Trypanosoma brucei Cytochrome c1 Is Imported into MitochondriaAlong an Unusual Pathway*

Received for publication, December 19, 2002, and in revised form, February 3, 2003Published, JBC Papers in Press, February 10, 2003, DOI 10.1074/jbc.M212956200

Jeffrey W. Priest‡ and Stephen L. Hajduk§

From the Departments of Biochemistry and Molecular Genetics, Schools of Medicine and Dentistry, University of Alabamaat Birmingham, Birmingham, Alabama 35294

In most eukaryotic organisms, cytochrome c1 is en-coded in the nucleus, translated on cytosolic ribosomes,and directed to its final destination in the mitochondrialinner membrane by a bipartite, cleaved, amino-terminalpresequence. However, in the kinetoplastids and eugle-noids, the cytochrome c1 protein has been shown to lacka cleaved presequence; a single methionine is removedfrom the amino terminus upon maturation, and the se-quence upstream of the heme-binding site is generallyshorter than that of the other eukaryotic homologs. Wehave used a newly developed mitochondrial protein im-port assay system from Trypanosoma brucei to demon-strate that the T. brucei cytochrome c1 protein is im-ported along a non-conservative pathway similar to thatdescribed for the inner membrane carrier proteins ofother organisms. This pathway requires external ATPand an external protein receptor but is not absolutelydependent on a membrane potential or on ATP hydrol-ysis in the mitochondrial matrix. We propose the cyto-chrome c1 import in T. brucei is a two-step process firstinvolving a membrane potential independent transloca-tion across the outer mitochondrial membrane followedby heme attachment and a membrane potential-depend-ent insertion into the inner membrane.

Most of the mitochondrial proteins are encoded within thenucleus, synthesized on cytosolic ribosomes, and directed totheir final mitochondrial destination by cleavable amino-termi-nal presequences (reviewed in Refs. 1 and 2). In many cases,mitochondrial protein import is receptor-mediated, is depend-ent upon an inner mitochondrial membrane potential, andrequires the hydrolysis of ATP both in the cytosol and in themitochondrial matrix. Constellations of proteins in the innerand outer mitochondrial membranes and in the intermem-brane space (IMS)1 have been implicated in the recognition,transfer, and integration of proteins from the cytosol (reviewed

in Ref. 3). Many of the imported mitochondrial proteins (nota-bly the Rieske iron-sulfur protein (ISP) subunit of the cyto-chrome c reductase complex) have a bipartite, amino-terminalpresequence that directs the proteins along a conservative sort-ing pathway; the first part of the presequence directs the pro-tein into the mitochondrial matrix, and the second part of thesignal causes the protein to be sorted into or through the innermitochondrial membrane to its final destination (4).

The cytochrome c1 subunit of the cytochrome c reductasecomplex of most eukaryotic organisms is synthesized as a pre-protein with a long (60–80 amino acids) amino-terminal exten-sion that has been shown to direct the import of the protein intothe mitochondrion in the presence of a membrane potential(5–8). Although the cytochrome c1 presequence does possessthe characteristics of a typical conservative sorting signal andis processed in two distinct steps on opposite sides of the innermitochondrial membrane, the precise pathway used for theimport of this protein has remained controversial (9, 10). Incontrast to the typical conservatively sorted preprotein, thecytochrome c1 preprotein does not appear to require the pres-ence of matrix ATP for import and sorting (11–13). To explainthis observation, some researchers (14) have suggested that thepreprotein may bypass the ATP-requiring matrix chaperonesby “looping” through the matrix on its way to the inner mem-brane along the conservative sorting pathway. Others (15–17)have suggested that a stretch of hydrophobic residues locatedwithin the amino-terminal presequence may act to arrest thetransfer of the preprotein within the inner mitochondrial mem-brane (stop-transfer pathway), thus preventing the maturepart of the protein from having any contact with the matrix.Regardless of the exact pathway, there is general agreementthat cytochrome c1 protein import requires a membrane poten-tial and external ATP (but not matrix ATP) and that thepreprotein is processed first by the matrix processing proteaseand then by the Imp2 protease on the outer surface of the innermitochondrial membrane (9, 12, 18–21). In addition, cyto-chrome c1 heme lyase, an enzymatic activity that is localized inthe IMS, is required to form two covalent thioester bondsbetween the vinyl groups of heme b and the cysteines in theCXXCH motif of the apoprotein (9, 22) before the protein can beassembled into the reductase complex. Recent work has shownthat the carboxyl-terminal region of the yeast cytochrome c1

protein contains a previously unrecognized targeting signalthat acts in a membrane potential-dependent fashion to insertthe �-helical membrane anchor of the protein into the innermembrane so that the mature protein has the proper Nout-Cin

orientation (23). Apparently, both the amino- and the carboxyl-terminal targeting sequences are required by the yeast systemfor proper membrane insertion and orientation because the

* This work was supported by National Institutes of Health Grants(to S. L. H.) and Postdoctoral Fellowship AI08259 (to J. W. P.). Thecosts of publication of this article were defrayed in part by the paymentof page charges. This article must therefore be hereby marked “adver-tisement” in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

‡ Present address: Division of Parasitic Diseases, Centers for DiseaseControl and Prevention, Mail Stop F-13, Bldg. 23, Rm. 1025, 4770Buford Hwy. N.E., Atlanta, GA 30341-3724.

§ To whom correspondence should be addressed. Tel.: 205-934-6033;Fax: 205-934-6096; E-mail: shajduk@uab.edu.

1 The abbreviations used are: IMS, intermembrane space; ISP, Rieskeiron-sulfur protein subunit of the mitochondrial cytochrome c reductasecomplex; AAC, ATP/ADP carrier protein; HSP70�RV, truncated form ofmitochondrial heat shock 70-kDa protein; MOPS, 4-morpholinepro-panesulfonic acid; CHAPS, 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid detergent; CCCP, carbonylcyanide m-chlorophen-ylhydrazone; PMSF, phenylmethylsulfonyl fluoride; PVDF,

polyvinylidene difluoride; MOM, mitochondrial outer membrane; MIM,mitochondrial inner membrane.

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 278, No. 17, Issue of April 25, pp. 15084–15094, 2003© 2003 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.

This paper is available on line at http://www.jbc.org15084

by guest on April 4, 2018

http://ww

w.jbc.org/

Dow

nloaded from

carboxyl-terminal signal, by itself, could direct the import tothe intermembrane space of only a limited fragment of theyeast cytochrome c1 protein.

Our previous analysis of the kinetoplastid c1 cytochromeshas shown that an essential element that is required by boththe stop-transfer pathway and the conservative sorting path-way models, namely a cleaved amino-terminal presequence, istotally absent; a single methionine is removed from the aminoterminus of the kinetoplastid c1 cytochromes to form the ma-ture protein (24). The absence of a cleaved presequence onthese c1 cytochromes suggested that they might be importedinto the mitochondrion along an alternative type of pathwaythat bypasses the mitochondrial matrix altogether.

In Neurospora crassa, several types of proteins are importedinto the mitochondrion in the absence of a cleavable prese-quence and along alternative pathways. The cytochrome c im-port pathway into the IMS does not utilize a receptor nor doesit require a membrane potential or the hydrolysis of ATP (re-viewed in Ref. 25). Import, which has been shown to require thepresence of heme, a source of reducing equivalents, and afunctional cytochrome c heme lyase, is thought to be driven bya conformational change that results from holoprotein forma-tion (26–28). Cytochrome c heme lyase is itself imported intothe IMS in the absence of a cleaved presequence. Import re-quires a membrane receptor but is not dependent upon eitherATP hydrolysis or a membrane potential (29, 30). The enzymeis thought to be associated with the mitochondrial membrane.The ATP/ADP carrier protein (AAC) and other inner mem-brane metabolite carrier molecules are imported along a recep-tor-mediated pathway to the outer surface of the inner mito-chondrial membrane in the absence of a membrane potentialand without matrix ATP hydrolysis (31–35). External ATP isrequired to maintain import competence (36). However, in con-trast to the other pathways, a membrane potential is requiredfor the assembly of the AAC dimer and insertion of the func-tional complex into the inner membrane (33, 34). Because thekinetoplastid cytochrome c1 must be inserted into the innermembrane, the AAC-type pathway seemed the most likely can-didate of the alternative pathways that have been describedso far.

In this work we demonstrate that the Trypanosoma bruceicytochrome c1 is indeed imported along an alternative path-way. The import, membrane insertion, and proper orientationof the mature protein in the membrane are likely directed by acarboxyl-terminal signal sequence similar to the one describedin the yeast system. Given that the T. brucei Rieske ISP sub-unit of the cytochrome c reductase complex is processed differ-ently in these organisms than in other eukaryotes (37), wesuggest that the assembly of the reductase complex may bedifferent in the trypanosomes.

MATERIALS AND METHODS

Construction of Clones, Transcription, and Translation—A PCR wasused to generate a coding sequence clone of the T. brucei cytochrome c1

using a TaqI genomic clone of the gene as target.2 Deoxyoligonucleo-tides XD-1 (5�-ATGGCAGGTAAGAAAGCTCAC-3�) and OR-21 (5�-CTCTCGCTCCACATCATACAGGC-3�) were used for hot start PCRwith AmpliTaq DNA polymerase as directed by the manufacturer(PerkinElmer Life Sciences). The resulting 816-bp product (GenBankTM

accession number AF102980; Fig. 1A) was cloned into the SmaI restric-tion site of Bluescript SK� plasmid (Stratagene). Cleavage of the clonewith BamHI generated a template for RNA synthesis with T7RNA polymerase.

A cDNA clone (pPNE-6) of the mitochondrial matrix HSP70 fromCrithidia fasciculata in Bluescript SK� (GenBankTM accession numberM95682) was kindly provided by Dr. Paul Englund (The Johns HopkinsUniversity, Baltimore) (38). Because the 1944-bp coding sequence did

not transcribe and translate efficiently in vitro, the clone was shortenedby excision of an EcoRV fragment from the 3� end. The open readingframe of the resulting clone included the first 678 bp of the HSP70coding sequence and terminated at an in-frame stop codon 54 bp down-stream of the EcoRV restriction site in the Bluescript plasmid sequence.Thus the 26-kDa truncated HSP70 translation product (HSP70�RV)contained the 20-amino acid, cleaved, amino-terminal presequence, 206amino acids of the mature HSP70 sequence, and 18 amino acids ofunrelated sequence from the plasmid. The clone was cleaved with HinfIto generate a template for RNA synthesis with T3 RNA polymerase. Theclone for the T. brucei Rieske ISP of the cytochrome bc1 complex wascleaved and transcribed as described previously (37).

5�-7-Methyl-GpppG capped mRNA was synthesized from the plasmidtemplates using either the mMessage mMachine in vitro transcriptionkit (Ambion) or the method described previously (37). If the lattermethod was used, the cap was often deleted from the cytochrome c1

reaction because this was not found to be necessary for efficient trans-lation. In vitro translations were performed using a rabbit reticulocytetranslation system (Promega) with [35S]methionine (PerkinElmer LifeSciences) as described previously (37).

Standard Mitochondrial Protein Import and Analysis—T. brucei(TREU 667) procyclic cells were grown in shaking cultures, collected bycentrifugation, and lysed in a nitrogen cavitation bomb as describedpreviously (37). A crude mitochondrial fraction was isolated by differ-ential centrifugation and stored in 50% glycerol storage buffer (37). Amodified lysis, storage, and dilution buffer (0.25 M sucrose, 20 mM

MOPS/KOH (MOPS/KOH, MOPS buffer adjusted to pH 7.2 with KOH),pH 7.2, and 2 mM EDTA) (SME-20) was used to improve the reproduc-ibility of the import reactions (39). Protein import reactions were initi-ated by the addition of 5 �l of rabbit reticulocyte translation mixture to20 �l of mitochondria (about 100 �g of protein from 1.5 � 108 cells) inbasal import buffer composed of 0.25 M sucrose, 20 mM MOPS/KOH, pH7.2, 80 mM KCl, 5 mM MgCl2, 5 mM dithiothreitol, and 1 mg/ml fattyacid-free bovine serum albumin (final concentrations). The “complete”import buffer used in some reactions was made by supplementing basalimport with an ATP-regenerating system (10 mM creatine phosphate,0.1 mg/ml creatine phosphokinase, and 2 mM ATP) and an energizingmixture (8 mM potassium ascorbate, 0.2 mM N,N,N�,N�-tetramethylphen-ylenediamine, and 5 mM NADH). Reactions were terminated by trans-fer to 4 °C, and the mitochondria were collected by centrifugation at14,000 � g for 10 min at 4 °C. Unless otherwise indicated, pellets wereresuspended in 50 �l of SME-20 buffer containing 30 �g/ml proteinaseK (Roche Molecular Biochemicals) and digested at room temperaturefor 15 min. Antimycin A (15 �M) and oligomycin (30 �M) or carbonyl-cyanide m-chlorophenylhydrazone (CCCP) (50 �M) and valinomycin (0.5�M) were included in the indicated reactions to dissipate the membranepotential and thereby decrease the possibility of continued import dur-ing the protease digestion step. Following the addition of 1 volume ofice-cold SME-20 containing 4 mM phenylmethylsulfonyl fluoride(PMSF), the mitochondria were again collected by centrifugation. La-beled proteins were resolved on 12% SDS-polyacrylamide gels, visual-ized by fluorography, and quantitated by storage phosphor autoradiog-raphy (Amersham Biosciences model 400E PhosphorImager) asdescribed previously (37). The analysis of imported proteins in thepresence of proteinase K and the detergent CHAPS was performed byacetone precipitation as described previously (37).

Import Followed by Sucrose Gradient and Western Blot Analysis—Scaled up import reactions (500-�l final volume containing 100 �l ofreticulocyte lysate and the mitochondria from �3.0 � 109 cells) wereperformed for ISP, cytochrome c1, and cytochrome c1 in the presence of100 �M CCCP using basal import buffer supplemented with the ATP-regenerating system described above. Following a 30-min incubation atroom temperature, the mitochondria were collected by centrifugationand resuspended in 500 �l of SME-20 buffer containing 30 �g/ml ofproteinase K, 30 �M oligomycin, and 15 �M antimycin A. After a 15-mindigestion at room temperature, 500 �l of SME-20 buffer containing 4mM PMSF were added, and the mitochondria were again collected bycentrifugation. The mitochondria from each import reaction were re-suspended in 500 �l of SME-20 containing 1 mM PMSF and loaded ontothe top of linear sucrose gradients as described by Opperdoes et al. (40)(1.18–2.0 M sucrose with 25 mM Tris, pH 7.4, and 1 mM EDTA on top ofa cushion of 0.8 ml of 2.5 M sucrose in the same buffer). The gradientswere centrifuged for 17 h at 30,000 rpm (60,000 � g at the sample zone)at 6 °C in a SW41Ti rotor (Beckman Instruments). Gradients werefractionated (12 times, 1 ml) from the bottom of the centrifuge tubesusing a pump and a thin capillary tube to minimize mixing. The frac-tions were numbered in reverse order so that fraction 1 represented thetop (least dense) fraction of the gradient. Each sucrose gradient fraction2 J. W. Priest and S. L. Hajduk, unpublished observations.

Import of Cytochrome c1 by Trypanosome Mitochondria 15085

by guest on April 4, 2018

http://ww

w.jbc.org/

Dow

nloaded from

was slowly diluted with 9 ml of 20 mM MOPS, pH 7.2, with gentlevortexing and then centrifuged for 15 min at 15,700 � g at 4 °C in aswinging bucket rotor (10,000 rpm; Beckman Instruments JS 13.1rotor). Mitochondrial pellets were dissolved in 50 �l of SDS sampleloading buffer with 2 mM PMSF.

The proteins from each fraction (20 �l/lane) were resolved on 12%polyacrylamide minigels and electrotransfered onto polyvinylidene di-fluoride (PVDF) membranes (Immobilon P, Millipore). The blots wereprobed with a rabbit polyclonal antibody raised against the T. bruceicytochrome c1.3 The rabbit serum was diluted 1:200 in phosphate-buffered saline (0.85% NaCl and 10 mM Na2HPO4, pH 7.2) (PBS) con-taining 0.3% Tween 20 (Tween/PBS). Bound antibodies were detectedwith a biotin-labeled rat anti-rabbit IgG monoclonal antibody (1:1000 inTween/PBS) (Zymed Laboratories Inc., S. San Francisco, CA) and alka-line phosphatase-labeled streptavidin (1:1000 in Tween/PBS) (Invitro-gen). Nitro blue tetrazolium and 5-bromo-4-chloro-3-indolylphosphatewere used to visualize the bound antibodies. As an additional confir-mation that the anti-cytochrome c1 antibody was indicating the correctlocation of the mitochondria on the sucrose gradient, one of the blotswas probed with a mouse monoclonal antibody against the Rieske ISPof the cytochrome bc1 complex (III.30.21.3 ascites diluted 1:125 inTween/PBS) (41). A biotin-labeled rat anti-mouse IgG monoclonal an-tibody was used as the secondary antibody for this blot (1:1000 inTween/PBS; Zymed Laboratories Inc.). The Western blots were sprayedwith En3Hance and exposed to x-ray film as directed by the manufac-turer (PerkinElmer Life Sciences).

Two additional 12% polyacrylamide gels (10 �l/lane from the cyto-chrome c1 import reaction gradient) were run to examine the proteinprofiles of the sucrose gradient fractions. One gel was transferred to aPVDF membrane, which was stained with AuroDye Forte as directed bythe manufacturer (Amersham Biosciences). For a more accurate indi-cation of the relative protein contents of each gradient fraction, thesecond gel was stained with Coomassie Brilliant Blue R-250 (42).

Pretreatment of Mitochondria with Proteinase K—Mitochondria from�3 � 108 cells (containing about 200 �g of protein) were resuspended in100 �l of SME-20 buffer with proteinase K at various concentrations(0.1, 0.5, 1.5, 4.5, and 15 �g/ml). After a 15-min incubation at 4 °C, thesamples were diluted with 9 volumes of cold SME-20 containing 1 mM

PMSF, and the mitochondria were collected by centrifugation for 10min at 4 °C. Control reactions (no addition during the initial incubationstep at 4 °C) were diluted with 900 �l of either SME-20 buffer alone,SME-20 buffer with 1 mM PMSF, or SME-20 buffer with 1 mM PMSFand 1.5 �g of proteinase K (equivalent to the amount of protease foundin the 15 �g/ml digestion). Mitochondrial pellets were washed once(without resuspension) with 150 �l of SME-20 with 1 mM PMSF. Themitochondrial pellets were then resuspended in 40 �l of import bufferwith an ATP-regenerating system and the energizing mixture and wereincubated with 10 �l of labeled rabbit reticulocyte lysate for 30 min atroom temperature. Mitochondria were collected by centrifugation at4 °C and resuspended in 100 �l of SME-20 buffer. One-half of eachsample was collected by centrifugation, run on a 12% SDS-polyacryl-amide gel, and stained with Coomassie Blue. The other half of eachsample was treated with 30 �g/ml proteinase K and analyzed as de-scribed for the standard import reaction.

Sequence Analysis—Sequences were aligned using the ClustalW pro-gram of Thompson et al. (43). The trypanosome protein sequence wasanalyzed for transmembrane �-helical domains using the PHDhtmprogram (44).

Miscellaneous Reagents—All reagents were purchased from Sigmaunless otherwise stated. ATP and ADP were made as 100 mM concen-trated stocks at pH 7.2 and kept at �80 °C until use. A concentratedstock (1000 units/ml) of potato apyrase (grade VIII) was made in 20 mM

MOPS/KOH, pH 7.2. Atractyloside (potassium salt) was dissolved inwater at a concentration of 10 mM. An inhibitory concentration of 400�M atractyloside was chosen for the assays based upon studies done inEuglena gracilis by Datta and Kahn (45). Valinomycin, CCCP, oligo-mycin, antimycin-A (mixture), and PMSF were made as 100-fold con-centrated stocks in ethanol.

RESULTS

T. brucei Cytochrome c1 Lacks a Presequence—Fig. 1A showsthe sequence of the T. brucei cytochrome c1 in alignment withthe Saccharomyces cerevisiae and E. gracilis homologs (Gen-BankTM accession numbers CAA25375 and JQ0021, respec-

tively) (5, 46). The yeast protein has a 61-amino acid prese-quence that is required for mitochondrial protein import (5). Aswith the E. gracilis cytochrome c1 and the other kinetoplastidc1 cytochromes that have been cloned to date (24), the T. bruceisequence does not have an amino-terminal extension thatmight act as a signal for protein import into the mitochondrion;the initial methionine (Met-1) of the protein aligns in closeproximity with the mature amino termini of the yeast and E.gracilis proteins (Met-62 and Gly-2, respectively) (indicated ingreen). If the T. brucei cytochrome c1 is processed in the samemanner as the C. fasciculata protein (24, 47), the mature aminoterminus is formed by the removal of a single methionineresidue to reveal Ala-2 (green arrowhead).

Given that the kinetoplastid and euglenoid c1 cytochromeslack the amino-terminal targeting sequence found in othereukaryotes, what sequence could act in its place? The trypano-some and other kinetoplastid cytochromes c1 (but not the E.3 Z. A. Wood and S. L. Hajduk, unpublished observations.

FIG. 1. Amino acid sequence comparison of T. brucei and N.crassa c1 cytochromes. A, the translated amino acid sequence of T.brucei (T.b.) cytochrome c1 from DNA sequence AF102980 was alignedwith the S. cerevisiae (S.c.) cytochrome c1 protein sequence (GenBankTM

accession number CAA25375) (5) and with the E. gracilis (E.g.) proteinsequence (JQ0021) (46) by using the ClustalW program (43). Amino-and carboxyl-terminal segments were optimized manually. Residuesthat are identical between the three sequences are indicated in yellow,and conservative substitutions are indicated in blue. The mature aminotermini of the yeast and E. gracilis proteins are shaded green, and thepredicted mature amino terminus of the trypanosome protein is indi-cated by a green arrowhead. Two clusters of positively charged residuesnear the amino terminus of the trypanosome protein sequence areindicated by italics and are shaded in red. Amino acids in the yeast (23)and trypanosome sequences that are predicted to span the lipid bilayerare indicated by an underline, and sequences that may form an am-phipathic �-helix are indicated in boldface. B, �-helical plot of residues239–256 of trypanosome cytochrome c1. Positively charged residues areindicated by � and hydrophobic residues are circled.

Import of Cytochrome c1 by Trypanosome Mitochondria15086

by guest on April 4, 2018

http://ww

w.jbc.org/

Dow

nloaded from

gracilis protein) do contain a cluster of positively charged res-idues at the amino terminus (Fig. 1A, italics) that might play arole in protein import (24). Although their precise role is un-clear, concentrations of positively charged residues have beenfound at the amino terminus of a number of imported trypano-some mitochondrial proteins (37, 48, 49). The T. brucei proteinalso has a region near the carboxyl terminus that is predictedto form a membrane-spanning �-helix (Fig. 1A, amino acids222–238 indicated by underlines). An �-helical plot of the 18residues that follow the transmembrane domain (amino acids239–256 indicated in boldface) shows that this region may forman amphipathic �-helix with the positively charged residuesand the hydrophobic residues localized on opposite sides of thestructure (Fig. 1B). The membrane spanning �-helix (Fig. 1A,underlines) and the amphipathic �-helix (boldface) identified inthe yeast protein by Arnold et al. (23) align with the predictedhelices in the trypanosome protein. In the yeast protein, thesecarboxyl-terminal regions have been implicated as a secondarysignal for the membrane potential-dependent insertion of cyto-chrome c1 into the inner mitochondrial membrane.

Cytochrome c1 and Truncated HSP70 Are Imported into Iso-lated Mitochondria—In earlier work, we demonstrated that acrude mitochondrial fraction isolated from T. brucei procycliccells by nitrogen cavitation lysis and differential centrifugationwas competent for the import of the ISP subunit of the cyto-chrome bc1 complex but did not import a non-mitochondrialprotein, luciferase (37). Aliquots of the crude mitochondrialfraction were incubated with in vitro translated, [35S]methi-onine-labeled T. brucei cytochrome c1, and the amount of im-ported protein was determined following digestion with in-creasing amounts of proteinase K (Fig. 2, Cyt c1). In parallelreactions (Fig. 2, HSP70�RV), we used the truncated form ofthe C. fasciculata mitochondrial HSP70 protein as a controlprotein for mitochondrial import. Like the ISP (37), the HSPshould be targeted to the mitochondrial matrix along an ATP-

dependent and membrane potential-dependent import path-way, and it contains a cleavable, amino-terminal presequencethat should be removed in a single digestion step inside thematrix. Based upon recovery of the [35S]methionine label, 42%of the input cytochrome c1 and 32% of the input HSP70�RVprotein were found in association with the mitochondria follow-ing centrifugation and a SME-20 buffer wash step (Fig. 2, lane1 and lane 5, respectively). Significant proportions of the boundcytochrome c1 and HSP70�RV proteins (7 and 50%, respec-tively) were resistant to digestion with 15 �g/ml proteinase K(Fig. 2, lane 2 and lane 6), and the recoveries did not changesignificantly even when the proteinase K concentration wasincreased to 60 �g/ml (5 and 43%; lanes 4 and 8, respectively).The observed level of cytochrome c1 import (5–7%) is similarto what has been reported for two other inner-membranecomplex subunit proteins (ISP, 6%; NdhK, 6%) in this system(37, 39). The efficiency of import of the HSP70�RV (43–50%)was consistently and significantly higher than that of thecytochrome c1.

No processing products were noted for the cytochrome c1.The HSP70�RV protein did occasionally demonstrate a novelweak band at a slightly reduced molecular weight that may bethe processed mature protein, but it was not consistent (seealso Fig. 7). The mitochondrial preparation was known to becompetent for processing because the ISP subunit of the cyto-chrome bc1 complex was cleaved to both the intermediate andmature forms in parallel experiments (data not shown). Theapparent lack of efficient processing of the HSP70�RV proteinmight be the result of either the truncation of the protein usedin our studies or the use of a Crithidia protein in a trypanosomeimport system.

Although we did not specifically test for the formation ofholocytochrome c1, we believe that the imported protein is onthe correct pathway for complex assembly. We observed an18–32% stimulation in the level of cytochrome c1 import whenNADH and flavin mononucleotide were included in the importbuffer at 5 mM and 10 �M concentrations, respectively (data notshown). Reducing equivalents and the FMN cofactor have beenreported as essential for the formation of holoprotein in othersystems (9). No stimulation in the level of ISP or HSP70�RVimport was observed in the presence of NADH and FMN (datanot shown). In summary, the crude mitochondrial preparationwas able to move both HSP70�RV and cytochrome c1 to aprotease-protected location.

Cytochrome c1 Is Imported in the Absence of a MembranePotential, but the Truncated HSP70 Is Not—Insertion of pro-teins into or across the inner membrane has been shown torequire the presence of an intact membrane potential (18). Toassess the requirement of a membrane potential for import ofthe truncated HSP70 and the cytochrome c1 proteins, importreactions were run in the presence of various membrane po-tential inhibitors. As with the conservatively sorted ISP (37),CCCP (a protonophore that uncouples oxidative phosphoryla-tion), valinomycin (a potassium ionophore), and a combinationof oligomycin plus antimycin A (inhibitors of the mitochondrialF1F0-ATPase and of electron transport at the cytochrome b,respectively) were 94–98% effective at blocking import of theHSP70�RV protein (Fig. 3A, HSP70�RV, lanes 8–10). In con-trast, import of the cytochrome c1 was decreased but was notcompletely eliminated by the presence of these membrane po-tential inhibitors (Fig. 3A, Cyt c1). Inclusion of CCCP (lane 3),valinomycin (lane 4), and antimycin A � oligomycin (lane 5) inthe import reactions resulted in a 53, 81, and 72% inhibition,respectively, as compared with the control level of import.

The observed resistance to protease digestion was not inher-ent to the translated protein; when CHAPS detergent was

FIG. 2. T. brucei cytochrome c1 and C. fasciculata HSP70�RVproteins are imported into isolated trypanosome mitochondria.Standard import reactions (25-�l final volume) in complete importbuffer as described under “Materials and Methods” were incubated for30 min at room temperature. Mitochondria were collected by centrifu-gation and then resuspended in 50 �l of SME-20 buffer containing noprotease (lane 1) or with 15 (lane 2), 30 (lane 3), or 60 (lane 4) �g/mlproteinase K (Prot. K). Reactions were incubated for 15 min at roomtemperature. Digestions were terminated by the addition of an equalvolume of ice-cold SME-20 containing 4 mM PMSF, and the mitochon-dria were collected and analyzed as described under “Materials andMethods.” Protease-resistant, labeled cytochrome c1 (lanes 1–4) andHSP70�RV (lanes 5–8) proteins were visualized by fluorography. Thepositions of molecular weight markers are indicated.

Import of Cytochrome c1 by Trypanosome Mitochondria 15087

by guest on April 4, 2018

http://ww

w.jbc.org/

Dow

nloaded from

included in the SME-20 buffer following import, at least 97% ofthe cytochrome c1 (Fig. 3B, lanes 2, 4, and 6) and 100% of theHSP70�RV (Fig. 3B, lane 8) became sensitive to digestion atthe lowest proteinase K concentration (15 �g/ml). While con-ducting these experiments, we noted that the levels of mito-chondrial protein import did not decrease significantly whenthe respiratory substrates of the energizing mixture, NADH,N,N,N�,N�-tetramethylphenylenediamine, and ascorbate, weredeleted from the reaction mixture (compare lanes 1 and 3 inFig. 3B). In general, these substrates were deleted from laterassays.

A titration of CCCP was performed to determine the concen-tration necessary to inhibit import. Reactions were terminatedafter 5 min so as to be in the linear range of the import assay(see Fig. 5). As shown in Fig. 4, 50% of the import of thetruncated HSP70 was inhibited by CCCP at a concentration of

about 21 �M. This compares favorably with the 17 �M value forthe Rieske ISP reported previously (37). In contrast, almost a10-fold increase in the CCCP concentration to about 185 �M

was required to achieve a similar level of inhibition of cyto-chrome c1 protein import. As shown in Fig. 3A, the ethanol usedas solvent for the CCCP (lanes 2 and 7) had little effect onimport by itself.

Of the kinetoplastid proteins that have thus far been used inimport assays (ISP, HSP70�RV, NdhK, and a dihydrolipoam-ide dehydrogenase construct), only the trypanosome cyto-chrome c1 has been shown to import in the absence of an innermitochondrial membrane potential (37, 39, 49). We think it isnot a coincidence that the cytochrome c1 is the only one of theseproteins that lacks a cleaved amino-terminal presequence.

Import of Cytochrome c1 Is Time- and Temperature-depend-ent—Fig. 5A shows that the accumulation of both the cyto-chrome c1 and the HSP70�RV proteins into the protease-inac-

FIG. 3. Cytochrome c1 is imported in the absence of a mem-brane potential. A, standard cytochrome c1 (upper panel) andHSP70�RV (lower panel) import reactions (25-�l final volume; 30-minincubations at room temperature) were conducted in basal importbuffer supplemented with an ATP regeneration system (No inhib., lanes1 and 6). The mitochondrial membrane potential was dissipated in theindicated reactions by preincubation (10 min at 4 °C) with 50 �M CCCP(lanes 3 and 8), 0.5 �M valinomycin (valin., lanes 4 and 9), or 15 �M

antimycin A with 30 �M oligomycin (Anti. A � Oli., lanes 5 and 10).Reactions containing 2% ethanol (EtOH, lanes 2 and 7) were included todetermined the effect of the solvent upon import. Post-import mitochon-drial pellets were digested in SME-20 buffer with 30 �g/ml proteinaseK for 15 min at room temperature and analyzed as in Fig. 2. B, standard25-�l cytochrome c1 (lanes 1 and 2) and HSP70�RV (lanes 7 and 8)import reactions in complete import buffer, cytochrome c1 reactions inbasal buffer with an ATP regeneration system (�ATP, lanes 3 and 4),and cytochrome c1 reactions in basal buffer with an ATP regenerationsystem and 50 �M CCCP (�ATP, �CCCP, lanes 5 and 6) were incubatedfor 30 min at room temperature. Post-import mitochondrial pellets wereresuspended in SME-20 buffer with 100 mM KCl and 15 �g/ml protein-ase K. CHAPS detergent (2%) was included in the indicated reactions(lanes 2, 4, 6, and 8). Digestions were incubated for 15 min at roomtemperature and then terminated by the addition of an equal volume ofice-cold SME-20 buffer with 4 mM PMSF. The proteins were precipi-tated with acetone (4 volumes of ice-cold acetone) at �20 °C overnightand collected by centrifugation (10 min at 4 °C). The resulting pelletswere dried for 5 min under vacuum before solubilization in SDS loadingbuffer and analysis by SDS-PAGE and fluorography.

FIG. 4. Import of cytochrome c1 and HSP70�RV in the pres-ence of different concentrations of CCCP. A, cytochrome c1 andHSP70�RV import reactions (25 �l) in basal import buffer supple-mented with an ATP-regenerating system were conducted in the pres-ence of various concentrations of CCCP as indicated (lanes 3–9 andlanes 12–18). The final concentration of the ethanol solvent in all of theCCCP reactions and in the control reactions in lanes 2 and 11 was 2%.In contrast to the reaction shown in Fig. 3, these reactions were incu-bated at room temperature for only 5 min so as to be in the linear rangeof the import assay (determined in Fig. 5). Mitochondria were collectedby centrifugation at 4 °C and resuspended in 50 �l of SME-20 buffercontaining 30 �g/ml proteinase K with 30 �M oligomycin and 15 �M

antimycin A to dissipate the membrane potential. After 15 min at roomtemperature, the samples were treated and analyzed by fluorographyas described previously. B, the amounts of the imported, labeled pro-teins in A were determined by storage phosphor autoradiography andexpressed as a percentage of the control level of import with 2% ethanol(A, lanes 2 and 11).

Import of Cytochrome c1 by Trypanosome Mitochondria15088

by guest on April 4, 2018

http://ww

w.jbc.org/

Dow

nloaded from

cessible compartment was time-dependent and was essentiallycomplete after 20–30 min. As demonstrated by the curves inFig. 5B, the two proteins were imported at similar rates asfollows: the t1⁄2 for cytochrome c1 was �8.0 min and that forHSP70�RV was �10.6 min. These times are slightly longerthan the 6.5-min t1⁄2 for the mitochondrial ISP reported previ-ously (37). The import of the cytochrome c1 protein was alsofound to be temperature-dependent. When the cytochrome c1

import reaction was maintained on ice for 30 min, the amountof protease-protected protein was decreased 9-fold relative tothe standard import reaction at room temperature (data notshown). Like the standard cytochrome c1 import reaction, theimport of cytochrome c1 in the presence of 200 �M CCCP (Fig.4) was also shown to be time-dependent (t1⁄2 � 8.8 min) and tobe temperature-sensitive (5-fold stimulation at room tempera-ture versus 4 °C) (data not shown). The similar kinetic profilesobserved for the import of HSP70�RV, cytochrome c1, andcytochrome c1 in the presence of CCCP as well as the markedreduction in import at low temperature would suggest that theimport of these proteins is occurring along a receptor-mediatedpathway.

Import of Cytochrome c1 Requires an Exposed, Protease-sen-sitive Component—With the notable exception of cytochrome c,the import of most proteins into the mitochondrion has beenshown to require a protein receptor that is exposed on theexterior surface of the outer mitochondrial membrane (50). Toconfirm that a surface protein receptor was required for cyto-chrome c1 import, the mitochondrial fraction was incubated at4 °C with low concentrations of proteinase K prior to the importreaction (Fig. 6A). Pretreatment of the crude mitochondrialfraction with proteinase K did not cause a drastic alteration inthe overall protein profile; only two proteins present in themitochondrial fraction were observed to be hypersensitive toprotease digestion (proteins with apparent molecular masses of70 and 17 kDa in the 4.5 and 15 �g/ml treatment lanes asindicated by arrows). In contrast, the ability of the mitochon-dria to import both cytochrome c1 and HSP70�RV proteins wasaffected by protease pretreatment (Fig. 6B). Import of bothproteins was essentially eliminated by pretreatment of themitochondria with 15 �g/ml proteinase K. Pretreatment ofmitochondria with an equivalent concentration of PMSF-inac-tivated proteinase K (15*, lane 3) had no effect on the proteinprofile or on the ability to import (compare Fig. 6A, lanes 3 and8 and the corresponding lanes in Fig. 6B). The level of proteinimport was also not affected by the presence of PMSF in theSME-20 wash buffer (lanes 1 and 2 in A and B). As shown bythe curve in Fig. 6C, protease pretreatment had a similar effecton the import of both proteins; the IC50 for cytochrome c1

import was �1.3 �g/ml whereas that for HSP70�RV importwas �1.0 �g/ml. We conclude that the import of both proteinsrequires a protein component that is exposed on the mitochon-drial surface and that is hypersensitive to protease digestion.

Cytochrome c1 Is Imported into Mitochondria in the Absenceof a Membrane Potential—To rule out the possibility that thecytochrome c1 was imported to a non-mitochondrial compart-ment in the absence of a membrane potential, we scaled up theimport reactions and then separated the membrane compo-nents by density on a linear sucrose gradient. Opperdoes et al.(40) demonstrated previously that mitochondria from isotoni-cally lysed T. brucei procyclic cells can be cleanly resolved fromthe glycosomes and other microbodies by isopycnic centrifuga-tion on a 1.18–2.0 M linear sucrose gradient (1.15–1.26 g/cm3).Mitochondria were shown to equilibrate over a broad densityrange centered at 1.17 g/cm3, whereas the glycosomes andother microbodies equilibrated as a sharp band at 1.23 g/cm3.

We imported both cytochrome c1 and the ISP in basal buffersupplemented with the ATP-regenerating system, and we alsoimported cytochrome c1 in the presence of the lowest CCCPconcentration that completely inhibited HSP70�RV (100 �M;see Fig. 4B) and then resolved the imported products into 12density fractions as shown in Fig. 7A. Fractions 7 and 8 (den-sity range 1.207–1.227 g/cm3) contained a well demarcated,visible band that, under microscopic examination, was ob-served to contain large numbers of flagella and vesicles. Nobands were obvious in the lower density range except in thetube containing the cytochrome c1 import in the presence ofCCCP. In that tube, a faint yellow band was apparent near thetop of the gradient. No 35S-labeled proteins were detected inthe most dense fractions (density �1.227 g/cm3; fractions9–12), and these fractions were not further analyzed.

Fig. 7B shows the protein profile (AuroDye), Western blotanalysis, and 35S-labeled protein distribution in the cyto-chrome c1 sucrose gradient. As demonstrated by the AuroDyepanel, the protein profile varied through the gradient; thelower density fractions were enriched for a different subset ofproteins than the higher density fractions, and fractions 7 and8 contained more total protein than the lower density fractions.

FIG. 5. Import of cytochrome c1 and HSP70�RV are time-de-pendent. A, 200-�l mitochondrial import reactions (mitochondria from�1.2 � 109 cells) in complete import buffer were conducted as describedunder “Materials and Methods.” Cytochrome c1 reactions were run with(middle panel) and without (top panel) 200 �M CCCP. At the indicatedtime points 25-�l aliquots were withdrawn from the reaction and di-luted with 75 �l of ice-cold SME-20 buffer containing sufficient CCCPand valinomycin for final concentrations of 50 and 0.5 �M, respectively.Mitochondria were collected by centrifugation at 4 °C and resuspendedin 50 �l of SME-20 buffer containing 30 �g/ml proteinase K. After a15-min incubation at room temperature, the samples were treated andanalyzed by fluorography as described previously. B, the amounts of theimported, labeled proteins in A were determined by storage phosphorautoradiography and expressed as a percentage of the maximum levelof import.

Import of Cytochrome c1 by Trypanosome Mitochondria 15089

by guest on April 4, 2018

http://ww

w.jbc.org/

Dow

nloaded from

FIG. 6. Import of both cytochrome c1 and HSP70�RV pro-teins requires an exposed, hypersensitive protein componenton the mitochondria. A, prior to use in import reactions, aliquots ofmitochondria from 3 � 108 cells were incubated in 100 �l of SME-20buffer with the indicated concentrations of proteinase K (Prot. K)(lanes 4–8) for 15 min at 4 °C. The asterisk in lane 3 indicates thatthe proteinase K in this reaction was added at the same time as thePMSF dilution buffer (see below). Samples were then diluted with900 �l of either SME-20 (lane 1), SME-20 with 1 mM PMSF and 1.5mg of proteinase K (lane 3; equivalent to 15 �g/ml in the originalincubation volume), or with SME-20 with 1 mM PMSF (lane 2 andlanes 4–8). The mitochondria were collected by centrifugation at 4 °C,washed once with SME-20 (lane 1) or SME-20 containing 1 mM PMSF(lanes 2–8), and used for 50-�l import reactions in complete buffer asdescribed under “Materials and Methods.” One-half of the post-import mitochondrial pellet was analyzed by SDS-PAGE and stainedwith Coomassie Blue. Arrows indicate the positions of two proteins atapparent molecular masses of 70 and 16 kDa that appear to behypersensitive to digestion. (The 16-kDa protein is the upper band ofa doublet.) Apparent molecular weights of the standards in lane Mare given to the left of the panel. B, the other half of each reaction wasdigested with 30 �g/ml proteinase K for 15 min at room temperatureand treated and analyzed as described previously. Cytochrome c1import is shown in the top panel, and HSP70�RV import is shown inthe bottom panel. C, the amounts of the imported, labeled proteins inlanes 3–8 in B were determined by storage phosphor autoradiographyand expressed as a percentage of the level of import observed in thecontrol lane 3 (PMSF-inactivated proteinase K).

FIG. 7. Most of the labeled cytochrome c1 protein is importedinto mitochondria. To determine the density of the vesicles thatcontained the imported, labeled proteins, scaled-up reactions (500 �lfinal volume; mitochondria from 3 � 109 cells) were conducted for 30min in basal import buffer with an ATP-regenerating system as de-scribed under “Materials and Methods.” After treatment with protease,the vesicles were centrifuged on a linear sucrose gradient with theprofile shown in A and fractionated as described under “Materials andMethods.” No proteins or radioactive counts were detected in fractions9–12, and these fractions were not further analyzed. A particulatepellet was obtained from each fraction and was dissolved SDS loadingbuffer. The proteins were resolved on a 12% polyacrylamide gel andtransferred to a PVDF membrane. B, the top panel shows an AuroDyeForte-stained blot of the sucrose gradient fractions (10 �l/lane) from thecytochrome c1 import reaction. The numbers above the panel indicatethe gradient fraction number from A. Apparent molecular weights ofthe standards run in lane M are indicated on the right side of the panel.The middle panel shows the locations of the native iron-sulfur protein(ISP) and cytochrome c1 (Cyt c1) proteins in the sucrose gradient (20�l/lane) as determined by Western blot. The Western blot was thensubjected to fluorography (bottom panel) to determine the location of thelabeled, imported cytochrome c1 in the gradient. C shows the sucrosegradient fractionation of labeled protein from an ISP import reaction. Thenumbers above the panel indicate the gradient fraction number from A. Dshows the sucrose gradient fractionation of labeled protein from a cyto-chrome c1 import reaction conducted in the presence 100 �M CCCP. Thenumbers above the panel indicate the gradient fraction number from A.The top panel shows the location of native cytochrome c1 protein (asdetermined by Western blot) in the sucrose gradient (20 �l/lane). TheWestern blot was then subjected to fluorography (bottom panel) to deter-mine the location of the labeled, imported cytochrome c1 in the gradient.

Import of Cytochrome c1 by Trypanosome Mitochondria15090

by guest on April 4, 2018

http://ww

w.jbc.org/

Dow

nloaded from

From a Coomassie Blue-stained polyacrylamide gel of the gra-dient fractions, we estimate that 80–90% of the total proteinrecovered from the gradient was present in fractions 7 and 8(data not shown). A Western blot analysis of the fractionatedproteins indicated that, although cytochrome c1 was foundthroughout the gradient, it was enriched in the lower densityfractions (fractions 2–5; density range 1.155–1.196 g/cm3). Tobe certain that the anti-cytochrome c1 antibody was giving anaccurate representation of the distribution of the mitochondrialmembranes in the gradient, the Western blot was also probedwith a monoclonal antibody against the Rieske ISP. Althoughthe monoclonal anti-ISP antibody was clearly not as sensitiveas the polyclonal anti-cytochrome c1 antibody (no protein wasdetected in fractions 1, 7, or 8), the Western blot demonstratesthat the ISP is distributed in a similar pattern in the gradi-ent with a peak indicated in fraction 4 (density range 1.176–1.186 g/cm3). Most of the imported, 35S-labeled cytochrome c1

protein (75%) was found in four upper fractions of the gradi-ent as follows: fraction 1 contained 10% of the total protectedprotein; fraction 2, 28%; fraction 3, 20%; and fraction 4, 17%(density range 1.15–1.186 g/cm3). Interestingly, although theimported cytochrome c1 protein overlapped with the mito-chondrial membranes as detected by Western blot, the frac-tion that contained the most imported cytochrome c1 (fraction2; density between 1.155 and 1.165 g/cm3) was shifted rela-tive to the fraction that contained the most mitochondrialmembranes (fraction 4; density between 1.176 and 1.186g/cm3). This observation may suggest that the majority of thewell sealed, actively respiring mitochondria are found in thelower density fraction.

An analysis of a gradient containing the mitochondria from alabeled ISP import reaction (Fig. 7C) gave similar results interms of distribution; most of the labeled, imported protein(59%) and most of the mitochondrial membranes were found infractions 2–4. The peak fraction for the import of labeled pro-tein, fraction 3, contained 32% of total labeled protein. Theimported, labeled ISP in these fractions was also processed tothe intermediate and mature size as expected.

The cytochrome c1 distribution in the gradient containing thecytochrome c1 � CCCP import reaction shows some differencesrelative to the standard cytochrome c1 and ISP gradients justdescribed. The Western blot in Fig. 7D shows that there are twopeaks of mitochondrial membranes as detected by the cyto-chrome c1 antibody: one peak in the top fraction of the gradient;a second, larger, and more broad peak in fractions 3–5. (Notethat fraction 1 in Fig. 7B contains very little native cytochromec1 protein.) The imported, labeled cytochrome c1 was also foundas two peaks in the gradient: fraction 1 contained 30% of thetotal imported protein, and fractions 3–5 contained a combined28% of the total protein (11, 9, and 8%, respectively). We thinkit possible that the observed shift of some of the mitochondriato a lighter density in the presence of the CCCP inhibitor(resulting in a second peak of imported protein in fraction 1that is not seen in the absence of inhibitor) may reflect theswelling of some of the well sealed mitochondria. CCCP-in-duced swelling of actively respiring mitochondria has beenreported previously (51) in other systems.

Thus, in each gradient we were able to separate a peak ofimported, labeled protein from the two fractions (7 and 8) thatcontained the dense glycosomes and microbodies (density range1.207 to 1.227 g/cm3). Given that the imported, labeled proteinsco-fractionate with the mitochondrial membrane marker pro-teins (as detected by Western blot) and that the densities ofthese fractions are consistent with the density range reportedpreviously for trypanosome mitochondria, we believe that thecytochrome c1 protein is imported into vesicles of mitochondrial

origin even in the absence of a membrane potential (CCCPreaction, Fig. 7D).

Cytochrome c1 Import Is Not Affected by Fluctuations inMatrix ATP Levels—By using various inhibitors of ATP syn-thesis and ATP transport, others (12) have demonstrated thatmatrix ATP is not an absolute requirement for the import ofnative cytochrome c1 into N. crassa and yeast mitochondria. Byusing atractyloside (a specific inhibitor of the ADP/ATP carrierprotein), oligomycin (an inhibitor of the mitochondrial F1F0-ATPase), and added external ATP and ADP, we attempted todetermine the relevance of matrix ATP levels to the import ofT. brucei cytochrome c1. In the first set of experiments, ATPand ADP were added to the import reaction in the absence of anATP regeneration system and in the absence of any additionalrespiration substrates in an attempt to manipulate the mito-chondrial matrix ATP pool via the ADP/ATP carrier protein. Asshown in Fig. 8A, inclusion of 2 mM ATP caused a slightincrease in the level of import of both the HSP70�RV andcytochrome c1 proteins. The addition of both 2 mM ATP and 400�M atractyloside to the import reaction caused a dramatic

FIG. 8. Import of cytochrome c1 is not significantly inhibitedby reduction of matrix ATP levels. A, 20-�l aliquots of mitochondriawere preincubated at room temperature for 10 min in basal buffereither with or without atractyloside (Atractyl.) (400 �M final concentra-tion). After the preincubation, 2 mM ADP and 2 mM ATP were added tothe indicated reactions, and import was initiated by the addition of 5 �lof labeled cytochrome c1 (upper panel) or HSP70�RV (lower panel) inreticulocyte lysate. Import reactions were incubated for 7.5 min at roomtemperature and then centrifuged at 4 °C for 10 min. Mitochondrialpellets were digested with 30 �g/ml proteinase K in SME-20 buffercontaining 30 �M oligomycin and 15 �M antimycin A (50 �l) for 15 minat room temperature and treated and analyzed as described previously.The amounts of the imported, labeled proteins were determined bystorage phosphor autoradiography and expressed as a percentage of thecontrol level of import (no additions). B, import assays were conductedas in A except that 30 �M oligomycin (Oligo.) and 2 mM ATP with anATP regeneration (ATP � regen.) system were added to the indicatedreactions after the preincubation.

Import of Cytochrome c1 by Trypanosome Mitochondria 15091

by guest on April 4, 2018

http://ww

w.jbc.org/

Dow

nloaded from

increase in the level of HSP70�RV import (from 115 to 315% ofthe control level) but had no similar effect on cytochrome c1

import. This atractyloside-specific stimulation was not peculiarto the HSP70�RV protein; the conservatively sorted ISP be-haved in a similar manner when ATP and atractyloside wereadded to the import reaction (about a 4-fold stimulation, datanot shown). Atractyloside by itself slightly stimulated the im-port of HSP70�RV protein (45% increase over basal level) buthad no such effect on cytochrome c1 import (Fig. 8B). Webelieve that the ADP present in the reticulocyte lysate trans-lation mixture may, in the presence of a functional ADP/ATPcarrier protein, rapidly deplete the ATP pool within the mito-chondrial matrix. In fact, when an additional 2 mM ADP wasadded to the import reaction, we observed a 60% decrease inthe level of HSP70�RV import, but only a 20% decrease in thelevel of cytochrome c1 import (Fig. 8A). The effect of ADP onHSP70�RV import could be completely reversed by the inclu-sion of atractyloside to block carrier protein-mediated ex-change, but no stimulation in the import of cytochrome c1 wasnoted. As would be expected for an equilibrium-driven phenom-enon, the effect of ADP on import of the HSP70�RV proteinwas largely negated by the inclusion of an equimolar amount ofATP in the reaction (Fig. 8A).

In a second set of experiments, we examined the effects of anexternal ATP regeneration system on import, and we at-tempted to reduce further matrix ATP levels with oligomycin.We noted that ATP and an ATP regeneration system togetherwere much more effective at stimulating import of theHSP70�RV protein than was ATP alone (compare the 15%increase over basal level in Fig. 8A and to the 665% increase inFig. 8B). In contrast, import of cytochrome c1 was only slightlymore efficient in the presence of the regeneration system (10%increase versus 50% increase, respectively). Inclusion of anATP regeneration system in the HSP70�RV import reactionalso significantly reduced the stimulatory effect observed withatractyloside. In the presence of ATP alone (shown in Fig. 8A),atractyloside caused a 1.7-fold increase in the level ofHSP70�RV import (115–315% of the basal level of import), but,in the presence of ATP and the regeneration system, only a0.4-fold increase was observed (765–1080% of the basal level).Taken together, these results are consistent with the proposedrole of external ADP in the depletion of the matrix ATP pooland further suggest that cytochrome c1 is not sensitive to thematrix ATP concentration.

Others (52, 53) have demonstrated that the T. brucei ATPasecan be inhibited by about 50% by oligomycin. As shown in Fig.8B, oligomycin in the absence of added ATP resulted in a 60%inhibition of HSP70�RV import but only a 15% inhibition ofcytochrome c1 import. In reactions containing either ATP andan ATP regeneration system to minimize the amount of effluxof ATP from the matrix, atractyloside to block ATP/ADP ex-change, or both ATP and atractyloside, HSP70�RV import wasconsistently more sensitive to inhibition by oligomycin thancytochrome c1 import (32% inhibition versus 6% stimulation,38% inhibition versus 21% inhibition, and 37% inhibition ver-sus 12% inhibition, respectively).

Depletion of the External ATP Blocks Cytochrome c1 Import—The 50% increase in import of cytochrome c1 observed whenboth ATP and an ATP-regenerating system were added to theimport reaction (shown above) suggested that the import of thisprotein might be sensitive to the external ATP concentration.When potato apyrase was used to hydrolyze the ATP in thereticulocyte lysate prior to the import reaction and efflux ofmatrix ATP was prevented by inhibition of the ATP/ADP car-rier protein with atractyloside, the import of HSP70�RV wasessentially abolished (5% of control) and that of the cytochrome

c1 was decreased to 15% of the control value (Fig. 9). Heat-treated apyrase enzyme had no effect on import (data notshown). Addition of ATP and the regenerating system essen-tially restored the import of the HSP�RV protein but had verylittle impact upon cytochrome c1 import. To rule out the possi-bility that contaminating activities in the apyrase preparationmight be destroying essential cofactors required for cytochromec1 import (9), we repeated the reactions in the presence of 5 mM

NADH and 10 �M FMN. Although the cofactors did stimulateimport in the control reaction (18–32% increase), they were notable to restore import in the reaction that was treated withapyrase and supplemented with ATP and the regenerationsystem (data not shown).

DISCUSSION

In the current work, we compared the import of the T. bruceicytochrome c1 protein to that of a truncated form of the C.fasciculata mitochondrial HSP70 protein. Based on our analy-sis of the trypanosome import system, we would suggest thatthe HSP70�RV protein is using a typical, presequence-directedmatrix import pathway identical to the first segment of theconservative sorting pathway reported previously for thetrypanosome ISP (37). In contrast, the cytochrome c1 protein islikely using an alternative pathway similar to that describedfor the yeast and N. crassa AAC and other inner membranecarrier proteins (31–36, 54). As shown in Fig. 10, HSP70�RVand cytochrome c1 both require external ATP in order to inter-act with protein receptors on the mitochondrial surface.Whereas the import of both the HSP70�RV and cytochrome c1

proteins was eliminated by apyrase pretreatment of reticulo-cyte lysate, only HSP70�RV import could be fully restored bythe addition of ATP and a regenerating system. We know thatboth proteins require the co-translational association of chap-erone-type factors found in the rabbit reticulocyte extract forimport because neither protein was imported from a wheatgerm extract translation even after the addition of reticulocytelysate.2,4 We hypothesize that the cytochrome c1 protein maybe released from these chaperone proteins upon ATP depletionand, perhaps because of the hydrophobic residues in the car-boxyl-terminal region of the protein or because of the lack of anamino-terminal presequence, may become locked in an import-incompetent conformation that cannot be restored upon there-introduction of ATP. This observation may be indicative ofsome unique characteristic of cytochrome c1 import into

4 K. Bertrand and S. L. Hajduk, unpublished observations.

FIG. 9. External ATP is required for import of both cyto-chrome c1 and HSP70�RV. Import assays were conducted as de-scribed in Fig. 8 except that 2 mM ATP and an ATP-regenerating system(ATP � regen.) were added to the indicated reactions after the prein-cubation. In addition, the labeled rabbit reticulocyte lysates of theindicated reactions were treated with apyrase (10 units/ml incubatedfor 15 min at 30 °C and 15 min at room temperature) prior to addition.Atractyl., atractyloside.

Import of Cytochrome c1 by Trypanosome Mitochondria15092

by guest on April 4, 2018

http://ww

w.jbc.org/

Dow

nloaded from

trypanosome mitochondria because apyrase inhibition of pro-tein import into N. crassa mitochondria could be fully restoredeven for the AAC and cytochrome c1 proteins (36, 55).

We do not yet know if the HSP70�RV and cytochrome c1

proteins utilize the same mitochondrial surface receptor, but ifthe trypanosome system is like the other eukaryotic systemsthus far examined, proteins that have cleaved presequencesuse a MOM-19-like receptor, whereas the AAC protein andothers lacking presequences use a MOM-72-like receptor (re-viewed in Refs. 2 and 3). Both proteins are then transferred tothe mitochondrial outer membrane (MOM) import complex (32,56). The major difference between the AAC pathway and thepresequence-directed conservative pathway (or the conserva-tive pathway segment leading to the matrix) is that the formercan import proteins in the absence of a membrane potential.Proteins imported along the AAC pathway are moved into theintermembrane space, either in association with a membrane-bound, multisubunit complex (ATP/ADP carrier protein) or as afreely soluble protein (dicarboxylate carrier protein) (33–35,54), whereas proteins imported into the matrix require theassociation of the mitochondrial inner membrane (MIM) com-plex, an intact membrane potential, and the ATP-dependentactivity of mitochondrial matrix HSP70. As noted in Fig. 1, theT. brucei cytochrome c1 protein (and other kinetoplastid pro-teins as well) does have a two-part, carboxyl-terminal sequencethat is consistent with the signal sequence identified by Arnoldet al. (23) at the end of the yeast cytochrome c1 protein. Thetrypanosome sequence contains a region predicted to span themembrane (residues 222–238) and an 18-amino acid sequencethat may form an amphipathic �-helix with the positivelycharged and hydrophobic residues distributed on opposite sidesof the structure. As in the yeast system, this sequence maymediate an interaction with a still undefined MIM complexand, in the presence of a membrane potential, may direct the

insertion of the protein from the IMS into the membrane in thecorrect orientation for complex assembly and holoenzymeformation.

As expected for a matrix-localized protein, HSP70�RV dem-onstrated a clear dependence upon matrix ATP for import;HSP70�RV import was quite sensitive to fluctuations in ma-trix ATP caused by the addition of external ADP (exchanged bythe mitochondrial ATP/ADP carrier protein and blocked byatractyloside) and was inhibited by about 50% by the additionof oligomycin. Given that isotonically isolated mitochondria areessentially depleted of internal ATP (53), these results wouldsuggest that the rabbit reticulocyte translation mixture, byitself, contains sufficient ATP and/or respiratory substrates tosupport a low level of import in basal buffer. The observed levelof inhibition of the mitochondrial F1F0-ATPase by oligomycin(30–60%) is similar to that reported by others (52, 53). Incontrast to the HSP70�RV, import of the cytochrome c1 waslargely unaffected by the addition of ADP or oligomycin. Eventhough neither of these experiments can be considered defini-tive (100% inhibition was not achieved), the results certainlysuggest that the trypanosome cytochrome c1 requires little orno matrix ATP for import. In this respect, the trypanosomecytochrome c1 appears to be similar to most other eukaryotic c1

cytochromes (11, 12, 14).In light of the recent discovery of the carboxyl-terminal sig-

nal in the yeast cytochrome c1 and our description of an alter-native cytochrome c1 import pathway in trypanosomes, it isinteresting to speculate that the other eukaryotic c1 cyto-chromes are, in fact, using a stop-transfer pathway (or somerelated version) for import; the first part of the pathway mayresemble the conservative pathway in that the amino-terminalpresequence may direct the membrane potential-dependentimport of a portion of the protein through the outer and innermembranes; the second part of the pathway may resemble the

FIG. 10. Proposed T. brucei mitochondrial import pathways for HSP70�RV and cytochrome c1. Both cytochrome c1 and HSP70�RV aretranslated in the cytosol and are associated with cytosolic factors or chaperones (not shown). Import of both proteins requires an association witha surface-exposed protein component (RECEPTOR) and external ATP. Import of HSP70�RV requires a membrane potential (��) and matrix ATPand likely occurs through contact sites where the mitochondrial outer membrane (MOM) and mitochondrial inner membrane (MIM) complexes arein association. Once in the matrix, the amino-terminal presequence of the HSP70�RV protein (open box) is cleaved by a processing protease to yieldthe mature protein. In contrast, import of cytochrome c1 does not require a membrane potential or matrix ATP. The protein is likely importedthrough a MOM complex into the intermembrane space (IMS), and when a membrane potential is restored, the protein is transported to the MIMcomplex and inserted into the inner membrane. In the presence of heme and reducing equivalents, cytochrome c1 heme lyase covalently attachesheme (H) to form the holoprotein that is then assembled into the cytochrome c reductase complex.

Import of Cytochrome c1 by Trypanosome Mitochondria 15093

by guest on April 4, 2018

http://ww

w.jbc.org/

Dow

nloaded from

AAC pathway in that, once across the outer membrane, thecarboxyl-terminal signal may direct membrane potential-de-pendent insertion of the mature portion of the protein into theinner membrane in the proper Nout-Cin orientation. How muchof the protein enters the matrix and whether the cytochrome c1

heme lyase might play a role in import (either by “pulling” theprotein across the outer membrane as with cytochrome c or bypreventing further translocation into the matrix) remain to bedetermined. This hybrid pathway would explain the absence ofa matrix ATP requirement and would be consistent with thecleavage of the bipartite, amino-terminal signal sequence onopposite sides of the inner mitochondrial membrane. Thetrypanosome system, therefore, is a “natural” experiment thatdemonstrates that the amino-terminal presequence-driven en-try of the protein into the matrix is not essential for cytochromec1 import and/or assembly.

Finally, it must be stated that, although we are certain thatthere is no cleaved amino-terminal presequence, we cannot becertain that there is no mitochondrial targeting signal at theamino terminus of the cytochrome c1. We and others (37, 47, 48)have noted that trypanosome amino-terminal, cleaved prese-quences are unusually short (as short as 8–9 residues) and thatthey often contain a MX1–2(K/R)(K/R) motif where X1–2 is one ortwo amino acid that are usually hydrophobic. All of the kineto-plastid c1 cytochromes thus far examined have 3–4 positivelycharged amino acids with the first 11 residues and a sequencethat would seem to satisfy this motif, M(A/G)G(K/R)(K/R) (seeRef. 24 and this work). We do not believe these residues arenecessary for cytochrome c1 import for two reasons. First, theE. gracilis cytochrome c1 protein also lacks a cleaved amino-terminal presequence yet does not have any positively chargedresidues in the first 20 residues of the sequence. Second, wehave mutated the MAGKK sequence of the trypanosome pro-tein to MAGQQ and have shown that it has no effect on importeither with or without a membrane potential.2 We tend tobelieve that the import signals for cytochrome c1 are internaland most likely downstream of the heme-binding site. We arecurrently working to elucidate further the signals involved inrecognition and import.

Acknowledgments—We thank Kathy Hancock, Karen Bertrand,Lynn Sherrer, and the members of the Hajduk laboratory for theirassistance, comments, and helpful discussions. We also thank the Adlerand Hajduk families for being gracious hosts during the completion ofthis work.

REFERENCES

1. Hartl, F.-U., Pfanner, N., Nicholson, D. W., and Neupert, W. (1989) Biochim.Biophys. Acta 988, 1–45

2. Voos, W., Martin, H., Krimmer, T., and Pfanner, N. (1999) Biochim. Biophys.Acta 1422, 235–254

3. Neupert, W. (1997) Annu. Rev. Biochem. 66, 863–9174. Hartl, F.-U., Schmidt, B., Wachter, E., Weiss, H., and Neupert, W. (1986) Cell

47, 939–9515. Sadler, I., Suda, K., Schatz, G., Kaudewitz, F., and Haid, A. (1984) EMBO J.

9, 2137–21436. Hartl, F.-U., Ostermann, J., Guiard, B., and Neupert, W. (1987) Cell 51,

1027–10377. Stuart, R. A., Nicholson, D. W., Wienhues, U., and Neupert, W. (1990) J. Biol.

Chem. 265, 20210–202198. Jensen, R. E., Schmidt, S., and Mark, R. J. (1992) Mol. Cell. Biol. 12,

4677–46869. Nicholson, D. W., Stuart, R. A., and Neupert, W. (1989) J. Biol. Chem. 264,

10156–1016810. Glick, B. S., Beasley, E. M., and Schatz, G. (1992) Trends Biol. Sci. 17,

453–45911. Stuart, R. A., Gruhler, A., van der Klei, I., Guiard, B., Koll, H., and Neupert,

W. (1994) Eur. J. Biochem. 220, 9–1812. Wachter, C., Schatz, G., and Glick, B. S. (1992) EMBO J. 11, 4787–479413. Wachter, C., Schatz, G., and Glick, B. S. (1994) Mol. Biol. Cell 5, 465–47414. Gruhler, A., Ono, H., Guiard, B., Neupert, W., and Stuart, R. A. (1995) EMBO

J. 14, 1349–135915. van Loon, A. P. G. M., and Schatz, G. (1987) EMBO J. 6, 2441–244816. van Loon, A. P. G. M., Brandli, A. W., Pesold-Hurt, B., Blank, D., and Schatz,

G. (1987) EMBO J. 6, 2433–9243917. Glick, B. S., Brandt, A., Cunningham, K., Muller, S., Hallberg, R. L., and

Schatz, G. (1992) Cell 69, 809–82218. Martin, J., Mahlke, K., and Pfanner, N. (1991) J. Biol. Chem. 266,

18051–1805719. Nunnari, J., Fox, T. D., and Walter, P. (1993) Science 262, 1997–200420. Schmidt, B., Wachter, E., Sebald, W., and Neupert, W. (1984) Eur. J. Biochem.

144, 581–58821. Schleyer, M., and Neupert, W. (1985) Cell 43, 339–35022. Ohashi, A., Gibson, J., Gregor, I., and Schatz, G. (1982) J. Biol. Chem. 257,

13042–1304723. Arnold, I., Folsch, H., Neupert, W., and Stuart, R. A. (1998) J. Biol. Chem. 273,

1469–147624. Priest, J. W., Wood, Z. A., and Hajduk, S. L. (1993) Biochim. Biophys. Acta

1144, 229–23125. Stuart, R. A., and Neupert, W. (1990) Biochimie (Paris) 72, 115–12126. Nicholson, D. W., Kohler, H., and Neupert, W. (1987) Eur. J. Biochem. 164,

147–15727. Nicholson, D. W., Hergersberg, C., and Neupert, W. (1988) J. Biol. Chem. 263,

19034–1904228. Nicholson, D. W., and Neupert, W. (1989) Proc. Natl. Acad. Sci. U. S. A. 86,

4340–434429. Lill, R., Stuart, R. A., Drygas, M. E., Nargang, F. E., and Neupert, W. (1992)

EMBO J. 11, 449–45630. Diekert, K., Kispal, G., Guiard, B., and Lill, R. (1999) Proc. Natl. Acad. Sci.

U. S. A. 96, 11752–1175731. Pfanner, N., and Neupert, W. (1987) J. Biol. Chem. 262, 7528–753632. Sollner, T., Pfaller, R., Griffiths, G., Pfanner, N., and Neupert, W. (1990) Cell

62, 107–11533. Koehler, C. M., Jarosch, E., Tokatlidis, K., Schmid, K., Schweyen, R. J., and

Schatz, G. (1998) Science 279, 369–37334. Ryan, M. T., Muller, H., and Pfanner, N. (1999) J. Biol. Chem. 274,

20619–2062735. Kubrich, M., Rassow, J., Voos, W., Pfanner, N., and Honlinger, A. (1998)

J. Biol. Chem. 273, 16374–1638136. Pfanner, N., Tropschug, M., and Neupert, W. (1987) Cell 49, 815–82337. Priest, J. W., and Hajduk, S. L. (1996) J. Biol. Chem. 271, 20060–2006938. Effron, P. N., Torri, A. F., Engman, D. M., Donelson, J. E., and Englund, P. T.

(1993) Mol. Biochem. Parasitol. 59, 191–20039. Bertrand, K. I., and Hajduk, S. L. (2000) Mol. Biochem. Parasitol. 106,

249–26040. Opperdoes, F. R., Borst, P., Bakker, S., and Leene, W. (1977) Eur. J. Biochem.

76, 29–3941. Priest, J. W., and Hajduk, S. L. (1994) Mol. Biochem. Parasitol. 65, 291–30442. Laemmli, U. K. (1970) Nature 227, 680–68543. Thompson, J. D., Higgins, D. G., and Gibson, T. J. (1994) Nucleic Acids Res. 22,

4673–468044. Rost, B., Fariselli, P., and Casadio, R. (1996) Protein Sci. 7, 1704–171845. Datta, D. B., and Kahn, J. S. (1977) J. Protozool. 24, 187–19246. Mukai, K., Wakabayashi, S., and Matsubara, H. (1989) J. Biochem. (Tokyo)

106, 479–48247. Priest, J. W., and Hajduk, S. L. (1992) J. Biol. Chem. 267, 20188–2019548. Priest, J. W., and Hajduk, S. L. (1995) Biochim. Biophys. Acta 1269, 201–20449. Hausler, T., Stierhof, Y.-D., Blattner, J., and Clayton, C. (1997) Eur. J. Cell

Biol. 73, 240–25150. Zwizinski, C., Schleyer, M., and Neupert, W. (1984) J. Biol. Chem. 259,

7850–785651. Cunarro, J., and Weiner, M. W. (1975) Biochim. Biophys. Acta 387, 234–24052. Williams, N., and Frank, P. H. (1990) Mol. Biochem. Parasitol. 43, 125–13253. Alleman, N., and Schneider, A. (2000) Mol. Biochem. Parasitol. 111, 87–9454. Zara, V., Palmisano, I., Rassow, J., and Palmieri, F. (2001) J. Mol. Biol. 310,

965–97155. Hartl, F.-U., Ostermann, J., Pfanner, N., Tropschug, M., Guiard, B., and

Neupert, W. (1987) in Cytochrome Systems: Molecular Biology and Ener-getics (Papa, S., ed) pp. 189–196, Plenum Publishing Corp., New York

56. Pfaller, R., Steger, H. F., Rassow, J., Pfanner, N., and Neupert, W. (1988)J. Cell Biol. 6, 2483–2490

Import of Cytochrome c1 by Trypanosome Mitochondria15094

by guest on April 4, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Jeffrey W. Priest and Stephen L. HajdukUnusual Pathway

Is Imported into Mitochondria Along an1 c CytochromeTrypanosoma brucei

doi: 10.1074/jbc.M212956200 originally published online February 10, 20032003, 278:15084-15094.J. Biol. Chem. 

  10.1074/jbc.M212956200Access the most updated version of this article at doi:

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/278/17/15084.full.html#ref-list-1

This article cites 55 references, 18 of which can be accessed free at

by guest on April 4, 2018

http://ww

w.jbc.org/

Dow

nloaded from