New technique unlocks protein secrets hidden in cancer tissue archives
Researchers have developed a faster, more reliable way to extract and analyze proteins from archived melanoma samples, revealing that acral melanoma tumors have distinct metabolic signatures. The breakthrough could accelerate cancer diagnostics and drug development by making it feasible to screen thousands of stored tissue samples at scale.
Originaltitel: High-throughput acoustic FFPE proteomics reveals ATP5IF1-associated mitochondrial alterations in acral melanoma.
Forskare vid Lund University och National Autonomous University of Mexico har utvecklat en höggenomströmnings-proteinanalys direkt från rutinhistologiska vävnadsprover, utan tidigare tids- och arbetskrävande förbehandling. Metoden analyserade 40 melanomfall och identifierade över 8 200 proteingrupper per tumor. Akral lentiginös melanom — en ovanlig, aggressiv hudcancerform — uppvisade karakteristisk proteinprofil med minskad ATP5IF1-nivå, en central regulator av mitochondriell energiproduktion. Detta samband indikerar omställning från normal cellrespiration till ökad glukosförbranning, en erkänd överlevnadsmekanism för cancerceller. Metodens format — 96-brunnsplatta, automatiserad — möjliggör retrospektiv analys av befintliga vävnadsarkiv utan nya provsamlingar. För sjukhuspatologer och inköpschefer innebär detta potential att snabbt klassificera melanomsubtyppr och identifiera behandlingsmål. Lund Universitys deltagande stärker svensk MedTech-position inom diagnostik.
Formalin-fixed, paraffin-embedded (FFPE) archives underpin dermatopathology and translational oncology, enabling clinically annotated melanoma cohorts, but cohort-scale proteomics remains limited by labor and variability in upstream processing. We established a plate-scale, acoustic FFPE proteomics workflow and integrated it with AI-assisted digital pathology to support composition-aware molecular profiling from routine sections. The optimized pipeline reduces handling steps, is designed to improve reproducibility, and supports rapid parallel processing in a 96-well format. Deep data-independent acquisition mass spectrometry of 40 primary melanomas spanning acral lentiginous, lentigo maligna, superficial spreading, and nodular subtypes quantified more than 8,200 protein groups and a mean of 5,200 proteins per tumor. Proteome profiles resolved melanoma subtypes and defined an acral lentiginous melanoma program enriched for translation/biogenesis and extracellular matrix/adhesion processes with relative depletion of lipid and fatty-acid metabolism and peroxisomal pathways. Supervised feature selection highlighted TNC, POSTN, EIF4A1, and ARF4 and uncovered selective depletion of ATP5IF1, a regulator of mitochondrial ATP synthase, in acral lentiginous melanoma. ATP5IF1 depletion persisted after adjustment for mitochondrial proxies and QuPath-derived tumor content, and coincided with higher glycolysis relative to Complex V. This pathology-integrated, high-throughput FFPE proteomics framework enables scalable retrospective discovery and suggests subtype-specific metabolic alterations consistent with mitochondrial remodeling in an underrepresented melanoma subtype.