Assembly of the peripheral stalk of ATP synthase in human mitochondria

Significance The production of ATP in mitochondria requires the oxidation of energy rich compounds to generate a proton motive force (pmf), a chemical potential difference for protons across the inner membrane. This pmf powers the ATP synthase, a molecular machine with a rotary action, to synthesize ATP. The assembly of human ATP synthase from 27 nuclear encoded proteins and two mitochondrially encoded subunits in the inner organellar membrane involves the formation of intermediate modules representing the F1-catalytic domain, the peripheral stalk, associated membrane subunits, and the c8 ring in the membrane part of the rotor. Here, we describe how components of the peripheral stalk and three associated membrane subunits are assembled and introduced into the enzyme complex.

SILAC DMEM media was also supplemented with 20 mM HEPES (pH 7.4) to aid buffering of the increased acidification of the media by these cell clones. Reciprocally labelled HEK293T-Δδ-bT (tagged subunit-b) and HEK293T-Δδ-jT (tagged j subunit) were mixed 1:1 (protein w/w). Mitoplast material was prepared by digitonin treatment (11,12) and then solubilized with digitonin (12 g/g protein) for affinity purification of tagged subunits and associated proteins. A sample was loaded at 4 o C onto a Pierce spin column (0.9 ml; Thermo Fisher Scientific) containing Strep-Tactin-Sepharose (IBA Lifesciences) and then washed with 5 column volumes of buffer [20 mM HEPES, pH 7.6, 150 mM NaCl, 2 mM dithiothreitol, 1x cOmplete EDTA-free protease inhibitor (Roche), and 0.05% (w/v) digitonin]. Bound protein was eluted with 6 portions of 0.5 column volumes of buffer containing 10 mM desthiobiotin. Eluates were analyzed by SDS-PAGE and quantitative mass spectrometry. A distinction was made between the endogenous subunits and the tagged form by the different migration positions on the SDS-PAGE gel. From the MaxQuant evidence file for the appropriate gel sections, the protein ratios of the tagged and endogenous subunits were determined manually (SI appendix, Datasets S10 and S11).
General Methods. Cell protein concentrations were determined by either the bicinchoninic acid assay (Thermo Fisher Scientific) or the detergent compatible protein assay (BioRad). Mitoplasts were prepared from cells with digitonin, as described before (11,12). Extracts of mitoplasts made with dodecylmaltoside (DDM; 1%, w/v) were fractionated by SDS-PAGE, and subunits of ATP synthase and citrate synthase were detected by Western blotting. The oligomeric states of ATP synthase and vestigial complexes in digitonin extracts of mitoplasts were examined by BN-PAGE or CN-PAGE (13,14), and Western blotting. Samples of mitoplasts were re-suspended to ca. 5 mg/ml in NativePAGE sample buffer (Thermo Fisher Scientific) containing digitonin (6-12 g/g protein), kept at 4°C for 15 min, and then centrifuged (10,500 x g, 20 min, 4°C). The supernatants were treated with benzonase (Merck Millipore) at room temperature, centrifuged again, and soluble complexes fractionated at 4°C in 3-12% acrylamide gradient Bis-Tris gels (Thermo Fisher Scientific) by CN-PAGE, or BN-PAGE according to the manufacturer's instructions for Western blotting. For CN-PAGE the cathode running buffer contained 0.05% (w/v) sodium deoxycholate plus 0.005% (w/v) DDM. The gel resolved complexes were transferred to polyvinylidene fluoride membranes, and the membranes were probed with subunit specific antibodies. The origins of the antibodies either have been described before (5,6), or they are listed in SI Appendix Table S3. ATP synthase was purified from digitonin solubilized mitoplasts with an immuno-capture resin (Abcam) as described before (7). Proteins in SILAC labelled mitoplast samples for quantitative mass spectrometric analysis were reduced and alkylated in gel sample buffer, fractionated by SDS-PAGE and stained with Coomassie blue R250 dye (15). Stained gel sections were excised and proteins digested in-gel with trypsin (16). SILAC labelled and affinity purified samples of ATP synthase were ethanol precipitated at -20ºC for 18 h with 20 vol. cold ethanol, centrifuged, and the pellet was digested in 50 mM ammonium bicarbonate for 18 h, with either trypsin at 37ºC or chymotrypsin at 30ºC. Protein Quantitation. Relative quantitation of proteins was derived from mass spectrometric data of SILAC samples (9). Peptide mixtures were analyzed by LC-MS-MS on a Proxeon EASY-nLC1000 system coupled directly to a Q-Exactive+ Orbitrap mass spectrometer (Thermo Fisher Scientific). Heavy and light peptide mass data were analyzed with MaxQuant version 1.6.5.0, and the integrated Andromeda search engine (17,18) employing a Swiss-Prot human protein database (March 2019) modified to include mature forms of ATP5IF1, denoted as IF1-M1, -M2 and -M3, with N-terminal residues Phe-25, Gly-26 and Ser-27, respectively (19,20), and a mature ATP synthase c-subunit. The ATP synthase c-subunit is encoded by three genes (21,22), with different mitochondrial targeting pre-sequences, but identical mature protein sequences. To aid identification of this subunit, using peptides derived from the mature protein N-terminal sequence, a representative sequence (lacking the mitochondrial N-terminal import sequence) was added to the human protein database employed in these analyses, with the identifier P48201-M. In addition, lysine trimethylation was included when interrogating chymotrypsin digest data, to aid identification of a characteristic methylated subunit-c peptide (23). Search parameters for in-gel trypsin digest samples were: MS tolerance 4.5 p.p.m.; MS/MS tolerance 20 p.p.m.; Trypsin/P with two missed cleavages; Fixed modification -Cys-carbamidomethyl; variable modifications -oxidation (Met) and acetyl (protein N-terminus); Arg-10 and Lys-8. In solution digest sample parameters were: MS tolerance 4.5 p.p.m.; MS/MS tolerance 20 p.p.m.; trypsin/P with two missed cleavages; chymotrypsin with four missed cleavages; variable modifications -oxidation (Met), acetyl (protein N-terminus) and trimethyl (lysine); Arg-10 and Lys-8. The MaxQuant output was managed further with Perseus (24). Protein ratios for ATP5IF1 were calculated manually with data for the unique Nterminal peptides of the various mature forms of ATP5IF1, located in the MaxQuant evidence file, and represent the median of the assigned specific peptide ratios, where MaxQuant ISO-MSMS peptide values were used only if fewer than three MULTI-MSMS peptide ratios were obtained. The basis of quantitative experiments using SILAC has been described previously (25).  (26). A, the membrane domain of subunit b, viewed orthogonal to the plane of the IMM, consisting of amphipathic α-helix bH1 (residues 19-29) in the head-group region on the matrix side of the IMM, transmembrane α-helix bH2 (residues 33-47) and the membrane region (residues 55-73) of bH3. In B, according to the assembly pathways in Fig. 7, subunits e and g associate with b to form the b-e-g intermediate. In C, subunit f is incorporated subsequently to complete the protein components of the wedge; D-H, possible points of incorporation of lipids. In E-G, CDL3, LGH4 and LGH5 could be incorporated before or after the addition of e and g to b to form b-e-g, but before incorporation of f. In H, CD1 is incorporated after the formation of the b-e-g intermediate to complete the b-e-g-f complex.         In the upper sections of (A) and (B), the disrupted DNA sequences are aligned with the wild-type (WT) sequence. The carets above the DNA sequences indicate the PAM (protospacer adjacent motif) sequence for the guide RNA, and solid lines the guide RNA target sequence. In the lower sections, the impact of the deletions on the protein sequence of the δ-subunit are shown. Asterisks indicate changes in the protein sequence arising from the deletions, and the dashes indicate that the protein sequence has been terminated by the introduction of a stop codon. In (A) and (B), respectively, truncated versions of the δ-subunit were produced consisting of residues 1-88 followed by 50 and 13 unmatched residues. By internal initiation of the translation from four cryptic AUG codons, it is possible that the insertion might lead also to the production of residues 124-168 of the δ-subunit and three unrelated polypeptides (90, 62 and 117 amino acids long, not shown).

Dataset legends
Dataset S1 (separate file). Proteins identified in SILAC experiments comparing immunopurified ATP synthase from wild-type and HAP1-ΔF6 cells. This information is the output from Perseus after processing of MaxQuant SILAC peptide pair data. ATP synthase was immunopurified from digitonin solubilized mitoplast material prepared from a 1:1 mixture of wildtype and HAP1-ΔF6 cells, and digested in-solution with trypsin or chymotrypsin. Experiment 1 refers to heavy isotope labelled HAP1-ΔF6 cells mixed with light labelled wild-type cells, and experiment 2 vice versa. Perseus processing removed proteins identified in both a decoy database (created in MaxQuant by reversing protein entries) and a contaminant database, experiment 2 ratios were inverted, and the protein ratios rendered base two logarithmic. Only protein groups (listed under 'Protein names') with ratios determined in both experiments are included. ATP5IF1 ratios were manually calculated using data obtained for unique peptides from the N-termini of two mature forms (IF1-M1 and -M3) of the protein (see Dataset S2).

Dataset S2 (separate file). Peptide data for the ATPase inhibitor protein obtained in SILAC experiments comparing ATP synthase from wild-type and HAP1-ΔF6 cells.
This information is obtained from the MaxQuant evidence file after the processing of SILAC peptide pair data. ATP synthase was immunopurified from digitonin solubilized mitoplast material prepared from a 1:1 mixture of HAP1 wild-type and HAP1-ΔF6 cells. Experiment 1 refers to heavy isotope labelled HAP1-ΔF6 cells mixed with light labelled wild-type cells, and experiment 2 vice versa. Only peptide data that provided ratio information is included. When fewer than three MULTI-MSMS ratios were obtained ISO-MSMS ratios were also included to determine the protein ratio, represented by the median peptide ratio.

Dataset S3 (separate file). Proteins identified in SILAC experiments comparing immunopurified ATP synthase from wild-type and HAP1-Δd cells.
This information is the output from Perseus after processing of MaxQuant SILAC peptide pair data. ATP synthase was immunopurified from digitonin solubilized mitoplast material prepared from a 1:1 mixture of wildtype and HAP1-Δd cells, and digested in-solution with trypsin or chymotrypsin. Experiment 1 refers to heavy isotope labelled HAP1-Δd cells mixed with light labelled wild-type cells, and experiment 2 vice versa. Perseus processing removed proteins identified in both a decoy database (created in MaxQuant by reversing protein entries) and a contaminant database, experiment 2 ratios were inverted, and the protein ratios rendered base two logarithmic. Only protein groups (listed under 'Protein names') with ratios determined in both experiments are included. The protein ratio is derived from a minimum of two peptide ratios from each experiment, except for the MT-ATP6 protein ratio for experiment 2 which is from a single peptide value. ATP5IF1 ratios were manually calculated using data obtained for unique peptides from the N-termini of two mature forms (IF1-M1 and -M3) of the protein (see Dataset S4).

Dataset S4 (separate file). Peptide data for the ATPase inhibitor protein obtained in SILAC experiments comparing ATP synthase from wild-type and HAP1-Δd cells.
This information is obtained from the MaxQuant evidence file after the processing of SILAC peptide pair data. ATP synthase was immunopurified from digitonin solubilized mitoplast material prepared from a 1:1 mixture of HAP1 wild-type and HAP1-Δd cells. Experiment 1 refers to heavy isotope labelled HAP1-Δd cells mixed with light labelled wild-type cells, and experiment 2 vice versa. Only peptide data that provided ratio information is included. When fewer than three MULTI-MSMS ratios were obtained ISO-MSMS ratios were also included to determine the protein ratio, represented by the median peptide ratio.

Dataset S5 (separate file). Proteins identified in SILAC experiments comparing mitoplasts from wild-type and HAP1-Δd cells.
This information is the output from Perseus after processing of MaxQuant SILAC peptide pair data. Digitonin solubilized mitoplast samples were prepared from a 1:1 mixture of wild-type and HAP1-Δd cells, analyzed by SDS-PAGE and gel sections digested with trypsin. Experiment 1 refers to heavy isotope labelled HAP1-Δd cells mixed with light labelled wild-type cells, and experiment 2 vice versa. Perseus processing removed proteins identified in both a decoy database (created in MaxQuant by reversing protein entries) and a contaminant database, experiment 2 ratios were inverted, and the protein ratios rendered base two logarithmic. Only protein groups (listed under 'Protein names') with ratios determined in both experiments are included. The protein ratio is derived from a minimum of two peptide ratios from each experiment, except for ATP5MF isoform 2 protein ratio which is from a single peptide value for both experiments, and MT-ATP6 protein ratio for experiment 2 which is from a single peptide value. ATP5IF1 ratios were manually calculated using data obtained for unique peptides from the N-termini of one mature form (IF1-M1) of the protein (see Dataset S6). The IF1-M1 protein ratio for experiment 1 is from a single peptide value.
Dataset S6 (separate file). Peptide data for the ATPase inhibitor protein obtained in SILAC experiments comparing mitoplasts from wild-type and HAP1-Δd cells. This information is obtained from the MaxQuant evidence file after the processing of SILAC peptide pair data. The peptide data is from digitonin protein extracts of mitoplast samples prepared with a 1:1 mixture of HAP1wild-type and HAP1-Δd cells. Experiment 1 refers to heavy isotope labelled HAP1-Δd cells mixed with light labelled wild-type cells, and experiment 2 vice versa. Only peptide data that provided ratio information is included.

Dataset S7 (separate file). Proteins identified in SILAC experiments comparing mitoplasts from wild-type and HAP1-ΔF6 cells.
This information is the output from Perseus after processing of MaxQuant SILAC peptide pair data. Digitonin solubilized mitoplast samples were prepared from a 1:1 mixture of wild-type and HAP1-ΔF6 cells, analyzed by SDS-PAGE and gel sections digested with trypsin. Experiment 1 refers to heavy isotope labelled HAP1-ΔF6 cells mixed with light labelled wild-type cells, and experiment 2 vice versa. Perseus processing removed proteins identified in both a decoy database (created in MaxQuant by reversing protein entries) and a contaminant database, experiment 2 ratios were inverted, and the protein ratios rendered base two logarithmic. Only protein groups (listed under 'Protein names') with ratios determined in both experiments are included. ATP5IF1 ratios were manually calculated using data obtained for unique peptides from the N-termini of one mature form (IF1-M1) of the protein (see Dataset S8).
Dataset S8 (separate file). Peptide data for the ATPase inhibitor protein obtained in SILAC experiments comparing mitoplasts from wild-type and HAP1-ΔF6 cells. This information is obtained from the MaxQuant evidence file after the processing of SILAC peptide pair data. The peptide data is from digitonin protein extracts of mitoplast samples prepared with a 1:1 mixture of HAP1wild-type and HAP1-ΔF6 cells. Experiment 1 refers to heavy isotope labelled HAP1-ΔF6 cells mixed with light labelled wild-type cells, and experiment 2 vice versa. Only peptide data that provided ratio information is included. When fewer than three MULTI-MSMS ratios were obtained ISO-MSMS ratios were also included to determine the protein ratio, represented by the median peptide ratio.

Dataset S9 (separate file). Proteins identified in SILAC experiments comparing overexpressed tagged-subunit b and overexpressed tagged-subunit j.
This information is the output from Perseus after processing of MaxQuant SILAC peptide pair data. Samples were affinity purified from mitoplasts prepared from a 1:1 mixture of Flp-In™ T-REx™ HEK293T-Δδ cells overexpressing either tagged-subunit b or tagged-subunit j, analyzed by SDS-PAGE and gel sections digested with trypsin. Experiment 1 refers to heavy isotope labelled tagged-subunit j cells mixed with light labelled tagged subunit-b cells, and experiment 2 vice versa. Perseus processing removed proteins identified in both a decoy database (created in MaxQuant by reversing protein entries) and a contaminant database, experiment 2 ratios were inverted, and the protein ratios rendered base two logarithmic. Only protein groups (listed under 'Protein names') with ratios determined in both experiments are included. Tagged and endogenous subunit-b or subunit j ratios were manually calculated using data obtained from different migration positions on the gel (see Datasets S10 and S11).
Dataset S10 (separate file). Peptide data for the ATP synthase subunit b protein obtained in SILAC experiments comparing mitoplasts from Flp-In™ T-REx™ HEK293T-Δδ cells overexpressing either tagged-subunt b or tagged-subunit j. This information is obtained from the MaxQuant evidence file after the processing of SILAC peptide pair data. Samples were affinity purified from digitonin solubilized mitoplast material prepared from a 1:1 mixture of Flp-In™ T-REx™ HEK293T-Δδ cells overexpressing either tagged-subunit b or tagged-subunit j, analyzed by SDS-PAGE and gel sections digested with trypsin. Experiment 1 refers to heavy isotope labelled subunit-j cells mixed with light labelled subunit-b cells, and experiment 2 vice versa. Only peptide data that provided ratio information is included. Gel section 7 corresponds to the tagged-b peptides and gel section 8 those from the endogenous subunit b. When fewer than three MULTI-MSMS ratios were obtained ISO-MSMS ratios were also included to determine the protein ratio, represented by the median peptide ratio.
Dataset S11 (separate file). Peptide data for the ATP synthase subunit-j protein obtained in SILAC experiments comparing mitoplasts from Flp-In™ T-REx™ HEK293T-Δδ cells overexpressing either tagged-subunt b or tagged-subunit j. This information is obtained from the MaxQuant evidence file after the processing of SILAC peptide pair data. Samples were affinity purified from digitonin solubilized mitoplast material prepared from a 1:1 mixture of Flp-In™ T-REx™ HEK293T-Δδ cells overexpressing either tagged-subunit b or tagged-subunit j, analyzed by SDS-PAGE and gel sections digested with trypsin. Experiment 1 refers to heavy isotope labelled subunit-j cells mixed with light labelled subunit-b cells, and experiment 2 vice versa. Only peptide data that provided ratio information is included. Gel section 10 corresponds to the tagged-subunit j peptides and gel section 11 those from the endogenous subunit-j. When fewer than three MULTI-MSMS ratios were obtained ISO-MSMS ratios were also included to determine the protein ratio, represented by the median peptide ratio.