Use of Arabidopsis thaliana as a model to understand specific carcinogenic events: Comparison of the molecular machinery associated with cancer-hallmarks in plants and humans

Abstract

Model organisms are fundamental in cancer research given that they rise the possibility to characterize in a quantitative-objective fashion the organisms as a whole in ways that are infeasible in humans. From this perspective, model organisms with short generation times and established protocols for genetic manipulation allow the understanding of basic biology principles that might guide carcinogenic onset. The cancer-hallmarks (CHs) approach, a modular perspective for cancer understanding, stands that underlying the variability among different cancer types, critical events support the carcinogenic origin and progression. Thus, CHs as interconnected genetic circuitry, have a causal effect over cancer biogenesis and might represent a comparison scaffold among model organisms to identify and characterize evolutionarily conserved modules to understand cancer. Nevertheless, the identification of novel cancer regulators by comparative genomics approaches relies on selecting specific biological processes or related signaling cascades that limit the type of detected regulators, even more, holistic analysis from a systemic perspective is absent. Similarly, although the plant Arabidopsis thaliana has been used as a model organism to dissect specific disease-associated mechanisms, given the evolutionary distance between plants and humans, a general concern about the utility of using A. thaliana as a cancer model persists. In the present research, we take advantage of the CHs paradigm as a framework to establish a functional systemic comparison between plants and humans, that allowed the identification not only of specific novel key genetic regulators, but also, biological processes, metabolic systems, and genetic modules that might contribute to the neoplastic transformation. We propose five cancer-hallmarks that overlapped in conserved mechanisms and processes between Arabidopsis and human and thus, represent mechanisms which study can be prioritized in A. thaliana as an alternative model for cancer research. Additionally, derived from network analyses and machine learning strategies, a new set of potential candidate genes that might contribute to neoplastic transformation is described. These findings postulate A. thaliana as a suitable model to dissect, not all, but specific cancer properties, highlighting the importance of using alternative complementary models to understand carcinogenesis.

Keywords: Arabidopsis thaliana; Cancer hallmarks; Comparative genomics; Machine learning; Model organisms for cancer research; Network biology.

Fig. 1

Implemented methodological workflow: A three-stages methodology was designed to establish a framework for the identification of conserved molecular processes and novel associated genes to understand the carcinogenic processes from a comparative genomics perspective. Stage 1: identification of orthologous genes linked to the CHs in A. thaliana, coupled to a gene ontology enrichment analysis to find the biological processes in which CHs-associated genes are functionally associated, and metabolic inference to characterize in which metabolic reactions CHs-associated genes can participate. Stage 2: obtention of Arabidopsis orthologs that could be suggested to further study specific cancer properties, implementing topological analyses over PPI networks and machine learning approaches, and stage 3: A differential expression analysis was carried out in order to validate with available transcriptomics data the potential of the detected genes of being involved in carcinogenic events.

Fig. 2

Cancer-hallmarks functional similarities according to gene sets composition. Principal coordinates analysis (PCoA) biplot using Jaccard distances for each of the ten cancer-hallmarks, analyzed in A. thaliana (green) and H. sapiens (blue). The closer two CHs appear plotted, the more similar GO-associated terms conformation they exhibit. AID (Avoiding Immune Destruction), AIM (Activating Invasion and metastasis), DCE (Deregulating Cellular Energetics), ERI (Enabling Replicative Immortality), EGS (Evading Growth Suppressors), GIM (Genome Instability Mutation), IA (Inducing Angiogenesis), RCD (Resisting Cell Death), SPS (Sustaining Proliferative Signaling) and TPI (Tumor Promoting Inflammation).

Fig. 3

Functional comparison among cancer-hallmarks. GO terms counts (BP) associated with each CH. Light blue: unique GO Terms for H. sapiens. Light green: Unique GO terms found for A. thaliana and Orange: shared GO terms between species. AID (Avoiding Immune Destruction), AIM (Activating Invasion and metastasis), DCE (Deregulating Cellular Energetics), ERI (Enabling Replicative Immortality), EGS (Evading Growth Suppressors), GIM (Genome Instability Mutation), IA (Inducing Angiogenesis), RCD (Resisting Cell Death), SPS (Sustaining Proliferative Signaling) and TPI (Tumor Promoting Inflammation).

Fig. 4

Metabolic classification for enzymatic reactions present in the gene sets associated with different cancer-hallmarks. According to BioCyc platform, the CHs associated genes were filtered out, exclusively selecting genes that code for proteins with catalytic activity. The enzyme-coding genes were classified into metabolic systems, subsystems, and sub-subsystems for A.H. sapiens and B.A. thaliana. Available at https://ccsosa.github.io/TEST_CH.

Fig. 5

Cancer-hallmarks associated genes mapped over the most connected fraction of the generated interactomes. Bars represent gene-counts for each CH. A. thaliana (green) and H. sapiens (blue). AID (Avoiding Immune Destruction), AIM (Activating Invasion and metastasis), DCE (Deregulating Cellular Energetics), ERI (Enabling Replicative Immortality), EGS (Evading Growth Suppressors), GIM (Genome Instability Mutation), IA (Inducing Angiogenesis), RCD (Resisting Cell Death), SPS (Sustaining Proliferative Signaling) and TPI (Tumor Promoting Inflammation).

Fig. 6

Gene expression analysis in thirty-one contrasting tissues. Codifying genes from the human proteins listed in Table 5, Table 6, Table 7, were contrasted against the GEPIA2 database to analyze gene expression changes when comparing normal and tumoral tissues: control tissue (N) and tumoral tissue (T). Rows are equivalent to genes and columns to tissues. The expression values in each cell are measured in log2 (TPM)+1 and were extracted only from those tissues with statistical significance (p < 0.05 and Log2FC = 1). Yellow indicates high expression and purple shows low expression. Adrenocortical carcinoma (ACC), Bladder Urothelial Carcinoma (BLCA), Breast invasive carcinoma (BRCA), Cervical squamous cell carcinoma (CESC), Cholangio carcinoma (CHOL), Colon adenocarcinoma (COAD), Diffuse Large B-cell Lymphoma (DLBC), Esophageal carcinoma (ESCA), Glioblastoma multiforme (GBM), Head and Neck squamous cell carcinoma (HNSC), Kidney Chromphobe (KICH), Kidney renal clear cell carcinoma (KIRC), Kidney renal papillary cell carcinoma (KIRP), Acute Myeloid Leukemia (LAML), Brain Lower Grade Glioma (LGG), Liver hepatocellular carcinoma (LIHC), Lung adenocarcinoma (LUAD), Lung squamous cell carcinoma (LUSC), Ovarian serous cystadenocarcinoma (OV), Pancreatic adenocarcinoma (PADD), Pheochromocytoma and Paraganglioma (PCPG), Prostate adenocarcinoma (PRAD), Rectum adenocarcinoma (READ), Sarcoma (SARC), Skin Cutaneous Melanoma (SKCM), Stomach adenocarcinoma (STAD), Testicular Germ Cell Tumors (TGCT), Thyroid carcinoma (THCA), Thymoma (THYM), Uterine Corpus Endometrial Carcinoma (UCEC), and Uterine Carcinosarcoma (UCS).