Daily Respiratory Research Analysis
Genetic analyses advance understanding of idiopathic pulmonary fibrosis and reveal shared etiologic signals with severe COVID-19, prioritizing therapeutic targets. Mechanistic work identifies Hv1 channel blockade as a strategy to curb neutrophil-driven acute lung injury. A prospective, interpretable ultrasound-based machine learning model achieves high accuracy for bedside diagnosis of COPD exacerbations.
Summary
Genetic analyses advance understanding of idiopathic pulmonary fibrosis and reveal shared etiologic signals with severe COVID-19, prioritizing therapeutic targets. Mechanistic work identifies Hv1 channel blockade as a strategy to curb neutrophil-driven acute lung injury. A prospective, interpretable ultrasound-based machine learning model achieves high accuracy for bedside diagnosis of COPD exacerbations.
Research Themes
- Genetic architecture of fibrotic lung disease and overlap with severe COVID-19
- Neutrophil-targeted immunomodulation in infectious acute lung injury
- Interpretable point-of-care ultrasound with machine learning for AECOPD diagnosis
Selected Articles
1. Common and rare variant analyses reveal genetic factors underlying idiopathic pulmonary fibrosis and its shared aetiology with severe COVID-19.
A large-scale GWAS meta-analysis (11,746 IPF cases and 1.4M controls) identified an additional IPF association at 1q21.2 and showed colocalization with severe COVID-19 genetic signals. Post-GWAS analyses, rare variant testing, and in vitro inhibition of a probable effector gene support shared aetiology and nominate druggable pathways for IPF with potential overlap in severe COVID-19.
Impact: By integrating WGS, meta-analysis, and colocalization with severe COVID-19, this work prioritizes IPF causal biology and highlights shared mechanisms that could enable repurposing or co-development of therapeutics.
Clinical Implications: Genetic signals and putative effector genes can guide target selection and stratification in IPF trials, and shared loci with severe COVID-19 suggest potential for common antifibrotic or immune-modulating therapies.
Key Findings
- Meta-analysis identified an additional IPF association at 1q21.2 (rs16837903, OR 0.88).
- Colocalization analyses revealed shared genetic architecture between IPF and severe COVID-19.
- Post-GWAS, rare variant analyses, and in vitro inhibition of a probable effector gene support causal pathways.
Methodological Strengths
- Integration of WGS, large-scale meta-analysis (11,746 cases; 1,416,493 controls)
- Colocalization and multi-trait analyses linking IPF and severe COVID-19, with functional follow-up
Limitations
- Limited functional validation across all loci and pathways
- Potential ancestry and cohort heterogeneity despite meta-analytic design
Future Directions: Mechanistic dissection of prioritized effector genes, target validation in relevant models, and genotype-informed clinical trials to test shared antifibrotic or immunomodulatory strategies.
BACKGROUND: Idiopathic pulmonary fibrosis (IPF) is a progressive and debilitating respiratory disease with limited therapeutic options. Genetic association studies for IPF have identified several associations and probable effector genes that could not only help understanding IPF pathogenesis but also develop effective treatments. Assessing genetic overlap between IPF and severe COVID-19, an acute respiratory disease that can trigger pulmonary fibrosis, may reveal shared aetiology and mechanisms, thereby supporting the development of common treatments. METHODS: We carried out genome-wide association studies (GWAS), post-GWAS, and rare variant analyses using whole genome sequencing data from the 100,000 Genomes Project IPF cohort (n = 586). We performed a meta-analysis combining 100 kGP with published IPF GWASs (total 11,746 cases and 1,416,493 controls). We tested inhibition in vitro for a probable effector gene of an identified association. We also investigated genetic colocalisation between IPF and severe COVID-19 and leveraged their genetic correlation through multi-trait meta-analysis for discovery. FINDINGS: IPF meta-analysis identified an additional association at 1q21.2 (rs16837903, OR [95% CI] = 0.88 [0.85, 0.92], P = 9.5 × 10
2. C6 peptide blockade of Hv1 channels inhibits neutrophil migration into the lungs to suppress Pseudomonas aeruginosa-induced acute lung injury.
In a live Pseudomonas aeruginosa infection model, the Hv1 blocker C6 reduced neutrophil alveolar infiltration (~86%), lung injury, and inflammatory cytokines while suppressing neutrophil ROS and calcium signals. Transcriptomics of BAL neutrophils and assays in human neutrophils confirm C6’s immunomodulatory mechanism, nominating Hv1 as a tractable target for infectious ALI.
Impact: This study translates a channel-targeting concept into an infectious ALI model and bridges to human neutrophils, providing mechanistic and translational evidence for Hv1 blockade as a therapeutic approach.
Clinical Implications: If safety and delivery can be addressed, Hv1 inhibition could complement antimicrobials by dampening neutrophil-driven injury in infectious ALI/ARDS, potentially improving oxygenation and reducing ventilator-induced damage.
Key Findings
- C6 reduced neutrophil alveolar infiltration by approximately 86% and improved lung injury scores in P. aeruginosa-induced ALI.
- BAL fluid proinflammatory cytokines, neutrophil ROS generation, and intracellular calcium were suppressed by C6.
- RNA-seq of BAL neutrophils showed 51 downregulated genes involved in migration, cytokine release, and ROS; human neutrophils exhibited reduced chemotaxis and activation with C6.
Methodological Strengths
- In vivo live bacterial infection model with quantitative histology and BAL analyses
- Mechanistic validation across RNA-seq of BAL neutrophils and human neutrophil functional assays
Limitations
- Preclinical study without pharmacokinetic/safety data or dosing strategies for humans
- Model focused on P. aeruginosa; generalizability to other ALI aetiologies needs testing
Future Directions: Define PK/PD, delivery routes (e.g., inhalation), and safety of Hv1 blockade; evaluate efficacy in other infectious and sterile ALI models and in combination with antibiotics.
BACKGROUND: Acute Lung Injury (ALI) and its most severe form, Acute Respiratory Distress Syndrome (ARDS), are critical pulmonary conditions characterized by life-threatening acute hypoxic respiratory failure, affecting over three million individuals globally each year. ALI involves alveolar inflammation and disruption of the alveolar-capillary barrier, primarily driven by neutrophil infiltration and the release of inflammatory mediators. In our previous study using a lipopolysaccharide (LPS)-induced mouse model of ALI, we demonstrated that C6, a peptide inhibitor of voltage-gated proton channels (Hv1), ameliorates lung injury, identifying Hv1 as a potential therapeutic target. However, (i) whether the anti-inflammatory effects of C6 are translatable to a clinically relevant live bacterial infection model, and (ii) the molecular mechanisms underlying these anti-inflammatory effects, remain unknown, and are a crucial next step towards targeted rational drug development. METHODS: To induce ALI, we used an intratracheal RESULTS: C6 mitigates P. aeruginosa-induced ALI in mice by reducing neutrophil infiltration into the alveolar space by ~ 86%, improving lung injury scores, decreasing BAL fluid proinflammatory cytokine levels, and suppressing neutrophil ROS production and intracellular calcium levels. RNA sequencing of BAL neutrophils revealed 51 downregulated genes, including key regulators of neutrophil migration, cytokine release, and ROS production; only three genes were upregulated and they also have roles in neutrophil immune defense. In human neutrophils, C6 similarly inhibited chemotaxis and reduced ROS and cytokine release, and calcium influx. CONCLUSIONS: Targeting Hv1 with C6 effectively protects against P. aeruginosa-induced ALI by limiting neutrophil recruitment and activation. These findings establish C6 as a promising therapeutic candidate against infectious ALI and provide important mechanistic insights into its immunomodulatory effects on neutrophils.
3. Interpretable machine learning model based on multimodal ultrasound for bedside diagnosis of acute exacerbations in COPD.
A prospective, single-center diagnostic study in 316 COPD patients shows that an SVM model using six routine variables (five ultrasound-derived) achieved AUC ~0.93 for AECOPD identification. SHAP interpretation highlighted lung ultrasound score, diaphragmatic dysfunction, and quadriceps atrophy as key contributors, enabling transparent bedside decision support.
Impact: Demonstrates a high-performing, interpretable, and implementable point-of-care diagnostic integrating lung, diaphragm, and muscle ultrasound—bridging physiology and AI for acute COPD care.
Clinical Implications: Supports real-time triage and early treatment decisions in suspected AECOPD using bedside ultrasound; may reduce reliance on radiography and expedite initiation of evidence-based therapies.
Key Findings
- SVM model achieved AUC 0.9321 (train) and 0.9302 (test) for diagnosing AECOPD.
- Six-variable model leveraged five ultrasound features; key predictors were lung ultrasound score, diaphragm dysfunction, and quadriceps atrophy via SHAP.
- Prospective acquisition and standardized multimodal ultrasound enabled bedside applicability.
Methodological Strengths
- Prospective diagnostic design with predefined train-test split
- Model interpretability via SHAP and use of routine, bedside-acquirable ultrasound variables
Limitations
- Single-center cohort; external validation needed for generalizability
- No comparison to biomarker-enhanced or CT-based diagnostic pathways
Future Directions: Multicenter validation, integration with clinical and biomarker data, and implementation trials to assess impact on time-to-treatment and outcomes.
BACKGROUND: Acute exacerbations of chronic obstructive pulmonary disease (AECOPD) are associated with accelerated lung function decline and increased mortality. However, early and accurate diagnosis remains clinically challenging due to nonspecific symptoms and limitations of existing diagnostic tools. This study aimed to develop an interpretable machine learning (ML) model integrating multimodal ultrasound indicators to facilitate real-time bedside diagnosis of AECOPD. METHODS: In this prospective, single-center study, 316 patients with COPD underwent standardized lung, diaphragmatic, and quadriceps ultrasound examinations upon hospital admission. Four ML algorithms were developed using a 7:3 training-to-test data split. Model performance was assessed by area under the receiver operating characteristic curve (AUC), and interpretability was enhanced using SHapley Additive exPlanations (SHAP). RESULTS: The support vector machine (SVM) model achieved the best diagnostic performance, with an AUC of 0.9321 in the training set and 0.9302 in the test set. The final model incorporated six routinely obtainable variables, five of which were ultrasound derived. SHAP analysis identified elevated lung ultrasound scores, diaphragmatic dysfunction, and quadriceps atrophy as the most influential predictors. CONCLUSIONS: This non-invasive and interpretable ML model, based on bedside ultrasound features, offers a clinically feasible tool for real-time AECOPD diagnosis. Further multicenter validation is warranted to confirm generalizability and explore integration with additional biomarkers or imaging modalities. CLINICAL TRIAL NUMBER: Not applicable.