Complementation of Saccharomyces cerevisiae coq7 mutants by mitochondrial targeting of the Escherichia coli UbiF polypeptide: two functions of yeast Coq7 polypeptide in coenzyme Q biosynthesis

Abstract

Coenzyme Q (ubiquinone or Q) functions in the respiratory electron transport chain and serves as a lipophilic antioxidant. In the budding yeast Saccharomyces cerevisiae, Q biosynthesis requires nine Coq proteins (Coq1-Coq9). Previous work suggests both an enzymatic activity and a structural role for the yeast Coq7 protein. To define the functional roles of yeast Coq7p we test whether Escherichia coli ubiF can functionally substitute for yeast COQ7. The ubiF gene encodes a flavin-dependent monooxygenase that shares no homology to the Coq7 protein and is required for the final monooxygenase step of Q biosynthesis in E. coli. The ubiF gene expressed at low copy restores growth of a coq7 point mutant (E194K) on medium containing a non-fermentable carbon source, but fails to rescue a coq7 null mutant. However, expression of ubiF from a multicopy vector restores growth and Q synthesis for both mutants, although with a higher efficiency in the point mutant. We attribute the more efficient rescue of the coq7 point mutant to higher steady state levels of the Coq3, Coq4, and Coq6 proteins and to the presence of demethoxyubiquinone, the substrate of UbiF. Coq7p co-migrates with the Coq3 and Coq4 polypeptides as a high molecular mass complex. Here we show that addition of Q to the growth media also stabilizes the Coq3 and Coq4 polypeptides in the coq7 null mutant. The data suggest that Coq7p, and the lipid quinones (demethoxyubiquinone and Q) function to stabilize other Coq polypeptides.

FIGURE 1. Proposed eukaryotic Q biosynthetic pathway

There are nine identified Coq proteins in S. cerevisiae ascribed the following functions: Coq1, hexaprenyl-diphosphate synthase; Coq2, 4-hydroxybenzoate:hexaprenyl-diphosphate transferase; Coq3, O-methyltransferase; Coq5, C-methyltransferase; Coq6, potential flavin-dependent monooxygenase; and Coq7/CLK-1, putative di-iron hydroxylase catalyzing the last hydroxylase step. The function of Coq4, Coq8, and Coq9 are unknown. The number of isoprene units in the tail of Q is designated by n; n = 6 in S. cerevisiae, 9 in C. elegans, and 10 in Homo sapiens.

FIGURE 2. Identification of mutations in yeast COQ7

Amino acid sequence of Coq7 from S. cerevisiae (NCBI accession number NP_014768), C. elegans (AAG00035), H. sapiens (H. sap; NP_057222), Mus musculus (M. mus; AAH38681), Drosophila melanogaster (D. mel; NP_651967), Pseudomonas aeruginosa (P. aer; AAG04044), Rickettsia prowazekii (R. pro; CAA14656) were aligned with the Clustal method in the Megalign program of DNASTAR. Identical residues are shaded black. The previously proposed Coq7 amino acid sequence of S. cerevisiae (X82930) was modified by the Saccharomyces Genome Data base, which predicts that translation initiated at the second in-frame ATG codon of the original reported ORF, leading to exclusion of the first 39 amino acids. Consequently, the G104D and E233K mutations noted previously (15, 16) are now designated as G65D and E194K, respectively. Seven distinct mutations in the COQ7 ORF were identified via sequencing of amplified genomic DNA segments of nine mutant strains from the G64 (coq7) complementation group (4) and amino acid substitution mutations are designated by arrows: G185A (G62D) (coq7-2, N183 and E2–196), G194A (G65D) (coq7-1, NM101) (15), G215A (G72D) (coq7-3, N558), G580A (E194K) (coq7-4, C104), and C589T (H197K) (coq7-5, W202). Asterisks designate nonsense mutations: G276A (W92STOP) (coq7-6, C291 and E4-140) and G360A (W120STOP) (coq7-7, N49). The region predicted to be adjacent to the DMQ binding pocket is designated by a bracket; predicted ligands to the di-iron center (EXXH) are in gray.

FIGURE 3. Expression of E. coli UbiF rescues the E194KCoq7 point mutant more effectively than the coq7 null mutant

Each strain was grown overnight in selective media and serial dilutions were spotted on SDC, YPG, YPE, and YPEG plate media, at a final optical density 600 nm (A₆₀₀) of 0.2, 0.02, 0.002, or 0.0002. Growth is depicted after 6 days at 30 °C. E194K, Δ, and WtCoq7, represent E194KCoq7, the coq7 null mutant (JM43Δcoq7-1) or the wild-type integrant, respectively. JM43 designates the parental wild-type strain. Yeast strains were transformed with pQMF, a low copy plasmid expressing E. coli UbiF; pCHF, a multicopy plasmid expressing E. coli UbiF; or pNMQ71, a low copy plasmid containing the yeast COQ7 gene.

FIGURE 4. Assays of Q₆ and DMQ₆ in coq7 null and E194KCoq7 mutants harboring low copy and multicopy plasmids expressing E. coli ubiF

All strains were cultured in YPGal media. Lipid extracts of cell pellets were prepared and analyzed for quinone content by HPLC/ECD as described under “Experimental Procedures.” The elution positions of Q₄, Q₆, and DMQ₆ are indicated: WtCoq7 (A), JM43Δcoq7 (B), E194KCoq7 (C), E194KCoq7:pQMF (D), E194KCoq7:pCHF (E), and E194KCoq7:pNMQ71 (F). The plasmid descriptions are as follows: pQMF, a low copy plasmid expressing E. coli UbiF; pCHF, a multicopy plasmid expressing E. coli UbiF; and pNMQ71, a low copy plasmid containing the yeast COQ7 gene.

FIGURE 5. Steady state levels of Coq1, Coq3, Coq4, Coq5, Coq6, and Coq7 polypeptides in wild-type and coq7 mutants

Purified mitochondria were isolated from yeast strains grown in YPGal media; parental wild type (JM43), JM43Δcoq7 (Δ), E194KCoq7 (E194K), and wild-type Coq7 integrant control (WtCoq7). Samples of mitochondria protein 20 (2×) or 10 µg (1×) were separated by SDS-PAGE and analyzed by Western blotting with antibodies against the Coq polypeptides (Coq1, Coq3, Coq4, Coq5, Coq6, and Coq7), and for other polypeptides of the mitochondrial respiratory chain including the cytochromes b₂, c, and c₁, and F₁β (the F₁-ATP synthase β subunit).

FIGURE 6. Steady state levels of Coq1, Coq3, Coq4, and Coq7 polypeptides in the presence and absence of UbiF or exogenous Q₆

Purified mitochondria were prepared from the following yeast strains grown in YPGal media: JM43Δcoq7:pCHF (Δ:pCHF), JM43Δcoq7:pQMF (Δ:pQMF), E194KCoq7:pQMF (E194K:pQMF), E194KCoq7:pCHF (E194:pCHF), E194KCoq7 with Q₆ supplement (E194K + Q₆), E194KCoq7 without Q₆ supplement (E194K), JM43Δcoq7 (Δ), and JM43Δcoq7 with Q₆ supplement (Δ + Q₆). Samples (20 µg of protein) were separated by SDS-PAGE and analyzed by Western blotting with antibodies against Coq1, Coq3, Coq4, and Coq7 proteins. The plasmid descriptions are as follows: pQMF, a low copy plasmid expressing E. coli UbiF; and pCHF, a multicopy plasmid expressing E. coli UbiF.

FIGURE 7. Coq7 polypeptide co-elutes with Coq3 and Coq4 as a high molecular mass complex

A, gel filtration analysis of Coq3 O-methyltransferase activity. Wild-type mitochondria (2 mg) were solubilized with digitonin (2:1, w/w, detergent/protein) and the 100,000 × g supernatant was subjected to gel filtration chromatography with calibration standards (denoted by arrows) as described under “Experimental Procedures.” Coq3 O-methyltransferase (pmol of methyl groups/fraction/h) is depicted. B, gel filtration analysis of Coq polypeptides. The Coq3, Coq4, Coq7, and isocitrate dehydrogenase (IDH) polypeptides were detected in eluate fractions by immunoblot analysis.

FIGURE 8. Coq7 polypeptide co-migrates with Coq3 and Coq4 as a high molecular mass complex

Mitochondria were isolated from the following yeast strains: JM43 (A), E194KCoq7 (B), JM43Δcoq7 (C), and JM43Δcoq7 with Q₆ supplement (D). Mitochondria (300 µg) were solubilized with digitonin and subjected to BN-PAGE (5–13.5%) in the first dimension and SDS-PAGE (10–13%) in the second dimension as described. The point of origin and size separation for the first BN dimension is designated by an arrow with the position of the molecular mass standards (in kDa) indicated. Aliquots of control mitochondria from JM43Δcoq7 (Δ) and wild type (JM43) were subjected only to the second dimension (SDS-PAGE) separation. The indicated that Coq polypeptides were detected via immunoblotting with specific antisera. A representative blot from each panel was stripped and re-probed with anti-Rip1.