Drosophila orthologue of WWOX, the chromosomal fragile site FRA16D tumour suppressor gene, functions in aerobic metabolism and regulates reactive oxygen species

Abstract

Common chromosomal fragile sites FRA3B and FRA16D are frequent sites of DNA instability in cancer, but their contribution to cancer cell biology is not yet understood. Genes that span these sites (FHIT and WWOX, respectively) are often perturbed (either increased or decreased) in cancer cells and both are able to suppress tumour growth. While WWOX has some tumour suppressor characteristics, its normal role and functional contribution to cancer has not been fully determined. We find that a significant proportion of Drosophila Wwox interactors identified by proteomics and microarray analyses have roles in aerobic metabolism. Functional relationships between Wwox and either CG6439/isocitrate dehydrogenase (Idh) or Cu-Zn superoxide dismutase (Sod) were confirmed by genetic interactions. In addition, altered levels of Wwox resulted in altered levels of endogenous reactive oxygen species. Wwox (like FHIT) contributes to pathways involving aerobic metabolism and oxidative stress, providing an explanation for the 'non-classical tumour suppressor' behaviour of WWOX. Fragile sites, and the genes that span them, are therefore part of a protective response mechanism to oxidative stress and likely contributors to the differences seen in aerobic glycolysis (Warburg effect) in cancer cells.

Figure 1.

Proteomic analysis of altered Wwox expression. 2D-DIGE protein spots that exhibited significant changes in abundance either in both of the independent Wwox mutants (Wwox¹ and Wwox^f04545) or with ubiquitous ectopic Wwox expression in 0- to 1-day adult Drosophila (see also Supplementary Material, Table S1). Spots 19 and 20 are superoxide dismutase (Sod) isoforms.

Figure 2.

Metabolic pathways impacted by altered Wwox levels. Alteration of Wwox levels resulted in changes to many enzymes with roles in the TCA cycle, glucose metabolism, ethanol metabolism, lipid metabolism and oxidation/reduction supportive of a contributing role for Wwox in the maintenance of aerobic metabolism. Arrows indicate up- or down-regulated Wwox interacting candidates: solid black arrows for those altered following ectopic over-expression of Wwox and open arrows for those altered in Wwox loss of function mutants. Side arrows indicate that changes were detected in different isoforms of the protein. Gene abbreviations are listed in Table 1.

Figure 3.

In vivo functional interactions of Wwox with CG6439/Idh. (A) Rationale of the in vivo genetic screens for Wwox functional interactors in Drosophila. Flies with ubiquitous knockdown of the candidate interactors alone were assayed for any resultant phenotypes that could be modified by either decreased or increased levels of Wwox. (B) Wwox interacts genetically with CG6439/Idh in viability assays. A deviation from the expected proportion (50%, indicated by the bold line) of non-TM6B progeny (see Supplementary Material, Fig. S1A) was observed when CG6439/Idh was knocked down ubiquitously (blue), indicative of a decrease in viability. This decrease in viability (blue) was enhanced when Wwox levels were decreased (red) and suppressed when Wwox levels were increased (green). Chi-square test was performed on each of the five separate experimental replicates and ‘*’ denotes statistical significance with P < 0.05. (C) Correlation of WWOX and IDH1 transcript levels in human cancer cell lines. qPCR analyses of WWOX and IDH1 transcript levels were determined for four different time points/confluencies in each of 15 exponentially growing human cancer cell lines. Regression analysis of WWOX and IDH1 mRNA levels revealed a positive correlation with a P-value of 2.5E−12 (see also Supplementary Material, Figs S2, S3 and Table S2).

Figure 4.

Wwox and Sod interact genetically. (A) A decrease in viability was observed when homozygous Wwox¹ or trans-heterozygous Wwox¹/Wwox^f04545 mutations were introduced into the Sodⁿ¹/Sodⁿ⁶⁴ mutant background (as indicated by the negative deviation from the expected 33.3% proportion of non-TM6B progeny, see Supplementary Material, Fig. S2). (B) A decreased lifespan was observed for the trans-heterozygous Sod mutations compared with wild-type and Wwox mutants. Each of the Wwox;Sod double mutants showed a further decrease in viability. Survival curves for the two Wwox;Sod double mutants overlap on the graph with all flies dead after 24 h. (C) qPCR showed decreased levels of Wwox transcript in Sod^n1/n64 compared with w¹¹¹⁸ larvae (P = 0.003) and increased levels of Wwox transcript in SOD1 overexpressing compared with control GFP overexpressing larvae (P = 0.043). (D) qPCR of WWOX levels in a HEK293 cells overexpressing SOD1 showed increased levels of endogenous WWOX transcript compared with the control line overexpressing GFP (P = 0.008, see also Supplementary Material, Fig. S4). No such increase in WWOX transcript was observed when cells were overexpressing a G37R mutant form of SOD1.

Figure 5.

Altered Wwox expression affects the levels of ROS in Drosophila larvae. (A) Fluorescence-activated cell sorter (FACS) analysis of ROS levels in cells from two independent lines over-expressing Wwox (da > Wwox#1 and da > Wwox#2) compared with control (da>+) and two Wwox mutant lines (Wwox¹ and Wwox^f04545) compared with control (w¹¹¹⁸) after 1 h incubation with 10 µm CM-H2DCFDA. (B) An arbitrary threshold of fluorescence intensity was set and the percentage of cells from the total population exhibiting fluorescence above that threshold was measured every 30 min for 3 h.