight, embedded in paraffin, sectioned at 4 μm, and stained with hematoxylin and eosin (H&E). Histological lung injury was assessed by blinded observers using a semi-quantitative scoring system evaluating alveolar edema, hemorrhage, leukocyte infiltration, and alveolar wall thickness (score range 0–4). Quantitative assessment of lung injury was performed using the standardized scoring system established by the American Thoracic Society workshop [16, 17].

Immunofluorescence Staining

Paraffin-embedded lung sections were blocked with 5% BSA and incubated overnight at 4 °C with primary antibodies against CD73 (Proteintech, 12231-1-AP; 1:200) or GBP5 as appropriate, followed by fluorophore-conjugated secondary antibodies. Nuclei were counterstained with DAPI. Images were acquired using a Leica SP8 confocal microscope. For THP-1 cell immunofluorescence, cells were fixed with 4% paraformaldehyde, permeabilized with 0.3% Triton X-100, and processed as above. Mean fluorescence intensity (MFI) of GBP5 (red channel) and CD73 (green channel) was quantified using ImageJ (v1.54, NIH) from at least 3 randomly selected fields per group. A cell-free region within each image was used for background subtraction, and values were normalized to the Ctrl group.

Transcriptome Sequencing, Gene Ontology(GO) and Kyoto Encyclopedia of Genes and Genomes ( KEGG) Enrichment Analysis

Total RNA from mouse lung tissues was extracted using TRIzol reagent (Invitrogen) and assessed for quality with an Agilent 2100 Bioanalyzer (Agilent Technologies). Sequencing libraries were prepared using the NEBNext Ultra RNA Library Prep Kit (NEB) and sequenced on an Illumina NovaSeq 6000 platform to generate 150 bp paired-end reads. Raw reads were filtered for quality, and clean reads were aligned to the mouse reference genome (GRCm38/mm10) using HISAT2. Differentially expressed genes (DEGs) were identified using DESeq2 with |log2 fold change| > 1 and adjusted p-value < 0.05 as thresholds.

To understand the function of DEGs, GO enrichment analysis (https://geneontology.org/) and KEGG pathway analyses (https://www.genome.jp/kegg/) were performed, GO enrichment analysis included biological process (BP), molecular function (MF), and cellular component (CC). KEGG pathway enrichment analysis of DEGs was performed using the ClusterProfiler R package. The p-value was corrected using the Benjamini–Hochberg method to obtain the false discovery rate (FDR). Both GO terms and KEGG pathways with FDR < 0.05 were considered significantly enriched.

Composite Score Calculation and Gene Set Enrichment Analysis (GSEA)

The Composite Score was used to assess the activation of adenosine signaling. Differentially expressed genes (DEGs) related to adenosine signaling, such as CD39 and CD73, were selected. The fold change in expression between the LPS-treated and control groups was calculated for each gene and normalized. The Composite Score for each sample was then determined by summing the normalized fold changes of the selected genes, reflecting the overall activation of the pathway.

GSEA was performed to evaluate the enrichment of predefined gene sets associated with adenosine signaling. Gene expression data were ranked by differential expression, and GSEA software was used to calculate an enrichment score (ES) for each gene set. The ES was normalized (NES) to account for gene set size. Statistical significance was determined using false discovery rate (FDR) and nominal p-values, with gene sets showing an FDR < 0.25 and p-value < 0.05 considered significantly enriched.

RNA Extraction and Quantitative Real-time PCR (RT-qPCR)

Total RNA was extracted from lung tissues or THP-1 cells using TRIzol reagent (Invitrogen). RNA concentration and purity were determined by NanoDrop 2000 (Thermo Fisher). First-strand cDNA synthesis was performed using PrimeScript RT reagent Kit (Takara). qPCR was performed using SYBR Green Master Mix (Applied Biosystems) on a QuantStudio 6 Real-Time PCR System. Expression levels were normalized to GAPDH and calculated using the 2^−ΔΔCt method. Primer sequences are listed in Supplementary Table 1.

Western Blotting

Total proteins were extracted using RIPA lysis buffer containing protease and phosphatase inhibitors (Beyotime, Shanghai, China). Protein concentrations were quantified by BCA assay (Thermo Fisher Scientific). Equal amounts (30 µg) were separated by 10–12% SDS-PAGE and transferred to PVDF membranes. After blocking with 5% non-fat milk, membranes were incubated overnight at 4 °C with the following primary antibodies: anti-GBP5 (Servicebio, Wab313390, 1:1000), anti-NLRP3 (CST, 15101, 1:1000),anti-ASC (CST, 67824, 1:1000),anti-Cleaved-caspase-1 (Abcam, ab179515, 1:1000),anti-CD73 (Servicebio, GB115174-100,1:1000), anti-CD39 (Abcam, ab223842, 1:1000),anti-ADORA2A (Proteintech, 51092-1-AP, 1:1000),anti-p-NF-κB (CST, 3033, 1:1000),anti-p-STAT1 (CST, 9167, 1:1000), anti-STAT1 (CST, 9172; 1:1000); anti-HIF-1α (CST, 36169, 1:1000), anti-p-CREB (Servicebio, GB114322-50,1:1000), anti-CREB (CST, 9197; 1:1000); anti-β-actin (Servicebio, GB13236-1-50,1:2000).After incubation with HRP-conjugated secondary antibodies (1:5000), bands were visualized using ECL reagent (Millipore) and quantified with ImageJ.

Enzyme-linked Immunosorbent Assay (ELISA)

Levels of TNF-α (Elabscience, E-EL-M3063,), IL-1β (Elabscience E-EL-M0037), IL-6 (Elabscience, E-EL-M0044), IL-18 (Elabscience, E-EL-M0730), and adenosine in mouse lung homogenates or THP-1 supernatants were measured using commercial ELISA kits according to the manufacturer’s protocols. Absorbance was measured at 450 nm using a microplate reader (BioTek).

THP-1 Cell Culture and Transfection

Human THP-1 monocytes (ATCC) were maintained in RPMI-1640 (Gibco) with 10% FBS and 1% penicillin-streptomycin. Cells were differentiated into macrophage-like cells with 100 ng/mL PMA (Sigma) for 24 h and rested for 24 h. Cells were transfected with oe-GBP5, oe-CD73, or control plasmids using Lipofectamine 3000 (Thermo Fisher). After 24 h, cells were treated with LPS (1 µg/mL) for an additional 8–24 h depending on experimental endpoints.

Luciferase Reporter Assay

Dual-luciferase reporter assays were performed to assess whether HIF-1α or NF-κB regulates CD73 promoter activity. The CD73 promoter fragment was cloned into pGL3-Basic to generate pGL3-CD73, while pGL3-Basic served as the promoterless control. THP-1-derived macrophages were co-transfected with reporter plasmids, transcription factor overexpression plasmids (HIF-1α or NF-κB/p65) or matched empty vector (EV), together with pRL-TK (Renilla luciferase) as an internal control.

For each transcription factor, four groups were included: (1) pGL3-Basic + EV + pRL-TK, (2) pGL3-Basic + TF overexpression plasmid + pRL-TK, (3) pGL3-CD73 + EV + pRL-TK, and (4) pGL3-CD73 + TF overexpression plasmid + pRL-TK. After 24–48 h, luciferase activities were measured using a Dual-Luciferase Reporter Assay System (Promega). Firefly luciferase activity was normalized to Renilla luciferase activity.

Flow Cytometry

Cells were harvested, washed with PBS, and stained with PE-conjugated anti-CD73 (Servicebio, WPE-FcA65564) for 30 min at 4 °C in the dark. After washing, samples were analyzed using a BD FACSCanto II flow cytometer. Data were processed using FlowJo (v10.8).

Intracellular cAMP Measurement

cAMP levels were quantified using a colorimetric cAMP ELISA Kit (Servicebio, WUCEA003Ge) following the manufacturer’s instructions. Cells were lysed in 0.1 M HCl, and OD was read at 450 nm. cAMP concentrations were calculated from standard curves.

Pharmacological Inhibitor Studies

THP-1 cells overexpressing GBP5 were treated with 10 µM BAY 11-7082 (NF-κB inhibitor, MedChemExpress, HY-13453), 20 µM Fludarabine (STAT1 inhibitor, MedChemExpress, HY-B0069), or 10 µM YC-1 (HIF-1α inhibitor, MedChemExpress, HY-14927) for 1 h before LPS stimulation. CD73 expression levels were assessed after 24 h by RT-qPCR and Western blotting.

Statistical Analysis

Data are presented as mean ± standard deviation (SD). Statistical analyses were performed using GraphPad Prism 9.0. One-way ANOVA with Tukey’s post hoc test was used for comparisons between multiple groups. Student’s t-test was used for pairwise comparisons. A p-value < 0.05 was considered statistically significant.

Results

Transcriptomic Profiling Reveals Coordinately Upregulated GBP5 and Adenosine Signaling in LPS-induced Lung Injury

To characterize the transcriptomic landscape of sepsis-associated lung injury, we performed RNA-seq on lung tissues from LPS-challenged and control mice. LPS induced a large number of DEGs compared with saline controls (Fig. 1A–B). Among the most prominently upregulated transcripts were multiple GBP family members (GBP2, GBP3, GBP4, GBP5, and GBP7), with GBP5 displaying the highest fold change (Fig. 1C), nominating it as the primary GBP isoform of interest in this model. GO analysis revealed enrichment of immune system activation, innate immune response, and cellular response to LPS (Fig. 1D). KEGG pathway analysis identified significant enrichment of the NOD-like receptor, JAK–STAT, and NF-κB signaling pathways (Fig. 1E–F), consistent with the known inflammatory signature of LPS-induced ALI.

Fig. 1.

Transcriptomic profiling identifies increased GBP5 and adenosine-signaling–related genes in LPS-induced lung injury. (A) Volcano plot of differentially expressed genes in lung tissues from LPS-induced ALI mice versus control mice based on RNA-seq. (B) Heatmap of global gene expression changes between LPS and control lung tissues. (C) Heatmap showing the expression of the GBP family in control and LPS-treated lungs, highlighting GBP5 upregulation. (D) Gene Ontology (GO) enrichment analysis of differentially expressed genes. (E) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis of differentially expressed genes. (F) Chord diagram illustrating the relationships between selected differentially expressed genes and enriched GO/KEGG pathways. n = 6 per group.

To explore whether metabolic reprogramming accompanies the inflammatory response, KEGG Level 2 analysis revealed prominent enrichment of both immune/signaling and metabolism-related pathways (Fig. 2A–C). GSEA demonstrated significant enrichment of the NOD-like receptor signaling gene set (enrichment score = 0.6; Fig. 2B). Further examination identified eight upregulated adenosine metabolism-related genes, including CD39 and CD73, and composite score analysis confirmed overall activation of the adenosine signaling pathway in LPS-injured lungs (Fig. 2D–E). Together, these data provide unbiased transcriptomic evidence that GBP5 upregulation co-occurs with activation of the CD73–adenosine axis during ALI, motivating direct investigation of a functional link between these two systems.

Fig. 2.

Further analysis of these differentially expressed genes revealed increased expression of genes related to NOD signaling and adenosine metabolism. (A) Heatmap of NOD-like receptor signaling pathway–related genes in lung tissues from LPS-induced ALI mice versus control mice based on RNA-seq. (B) Gene set enrichment analysis (GSEA) running enrichment plots for the NOD-like receptor signaling pathway–related gene set. (C) Distribution of upregulated and downregulated differentially expressed genes (DEGs) across KEGG Level 2 pathways.The x-axis shows the percentage ratio (%) of upregulated/downregulated DEGs annotated to each Level 2 pathway to the total number of upregulated/downregulated DEGs annotated to any KEGG pathway. The y-axis indicates the names of Level 2 pathways. Numbers displayed to the right of each bar represent the counts of upregulated/downregulated DEGs annotated to the corresponding Level 2 pathway. (D) Bar graph showing differential expression of adenosine-signaling–related genes between control and LPS groups. (E) Composite score analysis of adenosine-signaling–related genes identified by RNA-seq. n = 6 per group.

GBP5 Knockdown Reduces LPS-induced Pulmonary Inflammation in vivo

GBP5 was knocked down in mouse lungs via intratracheal AAV delivery, and animals were allocated to four groups: AAV-Ctrl + saline, AAV-Ctrl + LPS, AAV-shGBP5 + saline, and AAV-shGBP5 + LPS. Western blot and RT-qPCR confirmed efficient GBP5 knockdown (Fig. 3A-C). Because the AAV vector carries an intrinsic fluorescent reporter, fluorescence microscopy was performed on lung tissue sections from the four groups of mice to further confirm AAV vector expression (Fig. 3D). Compared with the AAV-Ctrl + saline group, the AAV-Ctrl + LPS group exhibited pronounced lung injury, as evidenced by hematoxylin and eosin (H&E) staining (Fig. 4A) and an increased lung injury score (Fig. 4B). In parallel, the lung wet-to-dry (W/D) ratio (Fig. 4C) was significantly elevated, and the percentage of neutrophils in BALF (Fig. 4D) was increased. Consistently, ELISA revealed markedly higher levels of pro-inflammatory cytokines, including IL-1β (Fig. 4E), IL-6(Fig. 4F), and TNF-α (Fig. 4G), confirming successful establishment of LPS-induced lung injury. Notably, there were no significant differences between the AAV-Ctrl + saline and AAV-shGBP5 + saline groups, indicating that GBP5 knockdown alone did not induce overt pulmonary inflammation under basal conditions. Importantly, compared with AAV-Ctrl + LPS, mice in the AAV-shGBP5 + LPS group displayed substantially attenuated lung injury. Histopathological examination demonstrated improved alveolar architecture with reduced inflammatory infiltrates and hemorrhage (Fig. 4A), accompanied by decreased lung injury scores(Fig. 4B) and a lower W/D ratio (Fig. 4C). Moreover, the BALF neutrophil percentage (Fig. 4D) was reduced, and lung homogenates levels of IL-1β (Fig. 4E), IL-6 (Fig. 4F), and TNF-α (Fig. 4G) were significantly decreased in the AAV-shGBP5 + LPS group.

Fig. 3.

Intratracheal AAV delivery efficiently knocks down GBP5 in mouse lungs. (A) Representative Western blot of GBP5 protein in lung tissues from four groups: AAV-Ctrl + saline, AAV-Ctrl + LPS, AAV-shGBP5 + saline, and AAV-shGBP5 + LPS. (B) Densitometric analysis of lung GBP5 protein expression in the four groups. (C) Relative Gbp5 mRNA expression in lung tissues of the four groups measured by qPCR (2^-ΔΔCt). (D) Immunofluorescence staining of GBP5 in paraffin-embedded lung sections from the four groups after AAV injection (blue, DAPI; red, GBP5; 40× magnification). n = 10 per group.Data are presented as mean ± SD. Statistical analysis was performed as described in the Methods. ns, not significant; *P < 0.05, **P < 0.01,***P < 0.001

Fig. 4.

AAV-mediated GBP5 knockdown attenuates LPS-induced acute lung injury. (A) Representative H&E staining of lung sections from the four groups (AAV-Ctrl + saline, AAV-Ctrl + LPS, AAV-shGBP5 + saline, AAV-shGBP5 + LPS) at 20× and 40× magnification. (B) Lung injury scores based on histopathological assessment of H&E-stained sections. (C) Lung wet-to-dry weight ratio in the four groups. (D) The percentage of neutrophils of bronchoalveolar lavage fluid (BALF) in the four groups. (E) Pulmonary IL-1β levels measured by ELISA in lung homogenates. (F) Pulmonary IL-6 levels measured by ELISA. (G) Pulmonary TNF-α levels measured by ELISA. n = 10 per group.Data are presented as mean ± SD. Statistical analysis was performed as described in the Methods. ns, not significant; *P < 0.05, **P < 0.01,***P < 0.001

GBP5 Promotes NLRP3 Inflammasome Activation and Upregulates the CD73–adenosine Axis in vivo

Western blot analysis of lung tissues showed that, compared with the AAV-Ctrl + saline group, AAV-Ctrl + LPS mice exhibited marked increases in NLRP3, ASC, and cleaved caspase-1 (Fig. 5A-D), indicating activation of the NLRP3 inflammasome after LPS challenge. This effect was significantly attenuated in the AAV-shGBP5 + LPS group, demonstrating that pulmonary GBP5 knockdown suppresses inflammasome activation (Fig. 5A-D).To assess the association between GBP5 and adenosine signaling, we quantified CD39, CD73, and ADORA2A by Western blotting (Fig. 5E-I) and qPCR (Fig. 6A-C). LPS significantly upregulated all three targets in the AAV-Ctrl + LPS group relative to AAV-Ctrl + saline, whereas GBP5 silencing reduced their expression in AAV-shGBP5 + LPS mice. Consistently, CD73 immunofluorescence (Fig. 6D) was strongest in AAV-Ctrl + LPS mice and was markedly diminished after GBP5 knockdown. Taken together, these in vivo data demonstrate that GBP5 simultaneously promotes NLRP3-driven inflammation and drives upregulation of the CD73–adenosine counter-regulatory axis.

Fig. 5.

GBP5 knockdown suppresses NLRP3 inflammasome activation and modulates adenosine-signaling proteins in vivo. (A) Representative Western blots of NLRP3, ASC and cleaved caspase-1 in lung tissues from AAV-Ctrl + saline, AAV-Ctrl + LPS, AAV-shGBP5 + saline and AAV-shGBP5 + LPS mice. (B–D) Densitometric analysis of lung NLRP3 (B), ASC (C) and cleaved caspase-1 (D) protein levels in the four groups. (E) Representative Western blots of CD39 and ADORA2A (A2A receptor) in lung tissues from the four groups. (F) Representative Western blot of CD73 protein in lung tissues from the four groups. (G–I) Densitometric analysis of lung CD39 (G), ADORA2A (H) and CD73 (I) protein expression in the four groups. n = 10 per group. Data are presented as mean ± SD. Statistical analysis was performed as described in the Methods. ns, not significant; *P < 0.05, **P < 0.01,***P < 0.001

Fig. 6.

GBP5 knockdown alters adenosine-signaling–related gene expression in mouse lungs. (A–C) Relative mRNA expression of CD39 (A), Adora2a (A2A receptor, B) and CD73 (C) in lung tissues from AAV-Ctrl + saline, AAV-Ctrl + LPS, AAV-shGBP5 + saline and AAV-shGBP5 + LPS mice, determined by qPCR (2^-ΔΔCt). (D) Immunofluorescence staining of CD73 in paraffin-embedded lung sections from the four groups (blue, DAPI; yellow, CD73; 40× magnification). (E) Representative Western blot of GBP5 protein in control (Ctrl), LV-Vector and LV-oeGBP5 THP-1 cells. (F) Densitometric analysis of GBP5 protein expression in Ctrl, LV-Vector and LV-oeGBP5 groups. (G) Relative GBP5 mRNA expression in Ctrl, LV-Vector and LV-oeGBP5 cells by qPCR (2^-ΔΔCt). A–D n = 10 per group, E–G n = 3 independent experiments. Data are presented as mean ± SD. Statistical analysis was performed as described in the Methods. ns, not significant; *P < 0.05, **P < 0.01,***P < 0.001

GBP5 Enhances CD73 Expression and NLRP3 Activation in Vitro

To validate the above findings in vitro, THP-1 macrophages were transfected with an LV-oeGBP5 plasmid and then stimulated with LPS. Successful GBP5 overexpression was confirmed by both Western blotting (Fig. 6E-F) and RT-qPCR (Fig. 6G). Six experimental groups were included: Ctrl, LPS, LV-Vector, LV-oeGBP5, LPS + LV-Vector, and LPS + LV-oeGBP5. RT-qPCR showed that LPS stimulation increased CD73 mRNA expression, and under LPS challenge, LV-oeGBP5 further elevated CD73 mRNA levels compared with the LPS + LV-Vector group (Fig. 7A). CD73 expression was then assessed by flow cytometry using an anti-CD73 antibody, and data were analyzed with FlowJo. The percentage of CD73-positive cells was higher in the LPS group than in the Ctrl group, and was further increased in the LPS + LV-oeGBP5 group compared with the LPS + LV-Vector group (Fig. 7B-D). Immunofluorescence staining for GBP5 and CD73 in all six groups further supported these results: LPS treatment increased both GBP5 and CD73 fluorescence signals, and LV-oeGBP5 under LPS conditions led to a further enhancement of both signals relative to LPS + LV-Vector (Fig. 7E-H).

Fig. 7.

GBP5 overexpression modulates CD73 expression in THP-1–derived macrophages. (A) Relative CD73 mRNA expression in six groups (Ctrl, LPS, LV-Vector, LV-oeGBP5, LPS + LV-Vector, LPS + LV-oeGBP5) measured by qPCR (2^-ΔΔCt). (B, C) Representative flow cytometry plots of CD73 surface expression in the six groups. (D) Quantitative comparison of CD73-positive cells across the six groups by flow cytometry. (E, F) Representative immunofluorescence images of THP-1 cells from the six groups showing nuclear staining (DAPI, blue), GBP5 (red), CD73 (green) and merged images. (G, H) Quantification of mean fluorescence intensity (MFI) of GBP5 (G) and CD73 (H) immunofluorescence signals across the six groups, measured from at least 3 randomly selected fields per group using ImageJ. n = 3 independent experiments. Data are presented as mean ± SD. Statistical analysis was performed as described in the Methods. ns, not significant; *P < 0.05, **P < 0.01,***P < 0.001

Western blotting showed that, compared with the Ctrl group, LPS treatment increased the protein expression of GBP5 (Fig. 8A-B) and CD73 (Fig. 8A, C), along with marked upregulation of ASC (Fig. 8D-E), NLRP3 (Fig. 8D, F), and cleaved caspase-1 (Fig. 8D, G), indicating activation of NLRP3 inflammasome signaling. Moreover, compared with the LPS + LV-Vector group, the LPS + LV-oeGBP5 group exhibited further increases in GBP5 and CD73 expression, accompanied by additional elevation of ASC, NLRP3, and cleaved caspase-1, suggesting that GBP5 overexpression enhances CD73 expression and further promotes inflammasome activation at the protein level (Fig. 8A-G). RT-qPCR (Fig. 8H-J) and ELISA (Fig. 8K-M) performed across the six groups showed consistent inflammatory changes. Relative to the Ctrl group, the LPS group displayed significantly increased TNF-α, IL-1β, and IL-18 mRNA expression and higher secretion of these cytokines in culture supernatants. Compared with the LPS + LV-Vector group, the LPS + LV-oeGBP5 group showed further increases in both mRNA expression and secreted levels of TNF-α, IL-1β, and IL-18.

Fig. 8.

GBP5 overexpression enhances CD73 expression and activates inflammasome- and transcription factor–related proteins in THP-1 cells. (A) Representative Western blots of GBP5 and CD73 proteins in Ctrl, LPS, LV-Vector, LV-oeGBP5, LPS + LV-Vector and LPS + LV-oeGBP5 groups. (B, C) Densitometric analysis of GBP5 (B) and CD73 (C) protein expression in the six groups. (D) Representative Western blots of ASC, NLRP3 and cleaved caspase-1 in the six groups. (E–G) Densitometric analysis of ASC (E), NLRP3 (F) and cleaved caspase-1 (G) protein levels in the six groups. (H-J) Relative mRNA expression of TNF-α (H), IL-1β (I) and IL-18 (J) in Ctrl, LPS, LV-Vector, LV-oeGBP5, LPS + LV-Vector and LPS + LV-oeGBP5 groups measured by qPCR (2^-ΔΔCt). (K-M) ELISA analysis of TNF-α (K), IL-1β (L) and IL-18 (M) concentrations in culture supernatants from the six groups. n = 3 independent experiments. Data are presented as mean ± SD. Statistical analysis was performed as described in the Methods. ns, not significant; *P < 0.05, **P < 0.01,***P < 0.001

GBP5-induced CD73 Expression Depends on Transcription Factor Activation

To dissect the molecular mechanism by which GBP5 induces CD73 expression, we first examined whether GBP5 regulates CD73 through direct protein–protein interaction, however, co-immunoprecipitation (co-IP) assays using anti-GBP5 and anti-CD73 antibodies failed to detect reciprocal pull-down of the partner protein in THP-1 macrophage lysates, indicating that GBP5 does not directly bind CD73 at the protein level. Based on prior literature, we hypothesized that GBP5 may regulate CD73 via transcription factors. We therefore performed additional Western blot analyses of candidate transcription factors (NF-κB/ HIF-1α/p-STAT1) across the six experimental groups. Compared with the LPS + LV-Vector group, cells in the LPS + LV-oeGBP5 group showed further increases in NF-κB and HIF-1α expression, whereas p-STAT1 showed no detectable difference (Fig. 9A-D).

Fig. 9.

Transcription factor inhibition reveals GBP5-dependent regulation of CD73 in THP-1 cells. (A) Representative Western blots of NF-κB, HIF-1α, STAT1 and phospho-STAT1 in the six groups. (B–D) Densitometric analysis of NF-κB (B), HIF-1α (C) and the phospho-STAT1/STAT1 ratio (D) in the six groups. (E) Representative Western blot of CD73 protein in eight groups: Ctrl, LPS, LV-Vector, LV-oeGBP5, LPS + LV-oeGBP5, LPS + LV-oeGBP5 + YC-1 (HIF-1α inhibitor), LPS + LV-oeGBP5 + fludarabine (STAT1 inhibitor), and LPS + LV-oeGBP5 + BAY11-7082 (NF-κB inhibitor). (F) Densitometric analysis of CD73 protein expression in the eight groups. (G) Relative CD73 mRNA expression in the eight groups determined by qPCR (2^-ΔΔCt). (H) Dual-luciferase reporter assay assessing the effect of HIF-1α on CD73 promoter activity. (I) Dual-luciferase reporter assay assessing the effect of NF-κB on CD73 promoter activity. n = 3 independent experiments.Data are presented as mean ± SD. Statistical analysis was performed as described in the Methods. ns, not significant; *P < 0.05, **P < 0.01,***P < 0.001.

Next, we used pharmacological inhibitors targeting these candidate pathways. THP-1 cells were transfected with a GBP5 overexpression plasmid and treated with the NF-κB inhibitor BAY11-7082, the STAT1 inhibitor fludarabine, or the HIF-1α inhibitor YC-1. GBP5 overexpression markedly increased protein levels (Fig. 9E-F) and CD73 mRNA(Fig. 9G). This upregulation was strongly attenuated by HIF-1α inhibition and moderately reduced by NF-κB inhibition, while STAT1 inhibition had no significant effect on CD73 expression. These findings indicate that GBP5-driven CD73 induction depends predominantly on HIF-1α, with NF-κB serving at most a secondary modulatory role, whereas STAT1 appears dispensable in this context.

Dual-luciferase reporter assays were used to confirm transcriptional regulation. HIF-1α overexpression (Fig. 9H), selectively increased CD73 promoter-driven luciferase activity without affecting promoterless pGL3-Basic activity. By contrast, NF-κB overexpression did not significantly alter luciferase activity in either the pGL3-Basic or pGL3-CD73 group (Fig. 9I). Together, these data support that HIF-1α, but not NF-κB, is a major transcriptional activator of CD73 under our experimental conditions.

CD73 Overexpression Enhances cAMP Signaling and Inhibits NLRP3 Activation

To investigate the role of CD73 in LPS-induced cellular injury, we overexpressed CD73 in LPS-treated THP-1 macrophages. Successful overexpression was confirmed by Western blotting (Fig. 10A-B) and RT-qPCR (Fig. 10C). Four groups were included: Ctrl, LPS, LPS + LV-Vector, and LPS + LV-oeCD73. LPS stimulation increased the expression of CD73 (Fig. 10D-E) and ADORA2A (Fig. 10D, F), whereas CD73 overexpression further elevated ADORA2A (Fig. 10D, F) and phosphorylated CREB (p-CREB) protein levels (Fig. 10G-I).

Fig. 10.

CD73 overexpression upregulates A2A receptor and CREB phosphorylation in THP-1 cells. (A) Representative Western blot of CD73 protein in Ctrl, LV-Vector and LV-oeCD73 THP-1 cells. (B) Densitometric analysis of CD73 protein expression in Ctrl, LV-Vector and LV-oeCD73 groups. (C) Relative CD73 mRNA expression in Ctrl, LV-Vector and LV-oeCD73 groups by qPCR (2^-ΔΔCt). (D) Representative Western blots of CD73 and ADORA2A (A2A receptor) proteins in Ctrl, LPS, LPS + LV-Vector and LPS + LV-oeCD73 groups. (E, F) Densitometric analysis of CD73 (E) and ADORA2A (F) protein expression in the four groups. (G) Representative Western blots of total CREB and phospho-CREB in LPS + LV-Vector and LPS + LV-oeCD73 groups. (H, I) Densitometric analysis of total CREB (H) and the phospho-CREB/CREB ratio (I) in LPS + LV-Vector and LPS + LV-oeCD73 groups. n = 3 independent experiments.Data are presented as mean ± SD. Statistical analysis was performed as described in the Methods. ns, not significant; *P < 0.05, **P < 0.01,***P < 0.001

Functionally, Western blot (Fig. 11A-D) and ELISA (Fig. 11E-H) analyses showed that CD73 overexpression suppressed the expression of NLRP3 inflammasome components (ASC, NLRP3, and cleaved caspase-1) and reduced the production of pro-inflammatory cytokines (TNF-α, IL-1β, and IL-18). In parallel, CD73 overexpression enhanced cAMP/p-CREB signaling. Collectively, these in vitro data indicate that CD73 overexpression alleviates inflammatory signaling by activating the cAMP/p-CREB pathway and inhibiting NLRP3 inflammasome signaling.

Fig. 11.

CD73 overexpression suppresses NLRP3 inflammasome activation and pro-inflammatory cytokine release via cAMP in THP-1 cells. (A) Representative Western blots of NLRP3, ASC and cleaved caspase-1 proteins in Ctrl, LPS, LPS + LV-Vector and LPS + LV-oeCD73 groups. (B–D) Densitometric analysis of NLRP3 (B), ASC (C) and cleaved caspase-1 (D) protein levels in the four groups. (E) ELISA analysis of intracellular cAMP levels in the four groups. (F–H) ELISA analysis of TNF-α (F), IL-1β (G) and IL-18 (H) concentrations in culture supernatants from Ctrl, LPS, LPS + LV-Vector and LPS + LV-oeCD73 groups. n = 3 independent experiments.Data are presented as mean ± SD. Statistical analysis was performed as described in the Methods. ns, not significant; *P < 0.05, **P < 0.01,***P < 0.001

Discussion

This study uncovers a previously unrecognized dual function of GBP5 in ALI: while GBP5 exacerbates pulmonary inflammation through NLRP3 inflammasome activation, it concurrently induces a HIF-1α–CD73–adenosine–cAMP–CREB counter-regulatory feedback loop that restrains excessive cytokine production. These findings reveal an unexpected crosstalk between innate immune activation and purinergic metabolic signaling in the injured lung and identify GBP5 as a molecular node at which pro-inflammatory and pro-resolving pathways converge.

Using RNA-seq, we first observed a broad upregulation of guanylate-binding protein (GBP) family members in the lungs of LPS-induced lung injury mice, with GBP5 showing the most pronounced increase. Guanylate-binding proteins are a family of interferon-inducible GTPases known for their roles in the innate immune response including GBP5 [18]. GBP5 was demonstrated to promote NLRP3 inflammasome assembly and caspase-1 activation in response to pathogen-derived signals [4]. NLRP3 inflammasome promotes inflammation in the lung tissue and exacerbates tissue damage and impairs normal lung function [5]. Our results align with previous studies showing that GBP5 overexpression in macrophages activates the NLRP3 inflammasome, leading to enhanced IL-1β and IL-18 production, which are key mediators of inflammation in ALI and sepsis. in vivo, GBP5 knockdown significantly reduced NLRP3, ASC, and cleaved caspase-1expression in lung tissues, corroborating its role as an inflammasome activator.These observations are consistent with findings that GBP5 facilitates lung inflammation and autophagy regulation [19].

RNA-seq analysis in our model further revealed that, in addition to GBP5, multiple genes involved in adenosine signaling were upregulated during LPS-induced lung injury, with CD73 showing a particularly marked increase. Functionally, inhibition of GBP5 not only ameliorated lung injury but also downregulated CD73 expression, suggesting a positive regulatory link between GBP5 and CD73. Prior studies have established that CD73 expression is controlled by a complex network of transcription factors and microRNAs [20], and that hypoxia-inducible factor-1α (HIF-1α), NF-κB and STAT family members are important upstream regulators of inflammatory and hypoxic responses [21]. Building on evidence that GBP5 is associated with hypoxia/HIF-1α signaling and can modulate NF-κB/STAT3 pathways in kidney and other tissues [22, 23], we hypothesized that GBP5 upregulates CD73 by influencing one or more of these transcription factors.

To test this hypothesis, we pharmacologically inhibited HIF-1α, NF-κB and STAT1 and evaluated CD73 expression. Both HIF-1α and NF-κB inhibition, particularly HIF-1α blockade, attenuated GBP5-induced CD73 expression at the mRNA and protein levels. However, dual-luciferase reporter assays using CD73 promoter constructs demonstrated that only HIF-1α enhanced CD73 promoter activity, whereas NF-κB did not significantly increase CD73 transcriptional activity. These data, together with the classic demonstration that HIF-1α directly regulates the CD73 promoter under hypoxic conditions, support a model in which GBP5 primarily upregulates CD73 through a HIF-1α–dependent transcriptional mechanism, rather than via direct NF-κB-mediated promoter activation.

Within the purinergic system, CD39 and CD73 act sequentially to degrade extracellular ATP/ADP to adenosine, thereby shaping the inflammatory milieu in injured tissues. Adenosine then signals through four G-protein–coupled receptors (A1, A2A, A2B and A3) with distinct expression patterns and downstream signaling pathways. Accumulating evidence indicates that adenosine can exert divergent effects in inflammatory diseases depending on receptor subtype, tissue context and disease stage [24]. In many acute injury models, A2A receptor activation is broadly anti-inflammatory and organ-protective, largely via Gs-coupled increases in cAMP and downstream inhibition of leukocyte adhesion, oxidative burst and pro-inflammatory cytokine production [25, 26]. In contrast, A2B receptor signaling is often associated with pro-inflammatory or pro-fibrotic responses in conditions of chronic hypoxia, metabolic stress and sustained tissue damage, although A2B can be protective in some forms of acute lung injury, underscoring the context-dependency of adenosine receptor biology [27, 28].

Several lines of evidence support a protective role of CD73-generated adenosine in lung injury. In the context of sepsis, CD73 deficiency has been shown to worsen pulmonary inflammation and organ injury, whereas enhancement of CD73-mediated adenosine production confers protection [29]. Similarly, CD73-derived adenosine has demonstrated protective effects in hemorrhagic shock models, attenuating vascular leakage and inflammatory cytokine release [30]. The anti-inflammatory effects of adenosine in macrophages are mediated predominantly through A2A receptors [31, 32] and, in certain inflammatory contexts, through A2B receptors [33], both signaling via Gs-coupled cAMP elevation to suppress pro-inflammatory gene expression. These prior findings strongly support the protective, pro-resolving role of the CD73–adenosine axis that we describe in our ALI model. In hyperoxic and endotoxin-induced lung injury models, loss of CD73-mediated extracellular adenosine production exacerbates pulmonary inflammation, impairs alveolar development or delays resolution of injury, indicating that controlled CD73 activity is required for limiting acute lung damage. CD73⁺ regulatory T cells have also been shown to contribute to adenosine-dependent resolution of ALI, further highlighting the importance of CD73 in restoring immune homeostasis in the lung [34, 35]. Beyond the lung, increased extracellular adenosine during high-altitude hypoxia activates erythrocyte A2B receptors and AMPK, enhancing 2,3-BPG production and oxygen unloading to peripheral tissues, thereby protecting against systemic hypoxia, inflammation and vascular leakage [36]. Together, these studies suggest that the CD73–adenosine axis functions as a critical endogenous brake on excessive inflammation and tissue hypoxia in diverse organs.

At the same time, CD73 has a well-recognized pro-tumorigenic role in the cancer setting, where its overexpression promotes immune evasion by generating an immunosuppressive adenosine-rich microenvironment. Small-molecule CD73 inhibitors such as AB680 (quemliclustat) have been developed to block extracellular adenosine production, restore T-cell proliferation and effector function, and enhance the antitumor efficacy of PD-1 blockade in preclinical models. Early phase clinical trials are now evaluating CD73 inhibitors in combination with immune checkpoint blockade in multiple solid tumors [37]. These apparently opposite roles of CD73 in cancer versus acute inflammatory injury strongly suggest that CD73/adenosine signaling may exert disease- and stage-specific effects: protective and pro-resolving in acute sterile or infectious injury, but immunosuppressive and tumor-promoting in chronic neoplastic contexts. Our data in ALI support the former, tissue-protective role.

Mechanistically, our cell experiments showed that CD73 overexpression increased A2A receptor expression, elevated intracellular cAMP levels and enhanced CREB phosphorylation. Combining these observations with our in vivo findings, we propose the following model: GBP5 promotes NLRP3 inflammasome activation and pro-inflammatory cytokine release in response to LPS, thereby aggravating lung injury. Concurrently, GBP5 induces HIF-1α–dependent upregulation of CD73, leading to increased adenosine production in the extracellular space. We observed a concomitant upregulation of A2A receptor (ADORA2A) expression alongside elevated cAMP and CREB phosphorylation; however, as we did not directly test A2A receptor function using pharmacological or genetic approaches, we cannot conclude that A2A receptor activation is solely responsible for these downstream effects. The elevated cAMP and p-CREB may reflect signaling through one or more adenosine receptor subtypes, and future studies using selective A2A agonists/antagonists or receptor knockdown will be needed to delineate the specific receptor contribution. Nonetheless, the cAMP/PKA–CREB axis appears to be engaged downstream of CD73, and its activation in turn suppresses further NLRP3 activation and cytokine release, forming a negative feedback loop that limits uncontrolled inflammation. This is in line with reports that CD73-generated adenosine augments cAMP/PKA/CREB signaling in several cell types and that pharmacological activation of cAMP/PKA/CREB dampens NLRP3 inflammasome-mediated inflammation in experimental colitis and other inflammatory models [38, 39].

Our findings also resonate with previous work describing an adenosine–A2B–PI3K–AKT–FoxO1 axis [40] and an A2B–Nrf2/HO-1 pathway [41] as important negative regulators of NLRP3 inflammasome activity. In particular, Thapa et al. showed that extracellular adenosine acting on A2B receptors upregulates the Nrf2/HO-1 axis in a cAMP-dependent manner, which negatively regulates NLRP3 expression and assembly and restrains hematopoietic stem/progenitor cell trafficking in an inflammasome-dependent fashion. CD73-generated extracellular adenosine mediates cardioprotection during ischemic preconditioning primarily through activation of A2B adenosine receptors [42]. These data, together with our GBP5–CD73–adenosine–cAMP–CREB axis, suggest that adenosine engages multiple receptor-specific pathways (A2A–cAMP–PKA–CREB, A2B–PI3K–AKT–FoxO1, A2B–Nrf2/HO-1) to dampen NLRP3-driven inflammation in a context-dependent manner. Although we did not directly dissect all of these downstream pathways in the present study, our results provide strong support for the concept that CD73-derived adenosine serves as a built-in compensatory mechanism to counterbalance GBP5-driven inflammasome activation in ALI.

Despite these advances, several important questions remain. First, our work focused on acute time points in LPS-induced lung injury, and the long-term consequences of modulating GBP5 and CD73 on lung repair, fibrosis and chronic remodeling are still unknown. Given that chronic adenosine elevation has been implicated in fibrotic lung disease, it will be important to determine whether sustained CD73 activation is beneficial or detrimental in the later phases of lung injury. Second, the effect of GBP5 on other immune cell populations, including macrophage polarization, neutrophil recruitment, and T-cell activation or exhaustion, was not fully explored here. Given that GBP5 expression is tightly linked to interferon signaling and broader immune landscape remodeling in cancer and infection, dissecting its cell type–specific roles in the lung microenvironment may reveal additional layers of regulation. Third, while we identified HIF-1α as a key transcriptional mediator of GBP5-induced CD73 upregulation, the possibility that GBP5 also modulates CD73 through epigenetic mechanisms or microRNA networks cannot be excluded and warrants further investigation. Notably, this HIF-1α-centered regulation may extend beyond direct transcriptional control of CD73: several HIF-1α-dependent microRNAs have been shown to suppress pro-inflammatory gene expression in macrophages [43–45], suggesting that GBP5-driven HIF-1α activation could simultaneously fine-tune CD73 induction and restrain NLRP3 activity through coordinated post-transcriptional mechanisms. The protective role of HIF-1α signaling in ARDS is moreover increasingly well-supported at both the preclinical and clinical level: HIF-1α activation mitigates alveolar permeability and attenuates cytokine release in LPS- and ventilation-induced lung injury models, and a recent randomized, double-blinded, multicenter phase II trial of the FDA-approved HIF stabilizer vadadustat in hypoxemic SARS-CoV-2 patients demonstrated that pharmacological HIF activation was well-tolerated and associated with reduced rates of severe lung injury [46]. These observations raise the intriguing possibility that GBP5-mediated HIF-1α activation may itself confer tissue protection through broader HIF-1α-dependent gene programs—beyond its role in CD73 upregulation—an avenue that warrants dedicated investigation. Fourth, the present study did not address the contribution of equilibrative nucleoside transporters (ENTs), particularly ENT1 and ENT2. ENTs terminate extracellular adenosine signaling by transporting it into the intracellular compartment, and importantly, their expression is repressed by hypoxia and inflammatory cytokines in a HIF-1α-dependent manner [47, 48]. This ENT suppression works in concert with CD73-mediated adenosine generation to amplify the extracellular adenosine signal, and prior work has demonstrated that pharmacological ENT1/2 blockade elevates lung adenosine levels, attenuates experimental ALI, and suppresses NLRP3 inflammasome activation through A2AR and A2BR signaling [49, 50]. The interplay between HIF-1α-driven ENT repression and the CD73–adenosine axis identified in our study therefore represents an important regulatory layer that may further amplify the feedback loop during pulmonary inflammation, and merits direct investigation in future work.

Finally, a central limitation of the present study is that the functional role of CD73 was not directly tested in vivo. Although we extensively characterized CD73 in THP-1 macrophages through overexpression, future studies using intratracheal delivery of CD73 inhibitors (e.g., AOPCP), CD73-overexpressing AAV vectors, or CD73-knockout mice in the LPS-induced ALI model will be necessary to establish the causal contribution of CD73-mediated adenosine production to the feedback loop described here. Beyond this, future work should (i) extend observations to chronic injury models such as bleomycin-induced fibrosis or viral lung remodeling to assess the long-term consequences of GBP5 and CD73 modulation, (ii) evaluate GBP5-targeted interventions combined with CD73 agonists or inhibitors to optimize the balance between inflammation and repair, and (iii) apply single-cell multi-omics approaches to map this signaling axis across specific immune and structural cell subsets in vivo. Together, these directions will be essential for translating the mechanistic insights described here into rational therapeutic strategies for ALI and ARDS.

Conclusion

In summary, our study reveals a previously unrecognized GBP5-driven mechanism of ALI regulation in which GBP5 activates the NLRP3 inflammasome to drive acute inflammation, while simultaneously inducing a protective CD73–adenosine–cAMP–CREB feedback loop that restrains excessive inflammasome activity. This dual action of GBP5 highlights the delicate balance between immune activation and resolution in the injured lung. Targeting GBP5 to reduce inflammasome overactivation, while preserving or selectively enhancing CD73-mediated adenosine signaling at appropriate disease stages, may offer a promising strategy for modulating inflammation and improving outcomes in ALI, sepsis and potentially other inflammatory lung diseases such as ARDS.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contributions

J.Z. and X.Z. conceived and designed the study. J.Z., L.X., X.Y., and Q.L. performed the experiments and analyzed the data. J.Z. and X.Z. wrote the manuscript with the input of other authors. The authors read and approved the final manuscript.

Funding

This work was supported by grants to Jia Zhang, Grant No. JQRC2024001, Henan Province Outstanding Young and Middle-aged Talent Program for Health Science and Technology Innovation.

Data Availability

RNA-sequencing data were uploaded to the China National Center for Bioinformation (CNCB) No. [subCRA063315](https:/ngdc.cncb.ac.cn/gsub/submit/gsa/subCRA063315).

Declarations

Consent for Publication

Not applicable.

Clinical Trial Number

Not applicable.

Competing Interests

The authors declare no competing interests.