An essential bacterial-type cardiolipin synthase mediates cardiolipin formation in a eukaryote

Mauro Serricchio and Peter Bütikofer1

Institute of Biochemistry and Molecular Medicine, University of Bern, 3012 Bern, Switzerland

Edited by Gottfried Schatz, University of Basel, Reinach, Switzerland, and approved February 23, 2012 (received for review December 30, 2011) Cardiolipin is important for bacterial and mitochondrial stability phosphate to yield phosphatidylglycerophosphate, which sub- and function. The final step in cardiolipin biosynthesis is catalyzed sequently is dephosphorylated to phosphatidylglycerol (PG). by cardiolipin synthase and differs mechanistically between The final biosynthetic step in CL formation, catalyzed by Cls, prokaryotes and eukaryotes. To study the importance of cardio- differs between prokaryotes and eukaryotes. In prokaryotes, PG lipin synthesis for mitochondrial integrity, membrane protein and a phosphatidyl moiety from a second PG are condensed to CL, complex formation, and cell proliferation in the human and animal whereas in eukaryotes, PG and CDP-diacylglycerol are used as pathogenic protozoan parasite, Trypanosoma brucei, we gener- substrates for CL formation. The catalytic sites of almost all pro- ated conditional cardiolipin synthase-knockout parasites. We karyotic Cls contain phospholipase D (PLD)-like motifs, whereas found that cardiolipin formation in T. brucei procyclic forms is those of eukaryotic Cls contain motifs of enzymes belonging to the catalyzed by a bacterial-type cardiolipin synthase, providing ex- CDP-alcohol phosphatidyltransferase family, such as phosphatidyl- perimental evidence for a prokaryotic-type cardiolipin synthase and phosphotransferases (22–24). in a eukaryotic organism. Ablation of enzyme expression resulted The protozoan parasite Trypanosoma brucei is the causative in inhibition of de novo cardiolipin synthesis, reduction in cellular agent of human African sleeping sickness and nagana in do- cardiolipin levels, alterations in mitochondrial morphology and mestic animals. Basic research to understand the biology of Af- function, and parasite death in culture. By using immunofluores- rican trypanosomes to combat the deadly diseases they cause led cence microscopy and blue-native gel electrophoresis, cardiolipin to the discovery of several key biological processes that later synthase was shown to colocalize with inner mitochondrial mem- were identified in other eukaryotes as well, such as antigenic brane proteins and to be part of a large protein complex. During variation (25), transsplicing (26), glycosylphosphatidylinositol- depletion of cardiolipin synthase, the levels of cytochrome oxidase anchoring of proteins (27), and RNA editing (28). In addition, subunit IV and cytochrome c1, reflecting mitochondrial respiratory very recently, a mitochondrial outer-membrane translocase of complexes IV and III, respectively, decreased progressively. bacterial origin mediating import of nuclear-encoded proteins into mitochondria has been identified and characterized in T. mitochondria | phospholipid | trypanosome brucei (29). The bacterial-type translocase is related to members of the Omp85 protein superfamily involved in membrane in- he anionic glycerophospholipid cardiolipin (CL) is a mem- sertion of bacterial outer-membrane proteins, providing experi- Tbrane constituent characteristic of prokaryotes and the inner mental evidence for the previously proposed placement of mitochondrial membrane of eukaryotes (1, 2). It has a dimeric trypanosomes at the root of the eukaryotic phylogenetic tree (30). structure consisting of two phosphatidyl moieties linked to glyc- T. brucei alternates between its hosts in a complex life cycle. In erol, providing CL with unique physical properties within a bi- the blood of mammals, T. brucei bloodstream forms meet their layer (reviewed in ref. 3). CL organizes into membrane domains energy requirement by metabolizing blood glucose for ATP and participates in the formation and maintenance of dynamic production via glycolysis (31). In contrast, T. brucei procyclic protein–lipid and protein–protein interactions (reviewed in ref. forms in the tsetse fly midgut use amino acid oxidation as the 4). In bacteria, CL levels have been found to increase in the major energy-generating pathway (32). This drastic change in stationary growth phase (5) as a result of up-regulation of car- substrate metabolism during parasite differentiation is accom- diolipin synthase (Cls) activity (6) and in response to osmotic panied by morphological and functional changes of the single stress (7) by up-regulating Cls transcription and osmosensory mitochondrion, involving a switch from a largely inactive or- protein recruitment (8). However, CL is not strictly necessary for ganelle in bloodstream-form trypanosomes to a fully developed bacterial growth, because Cls-deficient Escherichia coli (9) and mitochondrion with high rates of oxidative phosphorylation in Bacillus subtilis (10) are viable. However, Cls-deficient E. coli procyclic forms (33). Based on recent findings showing that mi- retain residual amounts of CL (9), suggesting that other enzymes tochondrial morphology and function in T. brucei are dependent may contribute to CL production. In eukaryotes, CL is required on mitochondrial membrane lipid composition (34), we investi- for proper function of key mitochondrial enzymes and proteins gated the biosynthetic pathway of the mitochondrion-specific involved in oxidative phosphorylation and certain mitochondrial membrane lipid, CL, in T. brucei and studied its role in the transport systems (reviewed in refs. 11–13). Mitochondrial CL formation of the mitochondrial membrane protein complex. We associates with respiratory complexes I (NADH-ubiquinone re- report that CL synthesis in T. brucei is catalyzed by a bacterial- ductase) (14), III (ubiquinol-cytochrome c reductase) (15, 16), IV type Cls, providing experimental evidence of a prokaryotic-type Cls in a eukaryote. We demonstrate that during ablation of Cls (cytochrome oxidase) (17), and V (FoF1-ATPase) (18) and sta- bilizes supercomplexes consisting of respiratory complexes III and IV (19). Despite its role in stabilizing protein complexes, CL is not essential for growth in yeast (20), although CL-deficient Author contributions: M.S. and P.B. designed research; M.S. performed research; M.S. and mutants show reduced growth when cultured on nonfermentable P.B. analyzed data; and M.S. and P.B. wrote the paper. carbon sources (21). The authors declare no conflict of interest. In both prokaryotes and eukaryotes, CL is synthesized from the This article is a PNAS Direct Submission. precursor glycerophospholipid, phosphatidic acid, and CTP to 1To whom correspondence should be addressed. E-mail: [email protected]. form the high-energy metabolite CDP-diacylglycerol (reviewed in See Author Summary on page 5927 (volume 109, number 16). ref. 22). The activated phosphatidyl group then is transferred to This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. the nucleophilic substitution 1 (sn1)-hydroxyl group of glycerol-3- 1073/pnas.1121528109/-/DCSupplemental.

E954–E961 | PNAS | Published online March 26, 2012 www.pnas.org/cgi/doi/10.1073/pnas.1121528109 Downloaded by guest on October 5, 2021 expression in inducible gene-knockout parasites, CL formation is (Fig. 1). These signature motifs also are found in putative Cls of PNAS PLUS inhibited, causing changes in mitochondrial morphology and, other kinetoplastids (Leishmania major, Trypanosoma cruzi, Try- ultimately, parasite death. In addition, we demonstrate that T. panosoma vivax, Trypanosoma congolense) and Apicomplexa brucei Cls localizes to mitochondrial membranes where it is part (Plasmodium falciparum, Toxoplasma gondii) (Fig. 1). The overall of a large protein complex. The presence of an essential bacte- amino acid sequence identity and similarity between TbCls and E. rial-type enzyme in T. brucei validates Cls as a potential drug coli Cls are 18% and 30%, respectively. target in trypanosomatids and other parasitic protozoa. TbCls Is Essential in T. brucei Procyclic Forms. Although the presence Results of bacterial-type Cls in certain early-diverging eukaryotes has Prediction of Cls Genes in Protozoan Parasites. To identify the gene been predicted previously in a comparative genomic report (23), encoding Cls in T. brucei, the predicted proteome of T. brucei CL synthesis in these organisms never has been addressed ex- fi strain 927 was blasted using eukaryotic-type Cls sequences as perimentally. In a rst attempt to study the importance of TbCls queries. Surprisingly, although CL, the product of the reaction for parasite survival, we used tetracycline-inducible RNAi to catalyzed by Cls, has been identified previously in T. brucei down-regulate TbCls expression. Unfortunately, we could not procyclic and bloodstream forms (35), our search revealed no observe any growth defect or change in CL levels after RNAi eukaryotic-type Cls homolog in T. brucei. A more refined blast against TbCls (SI Text and Fig. S1). Therefore, we decided to to identify predicted T. brucei proteins having a CDP-alcohol generate a conditional TbCls-knockout mutant lacking both phosphatidyltransferase domain, which is typical of eukaryotic- endogenous TbCls genes and expressing a tetracycline-inducible type Cls and other enzymes of the glycerophospholipid metab- ectopic copy of TbCls (Fig. 2A). First, we replaced the ORF of one TbCls allele by a blasticidin-resistance gene by homologous olism (23, 24), revealed three proteins: phosphatidylinositol recombination (using pMSBlastKO); we then introduced a tet- synthase (Tb09.160.0530), choline/ethanolamine phosphotrans- racycline-inducible ectopic copy of TbCls containing a C-termi- ferase (Tb927.10.8900), and ethanolamine phosphotransferase nal HA tag (using pMS2560HA); and finally we replaced the (Tb927.10.13290). However, all three enzymes have been char- second allele of TbCls by a phleomycin-resistance gene (using T. brucei acterized experimentally in and are involved in glycer- pMSPhleoKO) while maintaining the expression of the ectopic ophospholipid synthesis (36, 37) and thus are not part of the CL fi copy by the presence of tetracycline in the culture medium. biosynthetic pathway. These ndings suggested that T. brucei Replacement of both TbCls alleles in TbCls-knockout parasites may synthesize CL by a route not involving eukaryotic-type Cls. was analyzed by Southern blotting using a 32P-labeled DNA A subsequent blast using bacterial-type Cls sequences as fragment hybridizing to the 3′ UTR of gene Tb927.4.2560. The queries, which, in contrast to eukaryotic-type Cls contain phos- results demonstrate that the endogenous TbCls gene at 2 kbp phatidyltransferase and PLD signature domains (24), revealed (Fig. 2B, lane +/+) was replaced sequentially by the first (Fig. 2B, − − − a T. brucei gene (Tb927.4.2560) encoding a putative bacterial- lane +/ ) and then by the second (Fig. 2B, lane / ) antibiotic- type Cls with a predicted molecular mass of 73.8 kDa. Algo- resistance cassette. Because the two antibiotic-resistance genes rithmic prediction programs [Phobius (38) or transmembrane have similar lengths, the respective bands comigrate at around prediction using hidden Markov models predictor (39)] sug- 7 kbp. The presence and correct genomic integration of the two gested that the translated protein, named “TbCls,” contains two antibiotic-resistance genes in TbCls-knockout parasites was ver- C-terminal transmembrane domains and may be targeted to ified by PCR using gene-specific primers (Fig. S2). mitochondria (according to MitoCarta predictor) (40). Similar to To see if TbCls is essential for T. brucei parasites in culture, bacterial-type Cls, TbCls contains two highly conserved H(X)K TbCls-knockout cells were cultured in the absence of tetracy- (X)4D motifs found in the active sites of PLD, nucleases, and pox cline. The results show that parasite growth was reduced after envelope proteins (24). In addition, amino acid sequence align- 24 h of incubation without tetracycline and stopped after 48 h ments between TbCls and Cls from E. coli and Gram-positive (Fig. 3A). Northern blot analysis using total RNA demonstrated Streptococcus pneumoniae show that the minimal PLD-like motif that both endogenous and ectopic TbCls mRNA was absent after H(X)K(X)4D in TbCls is part of two much larger signature 48 h of growth in the absence of tetracycline (Fig. 3B). Exami- domains that are common to most bacterial Cls, R(X)HRK(X)4D nation by light microscopy revealed that parasites after 48 h (X)5–6G(X)2N and H(X)K(X)4D(X)6G(X)2N(X)D(X)2S(X)4–5E ablating TbCls expression showed reduced motility compared MICROBIOLOGY

Fig. 1. Multiple sequence alignment of bacterial-type CLs. Translated partial amino acid sequences of putative Cls genes from bacteria, kinetoplastids, and

apicomplexans containing the conserved R(X)HRK(X)4D(X)5–6G(X)2N and H(X)K(X)4D(X)6G(X)2N(X)D(X)2S(X)4–5E motifs. The minimal PLD-like motif typical for bacterial-type Cls is indicated in red; the extended conserved sequence is highlighted in blue.

Serricchio and Bütikofer PNAS | Published online March 26, 2012 | E955 Downloaded by guest on October 5, 2021 Fig. 2. Generation of inducible T. brucei TbCls-knockout parasites. (A) Strategy to generate inducible knockout parasites. In a first step, the ORF of TbCls was replaced with a blasticidin-resistance gene (BLAST) (1). After an- tibiotic selection and PCR confirmation of correct replacement, a tetracy- cline-inducible HA-tagged ectopic copy of TbCls was integrated into the genome (2). While the expression of the ectopic copy was maintained by the presence of tetracycline (TetO), the second allele of TbCls was replaced with a phleomycin-resistance gene (PHLEO) (3). (B) Southern blot analysis. Ge- − nomic DNA from the T. brucei 29–13 parental strain (+/+) and single- (+/ ) and − − double-allele ( / ) knockout cells was digested with NcoI and analyzed by Fig. 3. Characterization of TbCls-knockout parasites. (A) Growth curve. Southern blotting using a 32P-labeled probe hybridizing to the 3′ UTR of the Growth of TbCls-knockout parasites cultured in the presence (solid line) or TbCls ORF (sinuous line in A). absence (dashed line) of tetracycline to induce or shut down, respectively, expression of ectopic TbCls was followed for 4 d. (B) Northern blot analysis of TbCls-knockout parasites cultured in the presence or absence of tetracycline 32 with control parasites expressing TbCls. Because the ectopic for 48 h. TbCls mRNA was detected using a P-labeled 400-bp DNA probe hybridizing to the TbCls ORF. Ethidium bromide (EtBr)-stained rRNA was copy of TbCls contained a C-terminal HA tag, the presence of used as loading control. (C) Immunoblot analysis of total cell lysates. Proteins TbCls protein could be analyzed by immunoblotting (Fig. 3C). from parasites cultured in the presence (+) or absence (−) of tetracycline The results demonstrate that HA-tagged TbCls was expressed as were separated by SDS/PAGE. After electrotransfer to PVDF membranes, HA- an ∼70-kDa protein in TbCls-knockout parasites cultured in the tagged TbCls was detected using anti-HA antibody. Mitochondrial Vdac was presence of tetracycline, whereas the protein was absent in used as loading control. parasites cultured for 24–72 h in the absence of the antibiotic (Fig. 3C). Together, our results show that deletion of both en- previously that, after uptake into T. brucei procyclic forms, [3H] dogenous TbCls alleles, followed by ablation of ectopic TbCls 3 expression, leads to loss of TbCls protein and causes parasite glycerol is converted to [ H]glycerol-3-phosphate and is metab- olized to tritiated succinate or acetyl-CoA and can serve as death, demonstrating that TbCls is an essential protein in 3 T. brucei procyclic forms. In addition, our results show that HA- precursor for [ H]-labeled fatty acyl chain formation for mem- tagged TbCls functionally complements endogenous TbCls. brane lipid synthesis (41, 42). In line with these reports, we found that in TbCls-knockout parasites incubated in the presence of Ablation of TbCls Expression Decreases CL Levels and Blocks de Novo tetracycline, radioactivity was incorporated into all major glyc- CL Synthesis. Because Cls catalyzes the last reaction in the CL erophospholipid classes (Fig. S3B, Left), including PG and CL synthesis pathway, ablation of TbCls expression is expected to (Fig. 4B, Left). A similar labeling pattern for the glycer- result in reduced de novo CL formation and decreased levels of ophospholipid classes was observed in TbCls-depleted parasites CL in TbCls-knockout parasites. Analysis of the total phospho- cultured in the absence of tetracycline (Fig. S3B, Right), but the lipid composition in parasites cultured in the presence or ab- amount of radioactivity in CL decreased to almost undetectable sence of tetracycline using 1D TLC and lipid phosphorus levels (Fig. 4B, Right). Taken together, these results demonstrate determination showed that the relative amounts of CL in TbCls- that TbCls is involved in de novo CL synthesis and is required for depleted parasites decreased to ∼35% of the amounts in control the maintenance of CL levels in T. brucei procyclic forms. parasites after 24 h and remained at this level for a further 48 h (Fig. 4A). In contrast, no major changes were seen between TbCls in T. brucei Procyclic Forms Is Localized in the Mitochondrion. control and TbCls-deficient parasites in the relative amounts of To study the subcellular localization of TbCls in T. brucei pro- the biosynthetic precursor of CL, PG (Fig. 4A), and the other cyclic forms, we transiently expressed TbCls as a C-terminal GFP phospholipid classes (Fig. S3A). fusion protein. Fluorescence microscopy in parasites costained To study de novo synthesis of CL, TbCls-knockout parasites with the mitochondria-specific dye MitoTracker revealed a high cultured in the presence or absence of tetracycline were labeled degree of colocalization of GFP-tagged TbCls with the mito- with [3H]glycerol followed by analysis of radiolabeled phospho- chondrial dye, indicating mitochondrial localization of TbCls lipids by 1D TLC and radioisotope scanning. It has been shown (Fig. 5A). To avoid localization artifacts caused by a possibly

E956 | www.pnas.org/cgi/doi/10.1073/pnas.1121528109 Serricchio and Bütikofer Downloaded by guest on October 5, 2021 PNAS PLUS

Fig. 5. Subcellular localization of TbCls. (A) T. brucei strain 427 procyclic forms were transiently transfected with GFP-tagged TbCls (green) and stained with MitoTracker Red (red). DNA was visualized by DAPI (blue). (B and C) TbCls-knockout parasites cultured in the presence of tetracycline were costained with antibodies against HA-tagged TbCls (HA, B and C)and Cox4 (B) or Vdac (C). DNA was visualized by DAPI (blue). DIC, differential interference contrast.

panosomes showing normal staining dropped sharply after 24 h, reaching <6% of total cells by 48 h of incubation, whereas the number of parasites showing punctuate staining increased to ∼20% and that of parasites completely lacking MitoTracker Fig. 4. Analysis of CL and PG levels and de novo synthesis. (A) Lipid quan- staining increased to >70% (Fig. 6B). Together, these results fi ti cation. TbCls-knockout parasites were cultured in the presence or absence show that mitochondria in TbCls-depleted cells fragment into of tetracycline for 72 h. At indicated time points, lipids were extracted, separated by 1D TLC, and quantified by lipid phosphorus determination. The smaller vesicular structures and lose their membrane potential, data represent mean values ± SD from three independent experiments. (B) indicating that CL plays an important role in maintaining mito- Metabolic labeling of TbCls-knockout cells with [3H]-glycerol. TbCls-knock- chondrial morphology and function in T. brucei. out parasites cultured in the presence (Left) or absence (Right) of tetracy- cline for 24 h were labeled with [3H]-glycerol for 16 h. Lipids were extracted TbCls Resides in a Mitochondrial Protein Complex. Previous studies and separated by 1D TLC, and [3H]-labeled products were quantified by ra- in yeast and mammalian cells showed that CL interacts with and dioisotope detection. Only the traces for PG and CL are shown; the re- stabilizes respiratory chain protein complexes (14–19, 43). spective full scans are shown in Fig. S3. However, it is not known if Cls also is part of these (or other) inner-membrane protein complexes. To study a possible associ- ation of TbCls with mitochondrial protein complexes in T. brucei functionally inactive GFP-tagged enzyme, TbCls localization also and to assess the importance of CL in the assembly and/or sta- was studied in conditional TbCls-knockout parasites expressing bility of respiratory chain protein complexes, native protein functionally active HA-tagged TbCls (see above). Costaining of complexes were analyzed by blue-native polyacrylamide gel parasites with anti-HA antibody and an antibody against the electrophoresis (BN-PAGE) and immunoblotting (Fig. 7A). The mitochondrial inner-membrane protein cytochrome oxidase results show that, in T. brucei procyclic forms, respiratory com- subunit IV (Cox4) revealed almost complete colocalization (Fig. plex IV, detected using an antibody against Cox4, migrates with 5B). Less pronounced colocalization also was observed between the apparent molecular mass of about 740 kDa (Fig. 7A), HA-tagged TbCls and the mitochondrial outer-membrane pro- whereas respiratory complex III, detected using an antibody tein, voltage-dependent anion channel (Vdac) (Fig. 5C). To- against cytochrome c1 (Cyt c1), migrates at around 700 kDa (Fig. gether, these results demonstrate that TbCls is localized in the 7A). In addition, the outer mitochondrial membrane protein mitochondrion, where it may associate preferentially with the Vdac migrates in a protein complex of ∼200 kDa (Fig. 7A). The inner membrane. observed sizes of the complexes containing Cox4, Cyt c1, and Vdac are in good agreement with previous studies in T. brucei Ablation of TbCls Expression Affects Mitochondrial Morphology. To (29, 44, 45). Interestingly, using anti-HA antibody, we found that investigate a possible effect of reduced CL levels on mitochon- HA-tagged TbCls also is present in a high-molecular mass drial integrity, TbCls-knockout parasites during ablation of complex, migrating with an apparent molecular mass of ∼700 TbCls expression were stained with MitoTracker and the mito-

kDa (Fig. 7A). MICROBIOLOGY chondrial matrix protein Hsp60 (Fig. 6). In control parasites BN-PAGE also was used to analyze the stability of the dif- cultured in the presence of tetracycline, both markers revealed ferent protein complexes of the mitochondrial membrane in a tubular network typical of the T. brucei mitochondrion (Fig. parasites during ablation of TbCls expression and CL depletion 6A, Upper). In contrast, in parasites after ablation of TbCls ex- (Fig. 7B). In line with the above-mentioned results (Fig. 3C), we pression, both markers showed a punctuate staining (Fig. 6A, found that removal of tetracycline resulted in time-dependent Lower); in addition, we observed increasing numbers of parasites disappearance of HA-tagged TbCls (Fig. 7B, Top). Interestingly, with no detectable MitoTracker signal (see e.g., Fig. 6A, Lower a reduction also was observed for Cox4 and Cyt c1 levels (Fig. Left). Quantification of the MitoTracker signal in parasites after 7B, second and third blots); in contrast, Vdac levels increased ablation of TbCls expression revealed that the number of try- slightly (Fig. 7B, Bottom). Similar results, i.e., a time-dependent

Serricchio and Bütikofer PNAS | Published online March 26, 2012 | E957 Downloaded by guest on October 5, 2021 To study the role of TbCls for T. brucei proliferation, we generated a conditional TbCls-knockout cell line, in which ex- pression of the enzyme can be down-regulated by removal of tetracycline from the culture medium. In contrast to previous studies using Δcls strains in bacteria or eukaryotic cells, our ex- perimental approach enabled us to study possible cellular changes resulting from loss of TbCls action in an inducible and time-dependent way. Our results demonstrate that TbCls is re- sponsible for CL formation in T. brucei procyclic forms and is essential for survival of parasites in culture. The strict essentiality of TbCls for T. brucei survival is in contrast to previous studies in Saccharomyces cerevisiae (48) and E. coli (49), where Cls- knockout mutants were viable, although they showed reduced growth under certain culture conditions. Analysis of T. brucei conditional-knockout parasites by immu- nofluorescence microscopy and BN-PAGE showed that TbCls is localized in a large protein complex, most likely in the inner mi- tochondrial membrane. Based on the apparent molecular masses of the complexes, TbCls may associate with respiratory chain complex III but not with complex IV; alternatively, TbCls may be part of a different protein complex. Association of Cls activity with a mitochondrial protein complex has been demonstrated previously in yeast; however, the complex was not characterized biochemically (50). The fact that TbCls in our conditional- knockout cell line is tagged with HA may allow us to purify Fig. 6. Effect of TbCls depletion on mitochondrial structure. (A) TbCls- partially and to identify putative TbCls interaction partners. In- knockout parasites were cultured in the presence (+ tet) or absence (− tet) of terestingly, ablation of TbCls not only resulted in a decrease in tetracycline for 2 d and stained with MitoTracker Red or anti-Hsp60 as in- CL levels but also affected mitochondrial integrity. A block in dicated. DNA was visualized by DAPI (blue). (B) Quantification of mito- TbCls expression was followed by progressive loss of mitochon- chondrial staining. Mitotracker staining of TbCls-knockout parasites cultured drial membrane potential and fragmentation of mitochondria. A in the absence of tetracycline was quantified using the following criteria: similar observation has been made previously in T. brucei pro- “normal” refers to parasites showing typical tubular staining, with no more cyclic forms after blocking de novo synthesis of phosphatidyl- “ ” than two bright spots; punctuate refers to parasites showing two or more ethanolamine (34), another phospholipid constituent enriched in bright spots; “loss” refers to parasites showing no staining. The number of cells counted for quantification was 256 on day 0, 149 on day 1, 213 on day mitochondria. In addition, in parallel with the loss of TbCls, we 2, and 116 on day 3. noticed a decrease in Cox4 and Cyt c1, suggesting that TbCls affects the stability of the inner mitochondrial membrane com- plex by lowering CL levels. This interpretation is consistent with decrease in Cox4 and Cyt c1 and an increase in Vdac, were previous studies showing that CL is necessary for proper function obtained when proteins were analyzed by denaturing SDS/PAGE and stability of several mitochondrial proteins and protein com- (Fig. 7C); the mitochondrial matrix protein Hsp70, cytochrome plexes (19, 51). Alternatively, it is possible that depletion of CL c, and cytosolic elongation factor 1A (EF1a) were used as may affect mitochondrial protein levels in T. brucei by down- loading controls. The changes in protein levels were quantified regulating translation of complex proteins, as has been shown for by scanning the individual bands on immunoblots from three Cox4 translation in a yeast mutant lacking PG and CL (52). independent BN-PAGE and three independent SDS/PAGE Bacterial-type Cls catalyze CL formation using two PG mol- analyses, showing that Cox4 decreased to <35% of starting lev- ecules as substrates in a fully reversible reaction (reviewed in ref. els, whereas Cyt c1 decreased to ∼70% (Fig. 7D). Together, 22). Thus, depending on substrate availability and product re- these results show that TbCls associates with a mitochondrial quirement, bacterial-type Cls may catalyze not only CL forma- protein complex and that its depletion results in a concomitant tion but also CL degradation, as has been demonstrated in vivo reduction in Cox4 and Cyt c1. for E. coli in the presence of excess amounts of other phosphatidyl acceptors, such as mannitol (53, 54). In contrast, CL synthesis Discussion catalyzed by eukaryotic-type Cls using PG and the high-energy The phylum of Euglenozoa, to which trypanosomatids belong, intermediate CDP-diacylglycerol is an essentially irreversible re- recently has been placed at the bottom of the eukaryotic tree action. Thus, the fundamentally different reaction mechanisms (30). This repositioning was based, in part, on findings that the between bacterial- and eukaryotic-type Cls, together with the es- genes for the mitochondrial protein translocator of the outer- sentiality of TbCls for T. brucei survival, validate TbCls and other membrane complex and the origin recognition complex are ab- putative protist Cls as potential drug targets against the diseases sent from the genomes of T. brucei, T. cruzi, and Leishmania (46, caused by the respective organisms. 47). The very recent identification and functional characteriza- tion of a mitochondrial outer-membrane translocase of bacterial Materials and Methods origin in T. brucei (29) strongly supports this theory. In the Unless otherwise stated, all reagents were of analytical grade and were present report, we show that another protein of bacterial origin, purchased from Sigma Aldrich or Merck. Restriction enzymes were from Fermentas, and antibiotics were from Sigma Aldrich, Invivogen, or Invitrogen. TbCls, mediates the synthesis of the mitochondria-specific lipid, − − [1,2,3-3H]-glycerol (1 mCi mL 1,60Cimmol 1) was purchased from American CL, in T. brucei. The sequence similarity of TbCls in T. brucei Radiolabeled Chemicals Inc.. BioMax MS films were from GE Healthcare. and other parasitic protists, together with the absence of bacte- rial-type Cls in all eukaryotes except those belonging to the phyla Phylogenetic Analyses. For phylogenetic analyses and comparison of Cls pri- Euglenozoa and Apicomplexa, suggests that trypanosomatids mary sequences, the following translated protein sequences were used: may be among the most ancient living eukaryotes. Escherichia coli ZP_07688765, Streptococcus pneumoniae ZP_07341179.1,

E958 | www.pnas.org/cgi/doi/10.1073/pnas.1121528109 Serricchio and Bütikofer Downloaded by guest on October 5, 2021 PNAS PLUS

Fig. 7. PAGE analysis of mitochondrial protein complexes. (A and B) BN-PAGE and immunoblot analyses of digitonin-isolated mitochondria from TbCls- knockout parasites. Native protein complexes were detected using antibodies against HA, Cyt c1, Cox4, and Vdac. Parasites were cultured in the presence or absence of tetracycline as indicated (B). Proteins from 6 × 107 parasites were applied for TbCls, and proteins from 106 parasites were used to detect Cox4, cytochrome c, and Vdac. (C) SDS/PAGE and immunoblot analysis of TbCls-knockout parasites cultured in the presence or absence of tetracycline. Total proteins from 107 parasites were applied. (D) Quantification of proteins during ablation of TbCls. Bands detected by immunoblotting after BN-PAGE and SDS/PAGE were quantified. The values represent means ± SDs from four to six independent experiments. Control includes Hsp70, cytochrome c, and EF1a signals.

Trypanosoma brucei XP_844406.1, Trypanosoma congolense CCC90153.1, integration using primer 5′ UTR_control (SI Text) binding 50 bp upstream of Trypanosoma vivax CCC47581.1, Trypanosoma cruzi EFZ28535.1, Leishmania the 5′ recombination site. To remove tetracycline and shut down expression major XP_001686308.1, Plasmodium falciparum CAG25336.1, and Toxo- of ectopic TbCls, trypanosomes were washed twice and resuspended in plasma gondii XP_002364240.1. Sequences were aligned using the ClustalW2 culture medium lacking tetracycline. multiple sequence alignment tool (European Molecular Biology Laboratory, European Bioinformatics Institute). Construction of GFP-Tagged TbCls. To construct C-terminally enhanced GFP- tagged TbCls, the Tb927.4.2560 ORF was amplified by PCR using primers Trypanosomes and Culture Conditions. T. brucei strain 427 procyclic forms 2560_EGFP_fwd and 2560_EGFP_rev (SI Text). The PCR product was ligated were cultured at 27 °C in SDM-79 (Invitrogen) containing 5% (vol/vol) heat- into the HindIII- and XhoI-digested plasmid pG-EGFPΔLII β (57) (kindly pro- inactivated FBS. T. brucei 29–13 procyclic forms coexpressing a T7 RNA vided by Isabel Roditi, University of Bern, Bern, Switzerland), resulting in polymerase and a tetracycline repressor (55) were cultured at 27 °C in SDM- plasmid pMS2560GFP and allowing constitutive expression of C-terminally 79 containing 15% (vol/vol) heat-inactivated FBS, 25 μg/mL hygromycin, and GFP-tagged TbCls after transient transfection into T. brucei strain 427 15 μg/mL G418 (Invitrogen). Procyclic form RNAi cells were cultured in the procyclic forms. presence of an additional 2 μg/mL puromycin. TbCls conditional-knockout parasites were cultured in SDM-79 containing 15% (vol/vol) heat-inactivated Northern and Southern Blot Analyses. For Northern blot analyses, total RNA FBS, 25 μg/mL hygromycin, 15 μg/mL G418, 2 μg/mL puromycin, 5 μg/mL was extracted from 4 × 107 midlog-phase parasites using the Total SV RNA blasticidin, 0.2 μg/mL phleomycin, and 1 μg/mL tetracycline. Low-glucose Extraction Kit (Promega), and 20 μg were loaded on a 1% agarose gel. To medium (SDM-80) was prepared as described (56). control for equal loading, the gel was stained with ethidium bromide after RNA separation. After transfer onto Hybond-N+ nylon transfer membrane

TbCls Conditional-Knockout Mutants. To generate plasmids to replace the (Amersham Pharmacia Biotech) using 10× SSC buffer (150 mM Na3-citrate, endogenous TbCls genes, blasticidin- and phleomycin-resistance genes were pH 7.0, containing 1.5 M NaCl), the membrane was probed with a 32P-la- inserted into plasmid pBS SK+ (Agilent) using HindIII and BamHI sites to beled 450-bp probe of the TbCls ORF, generated using the Prime-a-Gene produce pBS-Phleo and pBS-Blast, respectively. With primers 5′ UTR_fwd, 5′ labeling system (Promega). For Southern blot analysis, 1.2 μg NcoI-digested UTR_rev, 3′ UTR_fwd. and 3′ UTR_rev (SI Text), 400-bp-long sequences genomic DNA was separated on a 1% agarose gel and transferred to flanking the Tb927.4.2560 locus were amplified. Digestion of the 5′ UTR a Hybond-N+ nylon transfer membrane. The membrane was probed with recombination site with XhoI and HindIII and of the 3′ UTR recombination a 400-bp 32P-labeled PCR product of the TbCls 3′ UTR generated with the MICROBIOLOGY site with XbaI and NotI and ligation into pBS-Phleo and pBS-Blast resulted in Prime-a-Gene labeling system. The hybridized probe was detected by auto- the final constructs pMSPhleoKO and pMSBlastKO, respectively. Before radiography using BioMax MS films in combination with intensifying screens. transfections, plasmids were digested with XhoI and NotI. The inducible HA- tagged ectopic copy of TbCls was constructed by PCR amplification of the Lipid Analyses. Total phospholipid classes of T. brucei were extracted (58), Tb927.4.2560 ORF using primers Cls_fwd and Cls_rev (SI Text) using Pfu separated by 1D TLC on Silica Gel 60 plates (Merck) in a solvent system polymerase (Promega). The resulting HA-tagged TbCls ORF was inserted into composed of chloroform:methanol:acetic acid (65:25:8, vol/vol/vol), and vi- a pLEW100-based vector (55), resulting in plasmid pMS2560HA, which allows sualized by exposure to iodine vapor. Radioactive lipids were detected on tetracycline-inducible expression of the tagged gene. Before transfection, dried TLC plates using a radioisotope detector (Berthold Technologies) and the vector was linearized with NotI. Clones were obtained by limiting di- quantified using the Rita Control software provided by the manufacturer. lution and antibiotic selection and subsequently were PCR-tested for correct For quantification of lipid phosphorus, individual spots were scraped from

Serricchio and Bütikofer PNAS | Published online March 26, 2012 | E959 Downloaded by guest on October 5, 2021 TLC plates and quantified as described previously (59). In each chromato- pended in 0.5 mL suspension buffer [20 mM Tris·HCl (pH 7.5) containing 600 graphic separation, appropriate lipid standards were carried along. mM sorbitol, and 2 mM EDTA] followed by the addition of 0.5 mL suspension buffer supplemented with 0.05% digitonin. After incubation on ice for 5 Fluorescence Microscopy. To localize GFP-tagged TbCls in T. brucei procyclic min, nonlysed cells were removed by centrifugation at 100 × g, and mito- 7 forms, 10 live parasites were labeled for 30 min with MitoTracker Red CM- chondria were isolated by centrifugation at 6,200 × g. For SDS/PAGE, pro- μ H2XRos (Invitrogen) at 0.5 M in trypanosome culture medium. After teins were separated on 10% polyacrylamide gels. BN-PAGE, which allows fi washing, parasites were xed for 3 min with 1% paraformaldehyde, the identification of protein complexes in their native, folded state, was washed, and spread on a microscope slide, followed by incubation for 5 min performed as described (61). Mitochondria from 6.2 × 107 cells were isolated in methanol at −20 °C, air drying, and mounting with Vectashield containing by digitonin fractionation (see above), suspended in 100 μL buffer A [20 mM DAPI (Vector Laboratories). For immunofluorescence microscopy, 106 cells Tris·HCl (pH 7.2), 600 mM sorbitol, 15 mM KH PO , 20 mM MgSO ] con- were allowed to adhere to a microscopy slide for 10 min. Parasites were 2 4 4 taining 1.5% (wt/vol) digitonin. Insoluble material was removed by centri- fixed with 4% paraformaldehyde for 10 min, washed with cold PBS (137 mM × NaCl, 2.7 mM KCl, 10 mM Na HPO , 1.76 mM KH PO , pH 7.4), and per- fugation for 30 min at 16,000 g. The cleared lysate was supplemented with 2 4 2 4 × meabilized with 0.2% (wt/vol) Triton X-100 in PBS. After blocking with 2% 10 BN loading buffer [5% (wt/vol)], Coomassie Brilliant Blue G-250, 500 mM BSA in PBS for 30 min, primary antibody in blocking solution was added for 6-aminocaproic acid, 100 mM Bis-Tris·HCl, pH 7.0) and was separated on 45 min. Antibodies used were mouse monoclonal anti-HA (Covance), rabbit a linear 4–13% acrylamide gel. After separation, proteins were transferred anti-Cox4, and rabbit anti-Vdac antisera (kindly provided by André to polyvinylidene fluoride membranes (Millipore) using a semidry blot ap- Schneider, University of Bern, Bern, Switzerland) at dilutions of 1:250, 1:500, paratus (Bio-Rad). Protein-specific antibodies were detected with HRP-con- and 1:1,000, respectively. After washing, the corresponding secondary flu- jugated secondary antibodies (Invitrogen) and visualized with SuperSignal orophore-conjugated antibodies goat anti-mouse Alexa Fluor 568 and goat West Pico Chemiluminescent Substrate (Thermo Scientific). Western blots anti-rabbit Alexa Fluor 488 (Invitrogen) at dilutions of 1:100 were added in were quantified using the gel analyzer function of ImageJ software (Na- blocking solution for 45 min. After washing, cells were mounted with Vec- tional Institutes of Health). Protein sizes were determined using PageRuler tashield containing DAPI. Fluorescence microscopy was performed with Plus Prestained Protein Ladder (Thermo Scientific) and HMW Native Marker a Leica SP2 using a 100× oil objective. Pictures were acquired, processed, (GE Healthcare). and 3D-deconvoluted with the Leica LAS AF Version 2.1.0 software (Leica Microsystems). ACKNOWLEDGMENTS. We thank M. Niemann and A. Schneider for helpful discussions and for providing antibodies and M. Rauch for technical SDS- and BN-PAGE. Mitochondria were isolated by digitonin extraction as assistance during part of the study. P.B. thanks R. Tedder for valuable input. described (60). Briefly, 1–3 × 108 trypanosomes were washed with wash This work was supported by Grant 31003A-130815 from the Swiss National buffer [150 mM Tris·HCl (pH 7.9), 20 mM glucose, 20 mM NaH2PO4], sus- Science Foundation.

1. Zinser E, et al. (1991) Phospholipid synthesis and lipid composition of subcellular 21. Jiang F, Rizavi HS, Greenberg ML (1997) Cardiolipin is not essential for the growth of membranes in the unicellular eukaryote Saccharomyces cerevisiae. J Bacteriol 173: Saccharomyces cerevisiae on fermentable or non-fermentable carbon sources. Mol 2026–2034. Microbiol 26:481–491. 2. White DC, Frerman FE (1967) Extraction, characterization, and cellular localization of 22. Schlame M (2008) Cardiolipin synthesis for the assembly of bacterial and mitochondrial the lipids of Staphylococcus aureus. J Bacteriol 94:1854–1867. membranes. J Lipid Res 49:1607–1620. 3. Lewis RN, McElhaney RN (2009) The physicochemical properties of cardiolipin bilayers 23. Lykidis A (2007) Comparative genomics and evolution of eukaryotic phospholipid and cardiolipin-containing lipid membranes. Biochim Biophys Acta 1788:2069–2079. biosynthesis. Prog Lipid Res 46:171–199. 4. Mileykovskaya E, Dowhan W (2009) Cardiolipin membrane domains in prokaryotes 24. Koonin EV (1996) A duplicated catalytic motif in a new superfamily of phos- and eukaryotes. Biochim Biophys Acta 1788:2084–2091. phohydrolases and phospholipid synthases that includes poxvirus envelope proteins. 5. Short SA, White DC (1971) Metabolism of phosphatidylglycerol, lysylphosphatidylglycerol, Trends Biochem Sci 21:242–243. – and cardiolipin of Staphylococcus aureus. J Bacteriol 108:219 226. 25. Cross GA (1977) Antigenic variation in trypanosomes. Am J Trop Med Hyg 26:240–244. 6. Hiraoka S, Matsuzaki H, Shibuya I (1993) Active increase in cardiolipin synthesis in the 26. Sutton RE, Boothroyd JC (1986) Evidence for trans splicing in trypanosomes. Cell 47: fi stationary growth phase and its physiological signi cance in Escherichia coli. FEBS Lett 527–535. – 336:221 224. 27. Ferguson MA, Homans SW, Dwek RA, Rademacher TW (1988) The glycosylphos- 7. Romantsov T, Stalker L, Culham DE, Wood JM (2008) Cardiolipin controls the osmotic phatidylinositol membrane anchor of Trypanosoma brucei variant surface glycoprotein. stress response and the subcellular location of transporter ProP in Escherichia coli. Biochem Soc Trans 16:265–268. – J Biol Chem 283:12314 12323. 28. Blum B, Bakalara N, Simpson L (1990) A model for RNA editing in kinetoplastid 8. Romantsov T, et al. (2007) Cardiolipin promotes polar localization of osmosensory mitochondria: “Guide” RNA molecules transcribed from maxicircle DNA provide the transporter ProP in Escherichia coli. Mol Microbiol 64:1455–1465. edited information. Cell 60:189–198. 9. Nishijima S, et al. (1988) Disruption of the Escherichia coli cls gene responsible for 29. Pusnik M, et al. (2011) Mitochondrial preprotein translocase of trypanosomatids has cardiolipin synthesis. J Bacteriol 170:775–780. a bacterial origin. Curr Biol 21:1738–1743. 10. López CS, Alice AF, Heras H, Rivas EA, Sánchez-Rivas C (2006) Role of anionic 30. Cavalier-Smith T (2010) Kingdoms Protozoa and Chromista and the eozoan root of phospholipids in the adaptation of Bacillus subtilis to high salinity. Microbiology 152: the eukaryotic tree. Biol Lett 6:342–345. 605–616. 31. Michels PA, Bringaud F, Herman M, Hannaert V (2006) Metabolic functions of 11. Klingenberg M (2009) Cardiolipin and mitochondrial carriers. Biochim Biophys Acta glycosomes in trypanosomatids. Biochim Biophys Acta 1763:1463–1477. 1788:2048–2058. 32. Vickerman K (1985) Developmental cycles and biology of pathogenic trypanosomes. 12. Schlame M, Ren M (2009) The role of cardiolipin in the structural organization of Br Med Bull 41:105–114. mitochondrial membranes. Biochim Biophys Acta 1788:2080–2083. 33. Brown RC, Evans DA, Vickerman K (1973) Changes in oxidative metabolism and 13. Bogdanov M, Mileykovskaya E, Dowhan W (2008) Lipids in the assembly of membrane ultrastructure accompanying differentiation of the mitochondrion in Trypanosoma proteins and organization of protein supercomplexes: Implications for lipid-linked brucei. Int J Parasitol 3:691–704. disorders. Subcell Biochem 49:197–239. 34. Signorell A, et al. (2009) Perturbation of phosphatidylethanolamine synthesis affects 14. Sharpley MS, Shannon RJ, Draghi F, Hirst J (2006) Interactions between phospholipids and NADH:ubiquinone oxidoreductase (complex I) from bovine mitochondria. mitochondrial morphology and cell-cycle progression in procyclic-form Trypanosoma – Biochemistry 45:241–248. brucei. Mol Microbiol 72:1068 1079. 15. Schägger H, et al. (1990) Phospholipid specificity of bovine heart bc1 complex. Eur J 35. Richmond GS, et al. (2010) Lipidomic analysis of bloodstream and procyclic form – Biochem 190:123–130. Trypanosoma brucei. Parasitology 137:1357 1392. 16. Hayer-Hartl M, Schägger H, von Jagow G, Beyer K (1992) Interactions of phospholipids 36. Martin KL, Smith TK (2006) Phosphatidylinositol synthesis is essential in bloodstream – with the mitochondrial cytochrome-c reductase studied by spin-label ESR and NMR form Trypanosoma brucei. Biochem J 396:287 295. spectroscopy. Eur J Biochem 209:423–430. 37. Signorell A, Rauch M, Jelk J, Ferguson MA, Bütikofer P (2008) Phosphatidylethanolamine 17. Robinson NC, Zborowski J, Talbert LH (1990) Cardiolipin-depleted bovine heart in Trypanosoma brucei is organized in two separate pools and is synthesized exclu- cytochrome c oxidase: Binding stoichiometry and affinity for cardiolipin derivatives. sively by the Kennedy pathway. J Biol Chem 283:23636–23644. Biochemistry 29:8962–8969. 38. Käll L, Krogh A, Sonnhammer EL (2007) Advantages of combined transmembrane 18. Eble KS, Coleman WB, Hantgan RR, Cunningham CC (1990) Tightly associated topology and signal peptide prediction—the Phobius web server. Nucleic Acids Res 35 cardiolipin in the bovine heart mitochondrial ATP synthase as analyzed by 31P nuclear (Web Server issue):W429–432. magnetic resonance spectroscopy. J Biol Chem 265:19434–19440. 39. Krogh A, Larsson B, von Heijne G, Sonnhammer EL (2001) Predicting transmembrane 19. Pfeiffer K, et al. (2003) Cardiolipin stabilizes respiratory chain supercomplexes. J Biol protein topology with a hidden Markov model: Application to complete genomes. Chem 278:52873–52880. J Mol Biol 305:567–580. 20. Tuller G, et al. (1998) YDL142c encodes cardiolipin synthase (Cls1p) and is non- 40. Zhang X, et al. (2010) The Trypanosoma brucei MitoCarta and its regulation and essential for aerobic growth of Saccharomyces cerevisiae. FEBS Lett 421:15–18. splicing pattern during development. Nucleic Acids Res 38:7378–7387.

E960 | www.pnas.org/cgi/doi/10.1073/pnas.1121528109 Serricchio and Bütikofer Downloaded by guest on October 5, 2021 41. Stephens JL, Lee SH, Paul KS, Englund PT (2007) Mitochondrial fatty acid synthesis in 52. Su X, Dowhan W (2006) Translational regulation of nuclear gene COX4 expression by PNAS PLUS Trypanosoma brucei. J Biol Chem 282:4427–4436. mitochondrial content of phosphatidylglycerol and cardiolipin in Saccharomyces 42. van Weelden SW, van Hellemond JJ, Opperdoes FR, Tielens AG (2005) New functions cerevisiae. Mol Cell Biol 26:743–753. for parts of the Krebs cycle in procyclic Trypanosoma brucei, a cycle not operating as 53. Shibuya I, Hiraoka S (1992) Cardiolipin synthase from Escherichia coli. Methods a cycle. J Biol Chem 280:12451–12460. Enzymol 209:321–330. 43. Acehan D, et al. (2011) Cardiolipin affects the supramolecular organization of ATP 54. Shibuya I, Yamagoe S, Miyazaki C, Matsuzaki H, Ohta A (1985) Biosynthesis of novel – synthase in mitochondria. Biophys J 100:2184 2192. acidic phospholipid analogs in Escherichia coli. J Bacteriol 161:473–477. 44. Acestor N, et al. (2011) Trypanosoma brucei mitochondrial respiratome: Composition 55. Wirtz E, Leal S, Ochatt C, Cross GA (1999) A tightly regulated inducible expression system – and organization in procyclic form. Mol Cell Proteomics 10:1 14. for conditional gene knock-outs and dominant-negative genetics in Trypanosoma brucei. 45. Sharma S, Singha UK, Chaudhuri M (2010) Role of Tob55 on mitochondrial protein Mol Biochem Parasitol 99:89–101. biogenesis in Trypanosoma brucei. Mol Biochem Parasitol 174:89–100. 56. Lamour N, et al. (2005) Proline metabolism in procyclic Trypanosoma brucei is down- 46. Schneider A, Bursać D, Lithgow T (2008) The direct route: A simplified pathway for regulated in the presence of glucose. J Biol Chem 280:11902–11910. protein import into the mitochondrion of trypanosomes. Trends Cell Biol 18:12–18. 57. Urwyler S, Studer E, Renggli CK, Roditi I (2007) A family of stage-specific alanine-rich 47. Godoy PD, et al. (2009) Trypanosome prereplication machinery contains a single proteins on the surface of epimastigote forms of Trypanosoma brucei. Mol Microbiol functional orc1/cdc6 protein, which is typical of archaea. Eukaryot Cell 8:1592–1603. 63:218–228. 48. Jiang F, Gu Z, Granger JM, Greenberg ML (1999) Cardiolipin synthase expression is 58. Bligh EG, Dyer WJ (1959) A rapid method of total lipid extraction and purification. essential for growth at elevated temperature and is regulated by factors affecting – mitochondrial development. Mol Microbiol 31:373–379. Can J Biochem Physiol 37:911 917. 49. Shibuya I, Miyazaki C, Ohta A (1985) Alteration of phospholipid composition by 59. Signorell A, Jelk J, Rauch M, Bütikofer P (2008) Phosphatidylethanolamine is the combined defects in phosphatidylserine and cardiolipin synthases and physiological precursor of the ethanolamine phosphoglycerol moiety bound to eukaryotic consequences in Escherichia coli. J Bacteriol 161:1086–1092. elongation factor 1A. J Biol Chem 283:20320–20329. 50. Zhao M, Schlame M, Rua D, Greenberg ML (1998) Cardiolipin synthase is associated 60. Schneider A, Bouzaidi-Tiali N, Chanez AL, Bulliard L (2007) ATP production in with a large complex in yeast mitochondria. J Biol Chem 273:2402–2408. isolated mitochondria of procyclic Trypanosoma brucei. Methods Mol Biol 372: 51. Zhang M, Mileykovskaya E, Dowhan W (2005) Cardiolipin is essential for organization 379–387. of complexes III and IV into a supercomplex in intact yeast mitochondria. J Biol Chem 61. Becker T, et al. (2008) Biogenesis of the mitochondrial TOM complex: Mim1 promotes 280:29403–29408. insertion and assembly of signal-anchored receptors. J Biol Chem 283:120–127. MICROBIOLOGY

Serricchio and Bütikofer PNAS | Published online March 26, 2012 | E961 Downloaded by guest on October 5, 2021