Insights from a Computational-Based Approach for Analyzing Autophagy Genes across Human Cancers

Abstract

In the last decade, there has been a boost in autophagy reports due to its role in cancer progression and its association with tumor resistance to treatment. Despite this, many questions remain to be elucidated and explored among the different tumors. Here, we used omics-based cancer datasets to identify autophagy genes as prognostic markers in cancer. We then combined these findings with independent studies to further characterize the clinical significance of these genes in cancer. Our observations highlight the importance of innovative approaches to analyze tumor heterogeneity, potentially affecting the expression of autophagy-related genes with either pro-tumoral or anti-tumoral functions. In silico analysis allowed for identifying three genes (TBC1D12, KERA, and TUBA3D) not previously described as associated with autophagy pathways in cancer. While autophagy-related genes were rarely mutated across human cancers, the expression profiles of these genes allowed the clustering of different cancers into three independent groups. We have also analyzed datasets highlighting the effects of drugs or regulatory RNAs on autophagy. Altogether, these data provide a comprehensive list of targets to further the understanding of autophagy mechanisms in cancer and investigate possible therapeutic targets.

Keywords: autophagy; cancer; cancer dataset; cell phenotype; systems biology.

Figure 1

Autophagy-related genes have differentiated expressions in solid tumors. Heatmaps showing the following values: (A) Log₂ fold-change (L2FC) in the medians of expression levels from normal to tumor tissues among TCGA participants; (B) L2FC in the medians of expression levels from normal to tumor tissues among TCGA + GTEX + TARGET participants; (C) medians of normal to tumor tissues L2FC matched by TCGA participants. White cells represent genes without statistical differences between tumor and normal (or normal-adjacent) tissues. The statistical test applied were Mann–Whitney’s test (A,B) and Wilcoxon’s test (C). BLCA: bladder urothelial carcinoma; BRCA: breast invasive carcinoma; CHOL: cholangiocarcinoma; COAD: colon adenocarcinoma; ESCA: esophageal carcinoma; GBM: glioblastoma multiforme; HNSC: head and neck squamous cell carcinoma; KICH: kidney chromophobe; KIRC: kidney renal clear cell carcinoma; KIRP: kidney renal papillary cell carcinoma; LGG: brain lower grade glioma; LIHC: liver hepatocellular carcinoma; LUAD: lung adenocarcinoma; LUSC: lung squamous cell carcinoma; OV: ovarian serous cystadenocarcinoma; PRAD: prostate adenocarcinoma; READ: rectum adenocarcinoma; SKCM: skin cutaneous melanoma; STAD: stomach adenocarcinoma; TGCT: testicular germ cell tumors; THCA: thyroid carcinoma; UCEC: uterine corpus endometrial carcinoma; COADREAD: colorectal adenocarcinoma (COAD + READ); KIPAN: pan-kidney cohort (KICH + KIRC + KIRP); STES: stomach and esophageal carcinoma (STAD + ESCA).

Figure 2

Autophagy-related genes can stratify solid tumors. (A) Clusterization of solid tumors based on the differential expression of autophagy genes. After a UMAP analysis, it is possible to recognize three classifications (B) of relevant tumors based on the expression of autophagy genes. BRCA: breast invasive carcinoma; COAD: colon adenocarcinoma; ESCA: esophageal carcinoma; GBM: glioblastoma multiforme; KIRC: kidney renal clear cell carcinoma; KIRP: kidney renal papillary cell carcinoma; LGG: brain lower grade glioma; LIHC: liver hepatocellular carcinoma; LUAD: lung adenocarcinoma; LUSC: lung squamous cell carcinoma; OV: ovarian serous cystadenocarcinoma; PAAD: pancreatic adenocarcinoma; PRAD: prostate adenocarcinoma; SKCM: skin cutaneous melanoma; STAD: stomach adenocarcinoma; TGCT: testicular germ cell tumors.

Figure 3

Genes upregulated in Cluster “0” differentiate tumor and normal adjacent tissues. Using the UALCAN tool, we compared a selection of genes stratifying solid tumors in Cluster “0” between tumor and normal-adjacent tissues. Herein, we represent data for BRCA, KIRC, LIHC, and PRAD datasets for the MAPT (A–D), NUPR1 (E–H), and TP53INP1 (I–L) genes. *** represents comparisons with p-value < 0.001 on Welch’s t-test. BRCA: breast invasive carcinoma; KIRC: kidney renal clear cell carcinoma; LIHC: liver hepatocellular carcinoma; PRAD: prostate adenocarcinoma.

Figure 4

Putative autophagy-related gene markers codify dysregulated proteins in LIHC. Data from Clinical Proteomic Tumor Analysis Consortium (CPTAC) and the International Cancer Proteogenome Consortium (ICPC) datasets via the UALCAN tool allow us to confirm putative gene markers upregulated in LIHC with their proteic version upregulated. Here is the shown data for EEF1A2 (A), MAPT (B), and NUPR1 (C) proteins. *** represents comparisons with p-value < 0.001 on Welch’s t-test. LIHC: liver hepatocellular carcinoma.

Figure 5

Putative autophagy-related gene markers are rarely mutated in solid tumors of Cluster “0”. Oncoprint produced by the cBioPortal for Cancer Genomics shows the frequency of somatic mutations per gene and cancer dataset related to Cluster “0”. Notably, a group of BRCA patients showed amplifications of all genes, whereas some PRAD patients showed deletions in MAPT, PRKAA2, and TUBA3E genes. * means that the mutational frequency was estimated about the number of profiled patients as this number can vary between genes.

Figure 6

Genes upregulated in Clusters “0” and “1” differentiate tumor and normal adjacent tissues. Using the UALCAN tool, we compared a selection of genes stratifying solid tumors in Clusters “0” and “1” between tumor and normal-adjacent tissues. Herein, we represent data for KIRC, KIRP, LUAD, and STAD datasets for the SREBF1 (A–D), TUBA3D (E–H), and FBXW7 (I–L) genes. p-values on Welch’s t-test are shown as *** (p < 0.001); * (p < 0.05); n.s. (p ≥ 0.05). KIRC: kidney renal clear cell carcinoma; KIRP: kidney renal papillary cell carcinoma; LUAD: lung adenocarcinoma; STAD: stomach adenocarcinoma.

Figure 7

ACBD5 protein levels in tumors belonging to Clusters “0” and “1”. Data from Clinical Proteomic Tumor Analysis Consortium (CPTAC) and the International Cancer Proteogenome Consortium (ICPC) datasets via the UALCAN tool allow us to confirm dysregulated levels of the ACBD5 protein in three tumor tissues (compared with their respective non-tumor adjacent tissues). Here is the shown data for BRCA (A), COAD (B), and LUAD (C) datasets. *** represents comparisons with p-value < 0.001 on Welch’s t-test. BRCA: breast invasive carcinoma; COAD: colon adenocarcinoma; LUAD: lung adenocarcinoma.

Figure 8

Putative autophagy-related gene markers are rarely mutated in solid tumors of clusters “0” and “1”. Oncoprint produced by the cBioPortal for Cancer Genomics shows the frequency of somatic mutations per gene and cancer dataset related to the clusters “0” and “1”. Notably, the FBXW7 accounts for the higher mutational frequency, mainly in the COAD dataset, whereas the PRAD cohort shows a high percentage of patients with deleted regions of analyzed genes. * means that the mutational frequency was estimated about the number of profiled patients as this number can vary between genes.