Introduction
Metabolic dysfunction-associated steatotic liver disease (MASLD) is a chronic liver condition estimated to affect one-third of the global adult population. As metabolic dysfunction-associated steatosis (MASL) progresses to steatohepatitis (MASH), the self-renewal capacity of Kupffer cells (KCs) becomes impaired, leading to a gradual depletion of the KC population. However, the phenotypic changes of the remaining KCs during disease progression remain poorly understood.
Aims & Methods
This study aims to characterise dynamic changes in KC identity during MASLD progression using a multi-omics approach, focusing on functional states and spatial organisation within the liver microenvironment.Spatial proteomic profiling was performed using Lunaphore COMET on MASLD liver samples obtained from UZ Leuven, Belgium (n = 8), enabling spatial phenotypic characterisation across different MASLD stages. High-resolution images were processed using the HORIZON software (Lunaphore, v2.5.0.0). Cell annotation and quantification were performed in QuPath (v0.5.1). Within CD163⁺ population, KCs were defined based on co-expression of CD5L, MARCO, and TIMD4. For each phenotypic markers, classifiers were developed to identify positive staining using Random Forest. Publicly available single-nucleus RNA sequencing data (n = 18) was analysed to assess KC-associated gene signatures across early and advanced disease stages. Association of key MASL- and MASH-associated KC markers with clinical features was evaluated using bulk RNA sequencing data from the European MASLD registry (n = 206).
Results
Multiplex fluorescent proteomic analysis of human liver biopsies across the MASLD spectrum revealed dynamic alterations in KC phenotypes during disease progression. Within the CD163⁺ population, we observed a gradual decrease in CD5L⁺MARCO⁺TIMD4⁺ KCs normal histology (63.2%) to MASL (36%) and further in MASH, with a progressive decrease across fibrosis stages (25.8% in F1 and 18.0% in F3). Interestingly, MASH samples showed an increased proportion of CD5L⁺MARCO⁻TIMD4⁻ cells compared to earlier stages with a subset co-expressing LYZ, suggesting the presence of monocyte-derived cells. Loss of KC identity was associated with increased CD45⁺ immune infiltration in the portal tract, parenchymal remodeling, hepatocyte marker loss (KRT8/18), and stellate cell activation (αSMA) with collagen deposition (COL1A1, COL1A3). snRNA-seq analysis further confirmed the gradual decrease in the number of CD5L⁺MARCO⁺TIMD4⁺ within CD163⁺ population from normal histology to MASL and MASH. Differential expression analysis comparing KCs from MASL and MASH patients identified 131 differentially expressed genes. MASH- associated KCs showed enrichment in genes important for cholesterol metabolism (FABP1, LPL, SAA1, SAA2 and PLTP), cell–cell junction formation (FMN1, DTNA), and complement regulation (CRP), while downregulating mitochondrial respiratory chain genes (MT-CYB, MT- ND2, MT-ND4 and MT-ND5), glucose metabolism (G6PC), and CYP2C19, suggesting metabolic reprogramming of KCs characterised by enhanced lipid-associated pathways and impaired mitochondrial function. Bulk RNA-seq analysis showed that key MASH-associated KC markers correlate with fibrosis stage, disease activity, FIB-4 score, and serum ALT/AST levels.
Conclusion
This study identifies dynamic changes in KC identity across MASLD progression, characterised by a shift towards a lipid-associated, inflammatory phenotype in MASH. Furthermore, key MASH-associated KC markers demonstrate clinical relevance for disease staging and activity across the MASLD spectrum.