Search

Additional file 2: Table S1. Key parameters of the AFADESI-MSI setting.

Additional file 1: Figure S1. Total metabolite distributions in the HC, LC and HCC tissue samples. a HE and MSI diagrams of HC, LC and HCC whole samples under the negative ionization mode. b PLS-DA comparison of AFADESI-MSI data under the negative ionization mode. c, d Alterations of metabolites detected by AFADESI-MSI method in the HC, LC and HCC samples based on positive (c) and negative (d) ionization modes. e, f K-means diagrams of the HC, LC and HCC tissue samples based on positive (e) and negative (f) ionization modes. Figure S2. The alterations of metabolites’ spatial distributions in HCC sub-regions under negative ionization mode. a HE and MSI images of different sub-regions of HCC. b PLS-DA analysis of different HCC sub-regions. c Number of metabolites detected from different HCC sub-regions. d Heatmap of significantly differentiated metabolites based on VIP > 1. e KEGG analysis of key altered metabolic pathways in the sub-regions of HCC. Figure S3. Metabolites’ spatial distributions in the non-cancerous regions (CV) in the HC, LC and HCC samples under negative ionization mode. a HE and MSI images of the CV regions in the HC, LC and HCC samples. b PLS-DA analysis for the CV regions. c Heatmap of significantly differentiated metabolites under positive ionization mode in the CV regions based on VIP > 1. d KEGG analysis of key altered metabolic pathways in the CV regions. e Heatmap of significantly differentiated metabolites based on VIP > 1 under negative ionization mode. f Sample time series analysis of the key metabolite expressions in the CV regions. Figure S4. Metabolites’ spatial distributions of the cancer-related regions in the HC, LC and HCC samples under negative ionization mode. a HE and MSI images (negative mode) of the cancer-related regions in the HC, LC and HCC samples. b PLS-DA analysis of the cancer-related regions. c Heatmap of significantly differentiated metabolites in the cancer-related regions based on variable VIP > 1 under positive ionization mode. d Heatmap of significantly differentiated metabolites in the cancer-related regions based on variable VIP > 1 under negative ionization mode. e KEGG analysis of key altered metabolic pathways in cancer-related regions. Figure S5. Metabolic changes in different tissues and its predictability of disease progression. a Sample time series analysis of key metabolite expressions (negative mode) in the cancer-related regions. b Examples of key metabolites’ spatial expressions in cancer-related regions under negative ion mode. c ROC analyses of key metabolites at distinguishing between diseased tissues and between different subregions (anabasine was detected under positive ionization mode, docosahexaenoic acid was detected under negative ionization mode).

Mitochondria play leading roles in initiation and progression of colorectal cancer (CRC). Proteogenomic analyses of mitochondria of CRC tumor cells would likely enhance our understanding of CRC pathogenesis and reveal new independent prognostic factors and treatment targets. However, comprehensive investigations focused on mitochondria of CRC patients are lacking. Here, we investigated global profiles of structural variants, DNA methylation, chromatin accessibility, transcriptome, proteome, and phosphoproteome on human CRC. Proteomic investigations uncovered greatly diminished mitochondrial proteome size in CRC relative to that found in adjacent healthy tissues. Integrated with analysis of RNA-Seq datasets obtained from the public database containing mRNA data of 538 CRC patients, the proteomic analysis indicated that proteins encoded by 45.5% of identified prognostic CRC genes were located within mitochondria, highlighting the association between altered mitochondrial function and CRC. Subsequently, we compared structural variants, DNA methylation, and chromatin accessibility of differentially expressed genes and found that chromatin accessibility was an important factor underlying mitochondrial gene expression. Furthermore, phosphoproteomic profiling demonstrated decreased phosphorylation of most mitochondria-related kinases within CRC versus adjacent healthy tissues, while also highlighting MKK3/p38 as an essential mitochondrial regulatory pathway. Meanwhile, systems-based analyses revealed identities of key kinases, transcriptional factors, and their interconnections. This research uncovered a close relationship between mitochondrial dysfunction and poor CRC prognosis, improve our understanding of molecular mechanism underlying mitochondrial linked to human CRC, and facilitate identifies of clinically relevant CRC prognostic factors and drug targets.

Additional file 2. The cox analyses of HIGD1A, SUCLG2 and SLC25A24; The overall survival rates of HIGD1A, SUCLG2 and SLC25A24 in colon and rectum cancer.

Additional file 1: Supplementary Figure 1. (A) UMAP plots for the disease and control groups. (B) UMAP cluster to perform dimensionality reduction and clustering on the ES and NC data. (C) Expression of genes in the ES and NC groups in cluster 0 of megakaryocyte erythroid cells that participate in the human T cell leukemia virus 1 infection KEGG pathway. (D) Expression of genes in the ES and NC groups in cluster 1 of megakaryocyte erythroid cells that participate in the apoptosis KEGG pathway.