CG7630 is the Drosophila melanogaster homolog of the cytochrome c oxidase subunit COX7B

Abstract

The mitochondrial respiratory chain (MRC) is composed of four multiheteromeric enzyme complexes. According to the endosymbiotic origin of mitochondria, eukaryotic MRC derives from ancestral proteobacterial respiratory structures consisting of a minimal set of complexes formed by a few subunits associated with redox prosthetic groups. These enzymes, which are the "core" redox centers of respiration, acquired additional subunits, and increased their complexity throughout evolution. Cytochrome c oxidase (COX), the terminal component of MRC, has a highly interspecific heterogeneous composition. Mammalian COX consists of 14 different polypeptides, of which COX7B is considered the evolutionarily youngest subunit. We applied proteomic, biochemical, and genetic approaches to investigate the COX composition in the invertebrate model Drosophila melanogaster. We identified and characterized a novel subunit which is widely different in amino acid sequence, but similar in secondary and tertiary structures to COX7B, and provided evidence that this object is in fact replacing the latter subunit in virtually all protostome invertebrates. These results demonstrate that although individual structures may differ the composition of COX is functionally conserved between vertebrate and invertebrate species.

Keywords: D. melanogaster; COX7B; cytochrome c oxidase; mitochondria; respiratory chain.

Figure 1. Identification of CG7630 as a novel subunit of cytochrome c oxidase (COX)

1. Scheme of a complexome profiling experiment. Mitochondrial from adult D. melanogaster individuals were isolated and subjected to native electrophoresis after solubilization with a mild detergent (digitonin). Natively separated complexes were excised from the gel in 60 slices, each subjected to MS analysis. Peptide mass fingerprinting profile distribution throughout the gel lane was analyzed and shown as complexome profiles.

2. MS profiles depicted as heatmaps and relative abundance of MRC enzymes in natively separated complexes from wild‐type fly mitochondria. Profiles of complexes I, III₂ and V (CI, CIII and CV) are reported as average migration profiles of their specific subunits identified by MS. Canonical COX subunits migrate as enzyme monomer (IV₁), dimer (IV₂) and together with dimeric complex III as supercomplex (IV₁ + III₂). A novel protein product of a gene with unknown function (CG7630) co‐migrates with canonical COX subunits.

Figure 2. CG7630 is a mitochondrial protein imported via an N‐terminal MTS

1. Scheme of synthetic constructs used to determine the subcellular localization of CG7630.

2. Confocal micrographs show the MTS‐dependent mitochondrial localization of CG7630. Embryonic S2R+ cells were transfected with GFP‐carrying the empty vector (EV:GFP), and vectors expressing HA‐tagged CG7630 protein (CG7630‐HA), a mutant CG7630‐HA carrying a deletion in the putative MTS (CG7630^Δ2–23‐HA) and a mutant consisting of CG7630 putative MTS fused with the GFP reporter (CG7630^1–23‐GFP). Scale bar: 5 μm.

Figure 3. CG7630 is a bona fide COX subunit

1. Relative quantification (RQ) of CG7630 transcript after RNAi with a 120‐bp inverted‐repeat sequence (act5c‐gal4>UAS‐CG7630^RNAi ) compared to control (act5c‐gal4>+). Data are plotted as mean ± SD (n = 3 biological replicates, Student’s t test ***P = 0.0005).

2. Relative percentage of egg to adult viability of CG7630 RNAi cross (act5c‐gal4/CyO.GFP x UAS‐CG7630^RNAi ) and control cross (act5c‐gal4/CyO.GFP x w¹¹¹⁸ ), calculated at three developmental stages (eggs, pupae, and adults), (n > 250, Chi‐square test 9.920 df(2), **P = 0.007).

3. Mendelian frequencies of adults obtained by crossing heterozygous act5c‐gal4>CyO.GFP flies with w¹¹¹⁸ genetic background flies (+/+) (n = 388 individuals, Chi‐square test 0.2527, df(1), P = 0.6152), and by crossing heterozygous act5c‐gal4>CyO.GFP flies with homozygous UAS‐CG7630^RNAi flies (n = 340 individuals, Chi‐square test 226.7, df(1), ****P < 0.0001).

4. Kinetic enzyme activity of individual MRC complexes in CG7630 RNAi and control individuals normalized by citrate synthase (CS) activity. Data are plotted as mean ± SD (n = 3 biological replicates, two‐way ANOVA with Sidak’s multiple comparisons, *P ≤ 0.05, ****P ≤ 0.0001).

5. In gel‐activity assays for MRC complexes II and IV in DDM‐solubilized mitochondria from RNAi (act5c‐gal4>UAS‐CG7630^RNAi ) and control larvae (act5c‐gal4>+).

6. BN‐PAGE, Western blot immunodetection of MRC complexes from CG7630 RNAi (act5c‐gal4>UAS‐CG7630^RNAi ) and control (act5c‐gal4>+) larvae using antibodies against specific subunits: anti‐UQCRC2 (complex III), anti‐NDUFS3 (complex I), anti‐COX4 (complex IV) and anti‐SDHA (complex II).

7. Western blot analysis of steady‐state levels (SDS‐PAGE) of MRC subunits COX4 (complex IV), SDHA (complex II) and UQCR‐C2 (complex III) in whole lysates from CG7630 RNAi (act5c‐gal4>UAS‐CG7630^RNAi ) and control (act5c‐gal4>+) larvae.

8. Western blot analysis of anti‐HA affinity purified samples from S2R+ D. melanogaster cells. EV = samples from cells carrying the empty expression vector. CG7630‐HA = samples from cells expressing HA‐tagged CG7630. Tot = total mitochondrial‐enriched fractions; unb = unbound material; IP:HA = immunoprecipitated material. Samples were probed with antibodies against MRC subunits of complexes I (NDUFS3), II (SDHA), III (UQCR‐C2), IV (COX4), and V (ATP5A).

Source data are available online for this figure.

Figure EV1. Alignment of COX7B with D. melanogaster proteins gomdanji and CG7630

HHpred was run with the mature COX7B protein against the D. melanogaster proteome using default settings.

Figure 4. Phylogenetic distribution of CG7630/COX7B suggests it emerged at the root of bilateria

1. Scheme of the evolution of subunits part of human cytochrome c oxidase, sorted by their evolutionary origin, with the oldest subunits on the left. Subunits colored in red date back to bacteria. Orange indicates subunits with a common ancestor at the root of eukaryotes. Subunits shared between animals and fungi (opisthokonts) are in yellow while subunits dating at the origin of bilateria are in green. Blue is for the subunit that can be traced back to the deuterostomia and purple is for the subunits originated after the origin of deuterostomia. Note that the most recent subunits originate from gene duplications.

2. Alignment of the sequence logos of COX7B and CG7630. The sequence logos were created with weblogo (https://weblogo.berkeley.edu/) using alignments obtained from Jackhmmer. The positions that are conserved between the profiles (Fig EV1) are indicated with vertical lines, and the (predicted) TM helix is indicated with a horizontal line. The amino acid positions are based on the mature protein. Relative to COX7B, CG7630 has a 20 amino acid insertion close to the N terminus of the mature protein.

Figure EV2. COX (5Z62) with subunits colored by their evolutionary origin

The color code is the same that was used in Fig 5. Note that youngest subunits COX7B and COX8A are close together in the structure, suggesting a common reason for their origin.

Figure 5. The 14‐subunit model of Drosophila melanogaster COX enzyme

1. A, B

   The front and side view of CG7630, panel (A and B), respectively, is reported using the red ribbon and its molecular surface is in transparent red against the other COX subunits in gray‐white. The inner mitochondrial membrane is reported in dark gray.

2. C

   In panel C, the ribbon representation is extended to the whole enzyme.

3. D

   In panel D, each subunit is shown as a ribbon in different colors.

4. E

   Panel E shows Drosophila melanogaster atomic model of CG7630 and COX4 helices (red) interactions compared to the human COX7B‐COX4 interactions (yellow). The conserved residues, located in the proximity of the intermembrane space and involved the helix‐helix interactions, are highlighted using stick representation and the distance of the observed H‐bonds are displayed.

Figure 6. Human COX7B can rescue COX deficiency caused by loss of CG7630

1. Relative quantification (RQ) of CG7630 transcript in wild‐type D. melanogaster S2R+ cells treated with a dsRNA targeting CG7630 (CG7630 KD) and a mock control dsRNA (mock). Data are plotted as mean ± SD (n = 3 biological replicates, Student’s t test ***P = 0.0005).

2. Relative quantification (RQ) of CG7630 and COX7B transcripts in stable S2R+ cell lines carrying the empty expression vector (EV:GFP;neo), and expression vectors carrying HA‐tagged forms of human COX7B (hCOX7B‐HA;neo) and CG7630 (CG7630‐HA;neo). Cells were treated with a dsRNA targeting CG7630 (striped bars, CG7630 KD) and a mock control dsRNA (solid bars, mock). N/D = not detected. Data are plotted as mean ± SD (n = 3 biological replicates, one‐way ANOVA with Dunnet’s multiple comparisons, *P = 0.0370, **P = 0.0074, ***P = 0.0003).

3. Immunoblot analysis of stable S2R+ cell lines carrying the empty expression vector (EV:GFP;neo), and expression vectors carrying HA‐tagged forms of human COX7B (hCOX7B‐HA;neo) and CG7630 (CG7630‐HA;neo). The CG7630 transcript was knocked‐down (KD) in cells by a dsRNA targeting CG7630 and compared with cells treated with a mock control dsRNA (mock). Samples were probed with an antibody anti‐HA tag and with an antibody against Hsp70 as an endogenous control.

4. Kinetic enzyme activity of COX normalized by citrate synthase activity (CS) in stable S2R+ cell lines carrying the empty expression vector (EV:GFP;neo), and expression vectors carrying HA‐tagged forms of human COX7B (hCOX7B‐HA;neo) and CG7630 (CG7630‐HA;neo). Cells were treated with a dsRNA targeting CG7630 (striped bars, CG7630 KD) and a mock control dsRNA (solid bars, mock). Data are plotted as mean ± SD (n = 3 biological replicates, two‐way ANOVA with Sidak’s multiple comparisons, *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001).

5. Steady‐state levels of MRC subunits in samples from rescue experiment performed as in (D).

Source data are available online for this figure.