Background:
Chronic diseases are increasingly recognized as systemic, with clinical evidence highlighting a strong interplay between pulmonary and hepatic health[1, 2]. However, the molecular mechanisms driving this hepato-pulmonary axis remain uncharacterized. A critical missing link is systemic metabolic coordination, as cellular metabolic shifts typically precede macroscopic tissue damage. While [18F]-fluorodeoxyglucose ([18F]-FDG) positron emission tomography (PET) offers an unprecedented window into in vivo metabolism[3, 4], the specific metabolic hierarchy between the liver and lung remains undefined. Consequently, it is unknown whether this molecular coupling translates into chronic physical implications at a population scale, or ultimately drives systemic metabolic collapse, such as incident type 2 diabetes (T2D).
Hypothesis/Objective:
The hypothesis of this project is that there is a fundamental, bidirectional metabolic coupling between the liver and lung, wherein the liver acts as the metabolic gatekeeper. Chronic dysregulation of this acute hepato-pulmonary metabolic axis drives reciprocal structural and functional decline, acting as a genetically determined, systemic conduit for T2D pathogenesis through the amplification of immune-inflammatory and metabolic signaling.
Methods:
To establish physiological metabolic crosstalk, [18F]-FDG PET was utilized in a cohort of metabolically healthy human subjects (n=154)[5]. Standardized uptake value ratios were extracted to evaluate inter-organ metabolic synchronization. Subsequently, structural and functional phenotypes, including liver proton density fat fraction (PDFF), liver iron, and functional phenotypes extracted from lung spirometry, were examined from over 20,000 subjects in the UK Biobank[6, 7]. Proteomics, canonical correlation analysis, genome-wide association study (GWAS) and post-GWAS analysis were employed to evaluate the predictive potential of the lung-liver axis for T2D[8]. Finally, to experimentally validate the bioinformatic findings, multi-channel liver- and lung-on-a-chip microphysiological systems were constructed to simulate T2D-associated high-glucose and inflammatory environments.
Results:
Pairwise correlation of [18F]-FDG PET uptake revealed that a robust metabolic synchronization across visceral organs occurs between the liver and lung (R = 0.66, p < 2.2e-16). Mediation analysis of the PET data demonstrated a unidirectional cascade: systemic blood glucose significantly modulates hepatic activity, which in turn acts as the primary driver of pulmonary glucose utilization, suggesting the role of the liver as a metabolic gatekeeper in the liver-lung axis.
At the population level, multi-organ MRI revealed that elevated liver PDFF is significantly associated with worsened lung spirometry performance. A generated lung-liver composite phenotype demonstrated high predictive capacity for incident T2D (AUC = 0.73), comparable to pancreatic markers, with Mendelian randomization confirming a causal relationship. Bayesian colocalization of GWAS data identified a shared genetic architecture between the axis and T2D at the 2p23.3 locus. This locus maps to the GCKR variant, an enzyme linked to glucose sensing, insulin release, and metabolic dysregulation, with multi-cohort evidence also associating GCKR with lung function[9].
In vitro validation using the microfluidic liver-lung-on-a-chip successfully modeled this directionality. When upstream hepatic tissue was subjected to T2D-simulated metabolic (high glucose) or endotoxic (LPS) stress, downstream pulmonary tissue exhibited a dramatically amplified inflammatory response, evidenced by spiked IL-6 secretion. Conversely, applying the initial stress to pulmonary tissue resulted in a blunted inflammatory and metabolic response in downstream hepatic tissue, demonstrating hepatic resilience to lung-derived stress [10].
Conclusion:
Molecular PET imaging identified a fundamental, hierarchical metabolic coupling between the liver and lung. By translating these in vivo findings into a controlled microphysiological model, this study confirms a directional amplification of systemic stress: the liver actively converts metabolic dysfunction into secondary pulmonary inflammation, while remaining resilient to reciprocal lung-derived stress. These observations establish hepato-pulmonary crosstalk as a genetically determined, directional driver of T2D, providing a novel framework for multi-organ therapeutic targeting.
Image/Figure:
Click to view full size
Click to view full size
Image/Figure Caption:
Figure. Schematic for deciphering the role of a liver-lung metabolic axis in Type 2 Diabetes.
(1) Bioinformatics analysis: Multi-modality in vivo imaging establishes the foundation of the liver-lung axis. [18F]-FDG PET analysis reveals a unidirectional metabolic hierarchy where systemic blood glucose dictates pulmonary metabolism primarily through hepatic mediation. Concurrently, population-scale MRI/CT and proteomics map the structural and immune-metabolic consequences of this axis, while canonical correlation analysis (CCA) and GWAS define a composite lung-liver phenotype with genetically inferred causality for incident Type 2 Diabetes (T2D). (2) Relationship query: This multi-modal population evidence highlights a critical mechanistic gap regarding the specific inter-organ crosstalk—specifically the directional flow of metabolic and inflammatory responses—that drives the hepato-pulmonary axis during disease progression. (3) Experimental exploration: To definitively isolate and validate these mechanisms, a customized microphysiological platform was engineered. Single liver- and lung-on-a-chip models were constructed to simulate controlled physiological and pathological environments (Control, T2DM, Inflammation, and combined T2DM/Inflammation). Cross-chip interaction protocols and downstream biomarker analyses were utilized to map the inter-organ inflammatory amplification and metabolic crosstalk, validating the macroscopic imaging findings in vitro.
Author
Yidan Gao
Fudan University