EXD2: A new regulator of mitochondrial translation and potential target for cancer therapy

ABSTRACT

In recent work we identified Exonuclease 3′−5′ domain-containing protein 2 (EXD2) as an RNase required for efficient mitochondrial translation. Here I describe in brief the cellular phenotypes caused by EXD2 deficiency and make the case that EXD2 inhibitors could be valuable agents for cancer therapy.

Commentary

In addition to generating the majority of cellular ATP, mitochondria are key sites for the generation of other critical metabolites and play important roles in signaling and programmed cell death. While the majority of mitochondrial proteins are encoded in the nucleus, mitochondria maintain their own genome that encodes rRNAs, tRNAs and 13 open reading frames encoding components of the oxidative phosphorylation (OXPHOS) complexes.¹ They therefore must coordinate the production and import of proteins with their own gene expression and translation in order to control the assembly of the OXPHOS complexes and meet the metabolic needs of the cell.

We recently reported the characterization of the Exonuclease 3′−5′ domain-containing protein 2 (EXD2).² EXD2 caught our attention due to its 3′−5′ exonuclease domain with homology to the Werner Syndrome ATP-dependent helicase (WRN). Mutations in WRN underlie Werner syndrome, a premature ageing and cancer predisposition syndrome characterized by DNA replication and repair defects. Several reports implicating EXD2 in DNA repair seemed to confirm our suspicions that EXD2 was a new replication or repair factor.^3,4 However, our results pointed to a predominantly mitochondrial role that appeared to be independent of mitochondrial DNA (mtDNA) replication. Cells with reduced EXD2 levels had increased reactive oxygen species (ROS), impaired glucose usage and reduced ATP levels. We identified the small subunit of the mitochondrial ribosome as a major interaction partner of EXD2 and mitochondrial translation was severely defective in EXD2 deficient cells. Biochemical studies indicated that EXD2's preferred substrate was single-stranded RNA, consistent with the fact that we observed aberrant mitochondrial ribosome fractions characterized by elevated mitochondrial mRNA levels. The reduction of EXD2 in Drosophila led to an increased lifespan, particularly in female flies where it correlated with the ROS-dependent premature loss of the germline stem cell population. Both the germline stem cell numbers and lifespan could be normalized if flies were fed an antioxidant diet.

While questions remain about the precise molecular function(s) of EXD2, we believe that the data in hand suggest that EXD2 is an interesting target to explore in cancer therapy, an idea that I will focus on for the remainder of this commentary.

Is EXD2 druggable? Human and insect EXD2 were identified as host factors for Dengue virus replication⁵ and a project to identify inhibitors of its exonuclease activity for antiviral therapy was funded by the NIH (grantome.com), although EXD2 inhibitors have not been reported. While not as common as kinase inhibitors, many nuclease inhibitors have been identified, particularly towards enzymes involved in DNA repair.⁶ The possibility that EXD2 could be targeted is therefore plausible, although homology amongst 3′−5′ exonuclease domain containing proteins could make the generation of specific inhibitors difficult.² As its molecular role becomes clearer, how EXD2 could be most effectively targeted will likely become more evident.

Why target EXD2 in cancer? For several decades, mitochondrial ATP generation through OXPHOS and the tricarboxylic acid (TCA) cycle was thought to be dysfunctional in cancer cells and metabolic interventions focused on targeting aerobic glycolysis. A rapidly expanding body of evidence has clarified that the TCA cycle is critical for the growth of many cancers where it accounts for a large proportion of the ATP generation. Metabolic reprogramming of cancer cells is now under intense study to identify potential vulnerabilities that can be used for therapy.⁷ While a complete review of this field is out of the scope of this commentary, I will briefly mention 3 areas of cancer metabolism research that support the potential of targeting EXD2 in cancer: inhibition of OXPHOS complexes, glutamine addiction and mitochondrial translation.

The retrospective observation that a large cohort of diabetes patients treated with Metformin, a drug that inhibits Complex I (NADH dehydrogenase) of the OXPHOS pathway, had reduced cancer incidence spurred new excitement, clinical trials and a search for additional compounds to target OXPHOS. As EXD2 deficient cells had reduced glucose flux into the TCA cycle and metabolomic alterations similar in many respects to those of Metformin treated cells, it stands to reason that EXD2 inhibitors could yield a similar effect as Metformin or related drugs, providing an additional tool for cancer therapy (Figure 1).^2,8 Notably, Metformin was recently shown to have an antiviral effect on Dengue,⁹ suggesting that promoting mitochondrial ATP generation or suppression of ROS may explain the requirements for EXD2 in Dengue infection.

Figure 1.

Predicted effects of targeting EXD2 in cancer. Similar to the OXPHOS inhibitors Metformin or Phenformin, Exonuclease 3′−5′ domain-containing protein 2 inhibitors (EXD2i) could reduce OXPHOS-mediated ATP production. Like antibiotics such as Tigecycline, EXD2i would be predicted to impair mitochondrial translation. Finally, EXD2i could render cells glutamine dependent, enhancing the efficacy of drugs such as C-968 and BPTES that target glutamine uptake.

The study of cancer metabolism has also revealed that many cancers are addicted to glutamine as a carbon source for the production of biomass. For example, metabolic reprogramming by the Myc oncogene results in increased glutamine uptake and catabolism. Therefore, inhibitors of glutamine uptake and catabolism are also under clinical investigation. We observed that EXD2 depletion conferred glutamine dependence for cell growth that was dependent on its exonuclease activity.² This again would suggest that EXD2 inhibitors could be useful for inducing glutamine addiction and enhancing the effect of drugs targeting glutamine usage (Figure 1).

Finally, mitochondrial translation, a field that has received decidedly less attention than its cytoplasmic parallel, has attracted new interest, as inhibitors of the mitochondrial ribosome, such as the antibiotic Tigecycline, have shown promise in numerous preclinical studies. Current evidence suggests that mitochondrial translation inhibitors may be effective anticancer agents, particularly affecting the cancer initiating cell population.¹⁰ Cells lacking EXD2 displayed a strong defect in mitochondrial translation that was dependent on its exonuclease and HNH domains.² This was comparable to control cells treated with antibiotics that inhibit mitochondrial translation, suggesting that EXD2 inhibition could mimic the tumor suppressive effects of these agents (Figure 1).

We have preliminary evidence using shRNA to target EXD2 in breast cancer xenografts that supports the potential of targeting EXD2 as an anti-tumoral strategy. Future work will be needed to further solidify these findings and determine if targeting EXD2 can impair OXPHOS, inhibit mitochondrial translation or enhance other existing therapeutics in vivo. If successful, these experiments may suggest that the identification of EXD2 inhibitors is worth pursuing as another tool to attack cancer metabolism.

Funding Statement

Ministerio de Economía y Competitividad.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Acknowledgments

T.H.S. is supported by the Spanish Ministry of Economy and Competitiveness (MINECO) [BFU2015-68354], Ayudas para incentivar la incorporación estable de doctores (IED) 2015 and institutional funding from MINECO through the Centres of Excellence Severo Ochoa award and from the CERCA Programme of the Catalan Government. We apologize to our many colleagues whose relevant work could not be referenced here due to space constraints.