Mutations in the
clk-1 gene of C. elegans affects lifespan and numerous physiological rates, including behavioral rates (defecation, pharyngeal pumping, and swimming), developmental rates (embryonic and post-embryonic), and reproduction.
clk-1 encodes a 187 amino acid protein localized in mitochondria. Interestingly,
clk-1 is implicated in the synthesis of ubiquinone (Q), an important electron transporter, especially in mitochondria. The involvement of CLK-1 in Q biosynthesis in C. elegans has been demonstrated by the finding that
clk-1 mutants do not synthesize Q (Miyadera et al., 2001; Jonassen et al., 2001). Rather, they accumulate demethoxyubiquinone (DMQ), a Q-synthesis intermediate that is able to sustain mitochondrial respiration in worms (Miyadera et al., 2001), as well as in mammals (Levavasseur et al., 2001). Remarquably, although only DMQ is present in all three
clk-1 alleles,
clk-1(
e2519), which is a point mutant, produces a much weaker phenotype. This suggests that the lack of Q cannot solely account for the Clk-1 phenotype. Recently, it has also been found that
clk-1 mutants are unable to grow on a Q-deficient bacterial strain (Jonassen et al., 2001). In order to understand the impact of Q metabolism on
clk-1 mutants, we systematically examined the phenotype of
clk-1 worms, when fed with bacteria deficient in each of the genes implicated in Q biosynthesis in E. coli (ubi genes). Our results confirm that exogenous Q is necessary for
clk-1 mutants fertility and development. This suggests that dietary Q from wild-type bacteria is capable of being used in a key cellular process that requires Q. To study this process further, we have generated a C. elegans knock-out (KO) in the
coq-3 gene, which encodes another enzyme that participates in ubiquinone biosynthesis. However, contrary to
clk-1 mutants,
coq-3 mutants do not appear to accumulate a Q-synthesis intermediate that is competent for respiration. In fact,
coq-3 KO mutants display a lethal phenotype, which suggests that exogenous Q cannot replace all of the Q functions. Taken together, our results suggest a model in which Q plays distinct roles at mitochondrial and non-mitochondrial sites (Hihi et al., 2002). In particular, Q is necessary at non-mitochondrial sites for development. Consistent with our model, we have found that
clk-1 is necessary for development in mammals, since a
clk-1 mouse KO is lethal early during development, even though respiration is almost normal (Levavasseur et al., 2001). This indicates that DMQ cannot functionally replace Q at some non-mitochondrial sites, and that these sites are probably conserved between species. Our model suggests how the Q biosynthesis defect relates to the phenotype of
clk-1 mutants. We are currently using genetic methods to determine the nature of the non-mitochondrial sites of Q action.