The Spatial Extracellular Proteomic Tumor Microenvironment Distinguishes Molecular Subtypes of Hepatocellular Carcinoma

• DOI: 10.1021/acs.jproteome.4c00099
• Publication Date: 2024-07-09T17:29:06Z
• Authors: Jade K. Macdonald; Harrison B. Taylor; Mengjun Wang; Andrew Delacourt; Christin Edge; David N. Lewin; Naoto Kubota; Naoto Fujiwara; Fahmida Rasha; Cesia A. Marquez; Atsushi Ono; Shiro Oka; Kazuaki Chayama; Sara Lewis; Bachir Taouli; Myron Schwartz; M Isabel Fiel; Richard R. Drake; Yujin Hoshida; Anand S. Mehta; Peggi M. Angel
• Keywords: Proteome; Discoidin domain

Hepatocellular carcinoma (HCC) mortality rates continue to increase faster than those of other cancer types due high heterogeneity, which limits diagnosis and treatment. Pathological molecular subtyping have identified that HCC tumors with poor outcomes are characterized by intratumoral collagenous accumulation. However, the translational post-translational regulation tumor collagen, is critical outcome, remains largely unknown. Here, we investigate spatial extracellular proteome understand differences associated defined Hoshida transcriptomic subtypes outcome (Subtype 1; S1; n = 12) better 3; S3; 24) show differential stroma-regulated pathways. Collagen-targeted mass spectrometry imaging (MSI) same-tissue reference libraries, built from untargeted targeted LC-MS/MS was used spatially define microenvironment clinically-characterized, formalin-fixed, paraffin-embedded tissue sections. Collagen α-1(I) chain domains for discoidin-domain receptor integrin binding showed distinctive distribution within microenvironment. Hydroxylated proline (HYP)-containing peptides triple helical regions fibrillar collagens distinguished S1 S3 tumors. Exploratory machine learning on multiple extracted could distinguish (with an area under receiver operating curve ≥0.98; 95% confidence intervals between 0.976 1.00; accuracies above 94%). An overall finding has a potential predict clinically relevant in HCC.