Examining Sporadic Cancer Mutations Uncovers a Set of Genes Involved in Mitochondrial Maintenance

Abstract

Mitochondria are key organelles for cellular health and metabolism and the activation of programmed cell death processes. Although pathways for regulating and re-establishing mitochondrial homeostasis have been identified over the past twenty years, the consequences of disrupting genes that regulate other cellular processes, such as division and proliferation, on affecting mitochondrial function remain unclear. In this study, we leveraged insights about increased sensitivity to mitochondrial damage in certain cancers, or genes that are frequently mutated in multiple cancer types, to compile a list of candidates for study. RNAi was used to disrupt orthologous genes in the model organism Caenorhabditis elegans, and a series of assays were used to evaluate these genes' importance for mitochondrial health. Iterative screening of ~1000 genes yielded a set of 139 genes predicted to play roles in mitochondrial maintenance or function. Bioinformatic analyses indicated that these genes are statistically interrelated. Functional validation of a sample of genes from this set indicated that disruption of each gene caused at least one phenotype consistent with mitochondrial dysfunction, including increased fragmentation of the mitochondrial network, abnormal steady-state levels of NADH or ROS, or altered oxygen consumption. Interestingly, RNAi-mediated knockdown of these genes often also exacerbated α-synuclein aggregation in a C. elegans model of Parkinson's disease. Additionally, human orthologs of the gene set showed enrichment for roles in human disorders. This gene set provides a foundation for identifying new mechanisms that support mitochondrial and cellular homeostasis.

Keywords: Caenorhabditis elegans; bioinformatics; cancer; development; mitochondria; mitophagy; neurodegeneration; pps-1.

Figure 1

Identification of a gene set underlying mitochondrial health. (A) Flowchart of testing process for genes identified from the NCI-60 hit list. The flowchart indicates the number of genes tested at each stage. (B) A chart illustrating the 1.4-fold cutoff (indicated by a red line) used to obtain the hits from screening for mitophagic activation. (C) A chart showing the 1.5-fold cutoff (indicated by a red line) used to identify hits in the cytotoxicity screen. (D) Visualization of the initial 50 gene ‘seed set’, showing 38 connections between the genes. (E) Connectivity of one representative set (of 100 sets) of 50 randomly-selected genes from the C. elegans genome, showing a total of three connections. p-values (D,E) were calculated by the WormNetV3 database and include a Bonferroni correction. (F) Bioinformatic analysis of the 50 gene set showed enrichment for germ cell proliferation, negative regulation of differentiation, and autophagic flux (autophagosome assembly, macroautophagy, response to starvation, etc.).

Figure 2

Refinement of the mitochondrial health gene set. (A) Flowchart for testing the 381 new genes identified by their connection to the seed set. (B) Connectivity of the expanded 139 gene set. (C) Connectivity of one representative set (of 100 sets) of 140 randomly selected genes from the C. elegans genome, showing a total of ~40 connections. p-values (B,C) were calculated by the WormNetV3 database and include a Bonferroni correction. (D) Bioinformatic analysis of the 139 gene set showed enrichment for cell proliferation and development, as well as signal transduction.

Figure 3

Genes in the mitochondrial health gene set support mitochondrial network connectivity. (A) Fluorescence micrographs of mitochondrially-targeted GFP (GFP^mt) expressed in body wall muscle cells showing four different qualitative states. Shown are filamentous (top left), intermediate fragmentation (upper right), highly punctate (bottom left), and completely punctate (bottom right). Scale bar represents 5 μm. (B) Quantification of fragmentation state of worms expressing GFP^mt in body wall muscles after RNAi targeting selected genes. Approximately 90 worms were examined for each genotype. All genes showed an increase in mitochondrial fragmentation after RNAi. Statistical significance was determined by χ² analysis, *** p < 0.001.

Figure 4

Genes in the mitochondrial health gene set support healthy mitochondrial surveillance. (A) Brightfield and GFP images of the 3XESRE::GFP reporter are quantified to compare ESRE expression of the validation RNAi panel. (B) Brightfield and GFP images of the Ptbb-6::GFP reporter are quantified to compare MAPK^mt expression across the RNAi validation panel, with pmk-3 being the positive control. (C) Brightfield and GFP images of the Phsp-6::GFP reporter are quantified to compare UPR^mt expression across the RNAi validation panel, with atfs-1 being the positive control. Statistical significance in all panels was calculated using one-way ANOVA analysis, followed by a Dunnett’s post hoc test. Statistical significance in all panels was calculated using one-way ANOVA analysis, followed by a Dunnett’s post hoc test. * p < 0.05, ** p < 0.01, *** p < 0.001. * indicates significant increase compared to EV, * indicates significant decrease compared to EV.

Figure 5

Genes in the mitochondrial health gene set support mitochondrial health. (A) Bar graph of mitochondrial mass, as measured by mitochondrial staining with MitoTracker Green FM in young adult worms reared on RNAi targeting the indicated gene. Fluorescence was measured using flow vermimetry. All subsequent mitochondrial functional assays were normalized to mitochondrial mass. (B) Bar graph of steady-state ratio of NADH–NAD+ using a conformation-dependent fluorometric reporter. Young adult worms carrying the reporter were reared on the indicated RNAi constructs. (C) Bar graph of steady-state reactive oxygen species, as measured by conversion of H₂DCFDA to a fluorescent state in young adult worms reared on RNAi targeting the indicated gene. Fluorescence was measured using Cytation5 Cell Imaging Multi-Mode Reader. (D) Bar graph of mitochondrial membrane potential, as measured by the dye MitoTracker Red CMXRos in young adult worms reared on RNAi targeting the indicated gene. Fluorescence was measured using flow vermimetry. (E) Bar graph of basal oxygen consumption in young adult worms reared on RNAi targeting the indicated gene as measured by a Clark electrode. Statistical significance in all panels was calculated using one-way ANOVA analysis followed by a Dunnett’s post hoc test. * p < 0.05, ** p < 0.01, *** p < 0.001. * indicates significant increase compared to EV, * indicates significant decrease compared to EV.

Figure 6

Genes in the mitochondrial health gene set support organism-level health parameters. (A) Bar graph of 8-day survival of worms reared on RNAi targeting the indicated gene. (B) Quantification of length of young adult worms reared on RNAi targeting the indicated gene. Worm length was measured using flow vermimetry. Circles represent average size per biological replicate. (C) Bar graph of worm fecundity in adult worms reared on RNAi targeting the indicated gene. (D) Bar graph of pharyngeal pumping rates for young adult worms reared on RNAi targeting the indicated gene. Statistical significance was calculated using one-way ANOVA analysis followed by Dunnett’s post hoc test. * p < 0.05, *** p < 0.001. * indicates significant increase compared to EV, * indicates significant decrease compared to EV.

Figure 7

Genes in the mitochondrial health gene set reduce proteostatic aggregation. (A) Representative fluorescence micrographs of a transgenic C. elegans strain containing an α-synuclein::YFP protein in worms reared on RNAi targeting the indicated gene. (B) Quantification of α-synuclein::YFP aggregates in young adult worms reared on the indicated RNAi. Statistical significance was calculated using one-way ANOVA analysis followed by a Dunnett’s post hoc test. * p < 0.05, *** p < 0.001. * indicates significant increase compared to EV, * indicates significant decrease compared to EV.

Figure 8

Human orthologs of the C. elegans mitochondrial health gene set illustrate increased co-association and are associated with disease. (A) A set of human genes comprising 153 orthologs of the C. elegans mitochondrial health gene set exhibits significant connectivity. (B) Connectivity of one representative set (of 100 sets) of 153 randomly selected genes from the human genome, showing a total of ~40 connections. p-values (B,C) were calculated by the STRING database. (C) Bioinformatic analysis of gene function for the human genes shows strong enrichment for autophagy and mitophagy, as well as cancer and cell division. Several pro-growth signaling pathways (such as p53, Wnt, Hedgehog, TGF-β, Notch, and ErbB signaling pathways) are also enriched. (D) Analysis of human orthologs also shows enrichment for a variety of diseases.