Breakthrough imaging platform detects disease markers before symptoms appear in tissues
Researchers have developed PathoPlex, a new multiplexed imaging technology capable of mapping over 140 proteins within cells at subcellular resolution. The framework identified kidney stress-related protein patterns in people with type 2 diabetes before clinical signs of disease appeared, potentially accelerating both diagnosis and treatment timelines.
A global research team has developed a groundbreaking imaging technology that can map the precise locations of more than 140 different proteins inside cells, offering unprecedented insights into disease processes before symptoms become apparent. The technology, called PathoPlex (pathology-oriented multiplexing), successfully identified kidney stress-related protein patterns in individuals with type 2 diabetes who showed no clinical signs of kidney disease.
Published in Nature, the research demonstrates how PathoPlex combines highly multiplexed imaging at subcellular resolution with sophisticated software to extract and interpret protein co-expression patterns across biological layers. The framework was optimised to map commercial antibodies at 80 nanometres per pixel across 95 iterative imaging cycles, enabling simultaneous processing of at least 40 archival biopsy specimens.
Advanced protein mapping reveals hidden disease patterns
“PathoPlex paves the way towards understanding and imaging complex tissues in human diseases like diabetes,” said Professor Matthias Kretzler from the University of Michigan’s Caswell Diabetes Institute. “We can finally develop atlases that describe changes in protein functions and how to improve them with new treatments.”
The technology addresses a fundamental challenge in medical research: understanding how protein organisation within tissues drives important functions and how abnormal changes in protein levels and patterns result in disease. Traditional imaging methods have been limited by the number of proteins they can simultaneously detect and their spatial resolution, typically restricting analysis to 30-60 antibodies at resolutions of 200-300 nanometres per pixel.
PathoPlex significantly advances these capabilities by enabling analysis of over 140 proteins at 80 nanometres per pixel resolution, generating datasets containing more than 600 billion available pixels. The tissues remained stable throughout 95 imaging cycles, suggesting this may not represent the technology’s limit.
Validation in experimental kidney disease models
To validate PathoPlex’s clinical potential, researchers first tested it on a well-characterised mouse model of immune-mediated kidney disease. Using 34 markers at 80 nanometres per pixel resolution, they identified cluster 21, featuring phosphorylated JUN (pJUN) as a top contributor, which was consistently increased in both acute and chronic disease states.
The spatial analysis revealed that this cluster was predominantly located in parietal epithelial cells and tubular cells, with increasing frequency in parietal epithelial cells during disease progression. Subsequent validation experiments confirmed that JNK inhibitor treatment reduced disease-associated cellular migration and significantly decreased proteinuria in preventive studies whilst mitigating glomerular damage in therapeutic interventions.
Clinical applications in diabetic kidney disease
The research team then applied PathoPlex to analyse clinical specimens from 38 individuals, comparing 18 controls with 20 patients with advanced diabetic kidney disease. Using 61 markers across 422 regions of interest, they generated over 100 billion pixels at 160 nanometres per pixel resolution.
This analysis identified 18 clusters with differential abundance between control and diabetic kidney disease samples. Notably, cluster 19, containing contributors from mitochondrial proteins AIFM1 and TRPC6, was increased in diabetic kidney disease tissues and localised primarily in proximal tubules – findings that correlated with conventional histopathological analysis.
Early detection capabilities demonstrated
Perhaps most significantly, PathoPlex successfully identified kidney stress-related changes in research biopsy samples from individuals with type 2 diabetes who showed no overt histopathological signs of kidney disease. The analysis included 18 specimens using 142 markers, revealing 24 significantly regulated clusters representing distinct biological processes.
The framework identified specific patterns including increases in stromal cell filopodia, mesangial matrix expansion, and vascular smooth muscle cells, alongside reductions in clusters associated with proximal tubule structural and functional features, peritubular capillary integrity, and mitochondrial function.
Treatment response assessment capabilities
The technology also demonstrated its potential for assessing therapeutic responses by analysing tissue samples from patients treated with SGLT2 inhibitors. PathoPlex revealed that these medications attenuated changes in peritubular capillaries and mitochondrial integrity in proximal tubules, whilst increasing gluconeogenesis markers.
However, the analysis showed that SGLT2 inhibitor treatment did not fully reverse diabetes-specific changes, suggesting “individuals with diabetes may benefit from additional interventions to reverse them to the healthy reference state,” according to the authors.
Democratising advanced tissue analysis
PathoPlex offers several practical advantages over existing multiplexed imaging approaches. The system requires no dependency on commercial equipment, provides open-access 3D printing-based solutions for sample preparation, and maintains compatibility with standard inverted fluorescence microscopes. The technology also introduces stringent quality-control measures and minimises batch effects through parallel processing of multiple clinical samples.
The research represents a significant advancement in spatial biology, moving beyond traditional cell-typing approaches towards data-driven identification of distinctive biological signatures across spatial scales based on individual pixel analysis.
Reference
Kuehl, M., Okabayashi, Y., Wong, M.N., et al. (2025). Pathology-oriented multiplexing enables integrative disease mapping. Nature, 644, 516-524. https://doi.org/10.1038/s41586-025-09225-2
