TCGAGGACATCCATCAGTACAG, HaeII; Wars: TTCCTCACTGACACGGCCAAG,

CAGAAGCACTGGAAGTGGAAGG, ATCAAGAGCAAGGTCAACAAGC, BstEII.

Designated primers for Xnp, Pdha1, Gla and Yy1 contain engineered base changes

(indicated by an asterisk) to facilitate allelic analysis. Changes do not affect relative

priming efficiencies between the alleles, as the changes occur on both (details available on

request). Xist and Hprt assays have been described29.

Received 28 October; accepted 20 November 2003; doi:10.1038/nature02222.

Published online 7 December 2003.

1. Lyon, M. F. Gene action in the X-chromosome of the mouse (Mus musculus L.). Nature 190, 372–373

(1961).

2. Mukherjee, A. B. Cell cycle analysis and X-chromosome inactivation in the developing mouse. Proc.

Natl Acad. Sci. USA 73, 1608–1611 (1976).

3. Epstein, C., Smith, S., Travis, B. & Tucker, G. Both X-chromosomes function before visible

X-chromosome inactivation in female mouse embryos. Nature 274, 500–503 (1978).

4. Kratzer, P. G. & Gartler, S. M. HGPRT activity changes in preimplantation mouse embryos. Nature

274, 503–504 (1978).

5. Adler, D. A., West, J. D. & Chapman, V. M. Expression of alpha-galactosidase in preimplantation

mouse embryos. Nature 267, 838–839 (1977).

6. Lyon, M. F. in Results and Problems in Cell Differentiation (ed. Ohlsson, R.) 73–90 (Springer,

Heidelberg, 1999).

7. Takagi, N. & Sasaki, M. Preferential inactivation of the paternally derived X-chromosome in the

extraembryonic membranes of the mouse. Nature 256, 640–642 (1975).

8. Marahrens, Y., Panning, B., Dausman, J., Strauss, W. & Jaenisch, R. Xist-deficient mice are defective in

dosage compensation but not spermatogenesis. Genes Dev. 11, 156–166 (1997).

9. Kay, G. F. et al. Expression of Xist during mouse development suggests a role in the initiation of X

chromosome inactivation. Cell 72, 171–182 (1993).

10. Matsui, J., Goto, Y. & Takagi, N. Control of Xist expression for imprinted and random X chromosome

inactivation in mice. Hum. Mol. Genet. 10, 1393–1401 (2001).

11. Huynh, K. D. & Lee, J. T. Imprinted X inactivation in eutherians: a model of gametic execution and

zygotic relaxation. Curr. Opin. Cell Biol. 13, 690–697 (2001).

12. Krietsch, W. K. et al. The expression of X-linked phosphoglycerate kinase in the early mouse embryo.

Differentiation 23, 141–144 (1982).

13. Latham, K. E., Patel, B., Bautista, F. D. M. & Hawes, S. M. Effects of X chromosome number and

parental origin on X-linked gene expression in preimplantation mouse embryos. Biol. Reprod. 63,

64–73 (2000).

14. Mayer, W., Niveleau, A., Walter, J., Fundele, R. & Haaf, T. Demethylation of the zygotic paternal

genome. Nature 3, 501–502 (2000).

15. Hall, L. L. et al. An ectopic human XIST gene can induce chromosome inactivation in

postdifferentiation human HT-1080 cells. Proc. Natl Acad. Sci. USA 99, 8677–8682 (2002).

16. Schultz, R. M. Regulation of zygotic gene activation in the mouse. Bioessays 15, 531–538 (1993).

17. Hartshorn, C., Rice, J. E. & Wangh, L. J. Differential pattern of Xist RNA accumulation in single

blastomeres isolated from 8-cell stage mouse embryos following laser zona drilling. Mol. Reprod. Dev.

64, 41–51 (2003).

18. Mak, W. et al. Mitotically stable association of polycomb group proteins Eed and Enx1 with the

inactive X chromosome in trophoblast stem cells. Curr. Biol. 12, 1016–1020 (2002).

19. Nothias, J. Y., Majumder, S., Kaneko, K. J. & DePamphilis, M. L. Regulation of gene expression at the

beginning of mammalian development. J. Biol. Chem. 270, 22077–22080 (1995).

20. Heard, E., Clerc, P. & Avner, P. X-chromosome inactivation in mammals. Annu. Rev. Genet. 31,

571–610 (1997).

21. Wutz, A. & Jaenisch, R. A shift from reversible to irreversible X inactivation is triggered during ES cell

differentiation. Mol. Cell 5, 695–705 (2000).

22. Chadwick, B. P. & Willard, H. F. SETting the stage: Eed-Enx1 leaves and epigenetic signature on the

inactive X chromosome. Dev. Cell 4, 445–447 (2003).

23. Duthie, S. M. et al. Xist RNA exhibits a banded localization on the inactive X chromosome and is

excluded from autosomal material in cis. Hum. Mol. Genet. 8, 195–204 (1999).

24. Lifschytz, E. & Lindsley, D. L. The role of X-chromosome inactivation during spermatogenesis. Proc.

Natl Acad. Sci. USA 69, 182–186 (1972).

25. Graves, J. A. M. Mammals that break the rules: Genetics of marsupials and monotremes. Annu. Rev.

Genet. 30, 233–260 (1996).

26. Lee, J. T. Molecular links between X-inactivation and autosomal imprinting: X-inactivation as a

driving force for the evolution of imprinting. Curr. Biol. 13, R242–R254 (2003).

27. Jegalian, K. & Page, D. C. A proposed path by which genes common to mammalian X and Y

chromosomes evolve to become X inactivated. Nature 394, 776–780 (1998).

28. Tanaka, S., Kunath, T., Hadjantonakis, A. K., Nagy, A. & Rossant, J. Promotion of trophoblast stem cell

proliferation by FGF4. Science 282, 2072–2075 (1998).

29. Stavropoulos, N., Lu, N. & Lee, J. T. A functional role for Tsix transcription in blocking Xist RNA

accumulation but not in X-chromosome choice. Proc. Natl Acad. Sci. USA 98, 10232–10237 (2001).

30. Carrel, L. et al. X inactivation analysis and DNA methylation studies of the ubiquitin activating

enzyme E1 and PCTAIRE-1 genes in human and mouse. Hum. Mol. Genet. 5, 391–401 (1996).

Acknowledgements We are grateful to S. Shinwa for instruction in manipulating pre-

implantation embryos, and to S. Shibata and Y. Ogawa for sharing reagents. We thank C. L. Tsai,

L. F. Zhang and B. K. Sun for critical reading of the manuscript, and all members of the laboratory

for discussion. This work was funded by the National Institutes of Health, USA, the Pew Scholar

Program, and the Howard Hughes Medical Institute.

Competing interests statement The authors declare that they have no competing financial

interests.

Correspondence and requests for materials should be addressed to J.T.L.

(lee@frodo.mgh.harvard.edu).

..............................................................

Evolutionary conservationof biogenesis of b-barrelmembrane proteinsStefan A. Paschen1, Thomas Waizenegger1, Tincuta Stan1, Marc Preuss1,Marek Cyrklaff2, Kai Hell1*, Doron Rapaport1* & Walter Neupert1

1Adolf-Butenandt-Institut fur Physiologische Chemie,Ludwig-Maximilians-Universitat Munchen, Butenandtstrasse 5,D-81377 Munchen, Germany2Max-Planck-Institut fur Biochemie, Abteilung fur Molekulare Strukturbiologie,Am Klopferspitz 18a, D-82152 Martinsried, Germany

* These authors contributed equally to this work

.............................................................................................................................................................................

The outer membranes of mitochondria and chloroplasts aredistinguished by the presence of b-barrel membrane proteins1,2.The outer membrane of Gram-negative bacteria also harboursb-barrel proteins3. In mitochondria these proteins fulfil a varietyof functions such as transport of small molecules (porin/VDAC),translocation of proteins (Tom40) and regulation of mitochon-drial morphology (Mdm10)4–7. These proteins are encoded by thenucleus, synthesized in the cytosol, targeted to mitochondria aschaperone-bound species, recognized by the translocase of theouter membrane, and then inserted into the outer membranewhere they assemble into functional oligomers8–11. Whereas someknowledge has been accumulated on the pathways of insertion ofproteins that span cellular membranes with a-helical segments,very little is known about how b-barrel proteins are integratedinto lipid bilayers and assembled into oligomeric structures12.Here we describe a protein complex that is essential for thetopogenesis of mitochondrial outer membrane b-barrel proteins(TOB). We present evidence that important elements of thetopogenesis of b-barrel membrane proteins have been conservedduring the evolution of mitochondria from endosymbioticbacterial ancestors13.

To find components involved in the import, folding and assemblyof b-barrel proteins we used a proteomic approach. On analysis bymass spectrometry of purified outer membranes of mitochondriafrom Neurospora crassa, we detected a protein of 55 kDa, hereaftercalled Tob55 (Supplementary Fig. S1a). Homologues are present inthe genomes of virtually all eukaryotes (Supplementary Fig. S1b).This protein displayed significant sequence similarity with theOmp85 protein, which is present in the outer membrane ofGram-negative bacteria (Supplementary Fig. S1c)14. Tob55, likeOmp85, is predicted to contain b-strands that are present mainlyin the carboxy-terminal part of the protein. For further analysis weinvestigated Tob55 in Saccharomyces cerevisiae. Subcellular fraction-ation experiments revealed that Tob55 is indeed located in mito-chondria (Supplementary Fig. S2). It represents an integralcomponent of the mitochondrial outer membrane, probably expos-ing an amino-terminal domain to the intermembrane space (Sup-plementary Fig. S3).

Cells harbouring the TOB55 gene under the control of the GAL10promoter were first grown on galactose-containing medium, andthen shifted to galactose-free medium. Their growth slowed downat about 40 h after the shift (Fig. 1a). This is in agreement withprevious reports demonstrating the essential nature of this gene15.Tob55 was virtually absent at 43 h after the shift (Tob55#) (Fig. 1b).Mitochondria from cells isolated 47 h after the shift (Tob55##) hadvery low levels of the b-barrel outer membrane proteins porin,Tom40 and Mdm10 for which the b-barrel structure was predicted(refs 5, 6; and S.A.P., unpublished results). The levels of other outermembrane proteins (Tom70, Tom20 and OM45) as well as thosemitochondrial proteins residing in the intermembrane space

letters to nature

NATURE | VOL 426 | 18/25 DECEMBER 2003 | www.nature.com/nature862 © 2003 Nature Publishing Group

(CCPO), inner membrane (Tim17) or the matrix (aconitase) wereessentially unaffected (Fig. 1b).

Tob55# mitochondria were analysed for their ability to importpre-proteins in vitro (Fig. 1c). They exhibited strongly reducedimport of the b-stranded outer membrane proteins porin, Tom40and Tob55 itself. On the other hand, proteins anchored to the outermembrane by a-helical segments (such as Tom20) were imported atroughly wild-type levels, as was the import of proteins destinedto the intermembrane space (CCHL), the inner membrane(pDLD-DHFR) and the matrix (pF1b). This suggests thatTob55 has an essential and specific role in the import of b-barrelproteins.

To determine the native molecular mass of Tob55, mitochondriawere lysed with digitonin and proteins were analysed by blue native

gel electrophoresis (BN–PAGE) (Fig. 2a, left panel) and sizingchromatography (Fig. 2a, right panel). A similar apparent molecu-lar mass of 220–250 kDa was obtained with both methods. Thus,Tob55 is part of an oligomeric complex called the topogenesis ofmitochondrial outer membrane b-barrel proteins (TOB) complex.It is distinct from the translocase of the outer membrane (TOM)complex as it migrates differently in BN–PAGE assays, and anti-bodies against Tob55 precipitate the TOB complex but not the TOMcomplex (Fig. 2a, left and middle panels).

We expressed a His-tagged version of Tob55 in Escherichia coli,purified it from inclusion bodies by Ni-NTA chromatography,solubilized the protein in 8 M urea and refolded it. The circulardichroism (CD) spectrum of the refolded Tob55 revealed a relativelyhigh content of b-strands (approximately 28%), comparable with

Figure 1 Tob55 is essential for the biogenesis of b-barrel proteins. a, Downregulation of

Tob55 affects cell growth. Cells from a wild-type strain (WT) or from a strain expressing

Tob55 under the control of the GAL10 promoter (Tob55(Gal10)) were shifted from

galactose- to glucose-containing medium at time zero. b, Levels of b-barrel proteins are

reduced in mitochondria from cells depleted of Tob55 for 43 h (Tob55#) or 47 h (Tob55##),

as determined by immunodecoration. c, Tob55 is required for the import of b-barrel

proteins. Radiolabelled pre-proteins were imported into wild-type or Tob55#

mitochondria. For each precursor, the results of at least three experiments are within

^15% of the indicated average value. Tom20(ext), Tom20 variant with N-terminal

extension, adopting wild-type topology.

Figure 2 Tob55 contains b-stranded segments, is present as an oligomeric assembly,

and forms channels in lipid bilayers. a, Tob55 occurs in a complex of approximately

220–250 kDa that is distinct from the TOM complex. Left panel: mitochondrial

analysis by BN–PAGE and immunodecoration with antibodies against Tob55 (left lane)

or Tom40 (right lane). Middle panel: mitochondria were lysed with digitonin and

added to protein A Sepharose beads with pre-bound Tob55 antibodies or pre-immune

serum (PI). Analysis was carried out as above. Right panel: gel filtration and

immunoblotting of fractions with antibodies against Tob55. b, Far-ultraviolet CD

spectrum of refolded recombinant Tob55 (evaluation of the spectrum suggests 28%

b-sheet, 24% a-helix and 12.5% turn structure). v, ellipticity. c, Single-channel

recordings of Tob55. Recombinant Tob55 was added directly or after pre-incubation

with anti-Tob55 IgG to a black lipid membrane. d, Histogram of channel

conductances. P(G) is the probability that a given conductance increment G was observed

(n ¼ 122 analysed). e, Voltage dependence of channel conductances. The ratio of the

conductance, G, at a given voltage, U, divided by G0 at 10 mV is shown as a function of

voltage.

letters to nature

NATURE | VOL 426 | 18/25 DECEMBER 2003 | www.nature.com/nature 863© 2003 Nature Publishing Group

that found in Tom40 and porin6 (Fig. 2b). The refolded Tob55inserted spontaneously into lipid bilayers (Fig. 2c). It formedchannels with an average conductivity of 3.7 nS at 1 M KCl(Fig. 2d). The channels partially closed at applied voltages higherthan ^70 mV (Fig. 2e). Thus, the Tob55 channel differs in itsconductivity and its voltage-dependence from mitochondrial andbacterial porins3,4,6.

We also isolated the His-tagged version of Tob55 by Ni-NTAchromatography after lysis of mitochondria from the yeast strainTob55(Gal10) with n-dodecyl-b-D-maltoside (DDM). Both nativeand recombinant Tob55 were analysed by electron microscopy aftercryo-negative staining16 (Fig. 3a–d). Ring-shaped assemblies werepredominant. Most of the rings enclosed a smaller density locatedeither centrally or slightly off-centre. Preliminary two-dimensionalaverage maps (Fig. 3e, f) revealed an outer diameter of approxi-mately 15 nm and an inner diameter of about 7–8 nm. The centraldensity measured approximately 4–5 nm. The rings exhibited a five-fold rotational symmetry (Fig. 3g, h).

In order to identify the molecular function of the TOB complex,mitochondria carrying His-tagged Tob55 were incubated withradiolabelled precursors of b-barrel proteins at low temperatureto accumulate intermediates17–19. During the import the precursorsbecame associated with Tob55 (Fig. 4a). In contrast to the precursorof porin in transit, endogenous fully assembled porin as well as theprecursor of Tom20—an outer membrane protein with a single a-helical transmembrane segment—were not co-purified with Tob55(Fig. 4a). This shows that Tob55 interacts with precursors ofb-barrel proteins on their import pathways.

To determine at which stage Tob55 is involved in the import

process, mitochondria after the import of various b-barrel proteinswere subjected to BN–PAGE, which leads to separation of importintermediates. In the case of the Tom40 precursor, 250 kDa and100 kDa complexes containing intermediates were reported thatkinetically precede formation of the mature TOM complex17,19.These intermediates were absent after import into mitochondriaisolated from Tob55-depleted cells (Fig. 4b). We chose conditionsthat favour accumulation of the 250 kDa and 100 kDa intermediatecomplexes, and investigated whether they contained Tob55. To thisend, mitochondria with accumulated intermediates were lysed,incubated with antibodies against Tob55 and then analysed byBN–PAGE. The 250 kDa but not the 100 kDa intermediate wasshifted to higher apparent molecular mass by formation of asupercomplex with the antibodies (Fig. 4c, left panel). When theantibodies were pre-incubated with an excess of Tob55 antigen, theshift was efficiently inhibited (Fig. 4c, middle panel). The presenceof Tob55 in the 250 kDa intermediate was also demonstrated by theability of antibodies against Tob55 to deplete this species from alysate of mitochondria (Fig. 4c, right panel). A similar behaviourwas observed with the precursor of Mdm10 (Fig. 4d). Thus, Tob55 isan integral component of import intermediates of b-barrel proteins.Mdm10 and Tom40 precursors associated with Tob55 were largelyresistant to degradation upon treatment of intact mitochondriawith proteases (Fig. 4e and data not shown), in agreement withprevious observations on the 250 kDa intermediate of Tom40(ref. 19). This suggests that the Tob55-associated precursor hadlargely or completely crossed the outer membrane.

To analyse whether this translocation occurs via the TOMcomplex, we blocked the latter by addition of an excess of recombi-nant pSu9(1–69)–DHFR, a precursor consisting of the targetingsignal of ATP synthase subunit 9 and dihydrofolate reductase18.Then porin was imported and Tob55 was isolated. Formation of theTob55-containing intermediate was strongly reduced when theTOM complexes were occupied (Fig. 4f). A similar effect wasobserved when Mdm10 was imported (Fig. 4g). Thus, b-barrelproteins use the TOM complex before they interact with Tob55. Infact, the TOB complex seems to be essential for complete transloca-tion across the TOM complex. In Tob55-depleted mitochondriathe 250 kDa intermediate was not formed (Fig. 4b) and Tom40precursor did not reach a protease-protected state (Fig. 4h).

A recent study suggested the participation of Mas37 in the sortingand assembly of outer membrane proteins in yeast20. As with Tob55,Mas37 was found to be present in the 250 kDa assembly inter-mediate in the Tom40 import pathway. The TOB complex inMas37-deleted cells was reduced in size by approximately 40 kDa(Supplementary Fig. S4a). Thus, Mas37 probably occurs in acomplex with Tob55, and a Tob55-containing core complex isformed in its absence. This TOB core complex maintained theability to form the assembly intermediate of Tom40 (Supplemen-tary Fig. S4b). Furthermore, the TOB core complex is able to act as afunctional insertion and assembly machinery, as cells deficient inMas37 are viable. Notably, no homologues of Mas37 were found inbacteria or eukaryotes other than yeast. Mas37 was also reported tointeract with pre-protein receptors of the outer membrane21. Thus,it might have a stabilizing effect on several outer membraneproteins. Extensive studies will be needed to understand its preciserole.

Why would b-barrel proteins of the mitochondrial membraneundergo such a complicated topogenic pathway? It seems reason-able to assume that b-barrel proteins of mitochondria are evolu-tionarily derived from bacterial outer membrane b-barrelproteins13,22. In view of the complex requirements for folding andassembly, it seems possible that the mitochondrial b-barrelproteins must follow a pathway conserved during evolution. Inbacteria, b-barrel proteins insert from the periplasmic side into theouter membrane3. Recently, a major role in this process in Neisseriameningitidis was attributed to Omp85 (ref. 14). Accumulation of

Figure 3 Electron microscopy of native and recombinant refolded TOB complexes using

cryo-negative staining with ammonium molybdate. a–d, Unprocessed fields. a, b, Images

of native TOB complexes attached to carbon support (a) and present over a hole in an

EM-carbon grid (b). c, d, Same as a and b, but for recombinant refolded Tob55.

e–h, Gallery of two-dimensionally averaged images; 15-nm particles appear most

frequently in both preparations. e, Average of 347 particles of native Tob55. f, Average of

207 particles of the recombinant refolded Tob55. g, h, Same as e and f, but after

imposing five-fold symmetry. Scale bar in a is 20 nm and pertains to a–d. The edges in

e–h are 27 nm.

letters to nature

NATURE | VOL 426 | 18/25 DECEMBER 2003 | www.nature.com/nature864 © 2003 Nature Publishing Group

b-barrel intermediates in the periplasmic space of Omp85-deficientcells was presented as the evidence for this hypothesis14, although acontradictory report postulated a role of Omp85 in lipid export23.Sequence similarity of Tob55 to the apparently ubiquitous Omp85-like proteins of Gram-negative bacteria14 is particularly high withrespect to Rickettsia prowazekii (Supplementary Fig. S1c). The latterbelongs to the a-Proteobacteria, which are thought to be evolution-ary ancestors of mitochondria13. Not only structure but alsotopology and function of the bacterial Omp85 machinery may beconserved in the TOB complex. This would require that b-barrelprecursors are also presented in mitochondria from the inner face ofthe outer membrane. Thus, the precursors first have to cross theouter membrane via the TOM complex, in order to then follow aconserved pathway of translocation into the outer membrane with

the help of the TOB complex.As the TOB complex forms a channel with an apparently large

diameter it is tempting to speculate that the channel may be largeenough to accommodate up to 16–22 b-strands, usually present inb-barrel proteins3. The b-strands could fold into their native ornear-native conformation in the channel, then to be released bylateral opening into the lipid phase of the outer membrane. Thus,the TOB complex might be viewed as an Anfinsen-type cage, inwhich the b-stranded membrane proteins can fold24.

The existence of a channel of large size in the mitochondrial outermembrane has been suggested on the basis of electrophysiologicalexperiments25. The TOB complex might indeed be the complex ofquest. Given that the Tob55 channel is large enough to allow thepassage of smaller folded proteins, it would be interesting to study

Figure 4 The TOB complex engages b-barrel proteins after their interaction with the TOM

complex. a, Tob55 interacts with precursors (prec.) of b-barrel proteins. Radiolabelled

pre-proteins were imported into the indicated mitochondria followed by Ni-NTA affinity

purification. Total, 15% of input. Authentic porin (endog.) was immunodecorated.

b, Tob55 is required for the import and assembly of Tom40. Import of radiolabelled

Tom40 was analysed by BN–PAGE. c, Tob55 is part of the 250 kDa assembly

intermediate of imported Tom40. Left and middle panels: lysed mitochondria were

incubated with antibodies or with antigen-saturated antibodies (anti-Tob55 þ Tob55p).

Right panel: mitochondrial lysate was added to Sepharose beads containing the indicated

antibodies. PI, pre-immune serum. d, Tob55 is present in an assembly intermediate of

Mdm10. Left panel: antibody shift; right panel: antibody depletion. Asterisks indicate

unspecific bands. e, The assembly intermediate of Mdm10 is protected from proteases.

T, trypsin; PK, proteinase K. f, g, b-Barrel precursors (f, porin precursor as in a;

g, Mdm10) are transported to the TOB complex via the TOM complex. De-energized

mitochondria were pre-incubated without (control) or with excess amounts of

pSu9(1–69)–DHFR (pSu9). h, Tom40 precursor is protease-sensitive in Tob55#

mitochondria. AE, alkaline extraction.

letters to nature

NATURE | VOL 426 | 18/25 DECEMBER 2003 | www.nature.com/nature 865© 2003 Nature Publishing Group

its possible additional role as a regulated conduit for the release ofcytochrome c, a key step in apoptosis. A

MethodsYeast strainsThe haploid Tob55(Gal10) strain, expressing a Tob55 protein with an N-terminal octa-histidine tag under control of the GAL10 promoter, was constructed by transformation ofthe wild-type strain YPH499 with a polymerase chain reaction fragment and homologousrecombination26. The growth rate of this strain was similar to that of the wild-type strain.For depletion of Tob55, Tob55(Gal10) cells were shifted from lactate medium containing0.1% galactose to lactate medium with 0.1% glucose.

Expression of recombinant Tob55 and refoldingA His-tagged version of Tob55 was cloned into the expression vector pQE60 (Qiagen) andexpressed in E. coli. The protein was purified from inclusion bodies by Ni-NTAchromatography and eluted with 8 M urea, 50 mM Na-phosphate, 200 mM imidazole and1 mM PMSF, pH 8.0. The purified protein was refolded by dialysis against decreasingconcentrations of urea in 50 mM Na-phosphate, 50 mM NaCl and 0.05% DDM, pH 8.0.

CD spectroscopyFar-ultraviolet CD measurements were performed using a Jasco J-715 spectrometer. FinalCD spectra were obtained by averaging eight consecutive scans, and the secondarystructure content was estimated27.

Purification of native Tob55Mitochondria were solubilized in 50 mM Na-phosphate, 300 mM NaCl, 20 mM imidazole,10% glycerol, 1 mM PMSF and 1% DDM, pH 8.0, for 30 min on ice. After a clarifying spin,solubilized material was added to Ni-NTA agarose beads. After 1 h agitation at 4 8C, beadswere washed with the same buffer containing 0.1% DDM and bound protein was elutedwith 300 mM imidazole. For co-purification of in vitro imported precursor proteins withTob55, mitochondria (250 mg) were treated as described above with the exception thatdigitonin was used as detergent.

Electron microscopySamples for electron microscopy were prepared using cryo-negative staining inammonium molybdate16. The electron microscopy grids were imaged on the CM200-FEGelectron microscope. Defocus pairs of 0.8 and 1.6 mm were recorded at the nominalmagnification of 38,000. The images were further analysed using the EM softwarepackage28. Particles for further analysis were interactively selected and two-dimensionalaverage maps were calculated without correction for contrast transfer function. A firstreference was created by reference-free alignment and then subjected to iterative rotationaland translational alignment. The particles with lowest correlation coefficient to thecurrent reference were gradually rejected. Resulting average maps were subjected torotational analysis by imposing the rotational symmetries in a range between two- toten-fold. The best symmetry fit was determined by eye.

Antibody-shift/depletion of protein complexesRadiolabelled precursor proteins were incubated with mitochondria. Samples werere-isolated and solubilized in buffer containing 1% digitonin. In the case of the antibodyshift experiments, immunoglobulin-g (IgG) was added directly to the mitochondriallysate and incubated for 1 h at 4 8C. For the immunodepletion experiments mitochondriallysate was added to protein A Sepharose beads that were pre-incubated with the indicatedantibodies. After 1 h incubation at 4 8C the beads were removed. The supernatants of theantibody depletion and the samples of the antibody shift were analysed by BN–PAGE17.

MiscellaneousIn vitro import of radiolabelled pre-proteins into isolated yeast mitochondria was performedessentially as described17,18. The ‘black lipid bilayer’ experiments were performed accordingto published procedures29. Recombinant pSu9(1–69)–DHFR was purified asdescribed18. Antibodies were raised in rabbits against the purified recombinant Tob55protein (anti-Tob55) and a peptide covering amino acid residues 1–15 (anti-Tob55-N).

Received 17 August; accepted 30 October 2003; doi:10.1038/nature02208.

1. Rapaport, D. Biogenesis of the mitochondrial TOM complex. Trends Biochem. Sci. 27, 191–197 (2002).

2. Schleiff, E. et al. Prediction of the plant b-barrel proteome: A case study of the chloroplast outer

envelope. Protein Sci. 12, 748–759 (2003).

3. Schulz, G. E. The structure of bacterial outer membrane proteins. Biochim. Biophys. Acta 1565,

308–317 (2002).

4. Blachly-Dyson, E. & Forte, M. VDAC channels. IUBMB Life 52, 113–118 (2001).

5. Hill, K. et al. Tom40 forms the hydrophilic channel of the mitochondrial import pore for preproteins.

Nature 395, 516–521 (1998).

6. Ahting, U. et al. Tom40, the pore-forming component of the protein-conducting TOM channel in the

outer membrane of mitochondria. J. Cell Biol. 153, 1151–1160 (2001).

7. Sogo, L. F. & Yaffe, M. P. Regulation of mitochondrial morphology and inheritance by Mdm10p, a

protein of the mitochondrial outer membrane. J. Cell Biol. 126, 1361–1373 (1994).

8. Endo, T. & Kohda, D. Functions of outer membrane receptors in mitochondrial protein import.

Biochim. Biophys. Acta 1592, 3–14 (2002).

9. Matouschek, A. & Glick, B. S. Barreling through the outer membrane. Nature Struct. Biol. 8, 284–286

(2001).

10. Pfanner, N. & Geissler, A. Versatility of the mitochondrial protein import machinery. Nature Rev. Mol.

Cell Biol. 2, 339–349 (2001).

11. Paschen, S. A. & Neupert, W. Protein import into mitochondria. IUBMB Life 52, 101–112 (2001).

12. Gabriel, K., Buchanan, S. K. & Lithgow, T. The alpha and the beta: protein translocation across

mitochondrial and plastid outer membranes. Trends Biochem. Sci. 26, 36–40 (2001).

13. Gray, M. W., Burger, G. & Lang, B. F. Mitochondrial evolution. Science 283, 1476–1481 (1999).

14. Voulhoux, R., Bos, M. P., Geurtsen, J., Mols, M. & Tommassen, J. Role of a highly conserved bacterial

protein in outer membrane protein assembly. Science 299, 262–265 (2003).

15. Brachat, A. et al. Analysis of deletion phenotypes and GFP fusions of 21 novel Saccharomyces cerevisiae

open reading frames. Yeast 16, 241–253 (2000).

16. De Carlo, S., El-Bez, C., Alvarez-Rua, C., Borge, J. & Dubochet, J. Cryo-negative staining reduces

electron-beam sensitivity of vitrified biological particles. J. Struct. Biol. 138, 216–226 (2002).

17. Rapaport, D. & Neupert, W. Biogenesis of Tom40, core component of the TOM complex of

mitochondria. J. Cell Biol. 146, 321–331 (1999).

18. Krimmer, T. et al. Biogenesis of porin of the outer mitochondrial membrane involves an import pathway

via receptors and the general import pore of the TOM complex. J. Cell Biol. 152, 289–300 (2001).

19. Model, K. et al. Multistep assembly of the protein import channel of the mitochondrial outer

membrane. Nature Struct. Biol. 8, 361–370 (2001).

20. Wiedemann, N. et al. Machinery for protein sorting and assembly in the mitochondrial outer

membrane. Nature 424, 565–571 (2003).

21. Gratzer, S. et al. Mas37p, a novel receptor subunit for protein import into mitochondria. J. Cell Biol.

129, 25–34 (1995).

22. Marcotte, E. M., Xenarios, I., van Der Bliek, A. M. & Eisenberg, D. Localizing proteins in the cell from

their phylogenetic profiles. Proc. Natl Acad. Sci. USA 97, 12115–12120 (2000).

23. Genevrois, S., Steeghs, L., Roholl, P., Letesson, J. J. & van der Ley, P. The Omp85 protein of Neisseria

meningitidis is required for lipid export to the outer membrane. EMBO J. 22, 1780–1789 (2003).

24. Ellis, R. J. Revisiting the Anfinsen cage. Fold. Des. 1, R9–R15 (1996).

25. Pavlov, E. V. et al. A novel, high conductance channel of mitochondria linked to apoptosis in

mammalian cells and Bax expression in yeast. J. Cell Biol. 155, 725–731 (2001).

26. Lafontaine, D. & Tollervey, D. One-step PCR mediated strategy for the construction of conditionally

expressed and epitope tagged yeast proteins. Nucleic Acids Res. 24, 3469–3471 (1996).

27. Raussens, V., Ruysschaert, J. M. & Goormaghtigh, E. Protein concentration is not an absolute

prerequisite for the determination of secondary structure from circular dichroism spectra: a new

scaling method. Anal. Biochem. 319, 114–121 (2003).

28. Hegerl, R. The EM Program Package: A platform for image processing in biological electron

microscopy. J. Struct. Biol. 116, 30–34 (1996).

29. Popp, B., Court, D. A., Benz, R., Neupert, W. & Lill, R. The role of the N and C termini of recombinant

Neurospora mitochondrial porin in channel formation and voltage-dependent gating. J. Biol. Chem.

271, 13593–13599 (1996).

Supplementary Information accompanies the paper on www.nature.com/nature.

Acknowledgements We thank U. Gartner and P. Heckmeyer for technical assistance; L. Peters,

T. Silberzahn and A. Weinzierl for their help in some experiments; S. Schmitt for outer

membranes of N. crassa mitochondria; and C. Bornhovd for help in subcellular yeast

fractionation. We are grateful to W. Baumeister for support and access to the electron microscope

facility. We thank E. Weyher-Stingl for help with the CD measurements, and R. Benz and

S. Nussberger for support with the electrophysiological experiments. This work was supported by

the Deutsche Forschungsgemeinschaft (D.R.), Sonderforschungsbereich 594, the

Bundesministerium fur Bildung und Forschung (MITOP), the Fonds der Chemischen Industrie

(W.N.), and predoctoral fellowships from the Boehringer Ingelheim Fonds (S.A.P. and T.W.).

Competing interests statement The authors declare that they have no competing financial

interests.

Correspondence and requests for materials should be addressed to W.N.

(Neupert@bio.med.uni-muenchen.de).

..............................................................

An ABC transporter with asecondary-active multidrugtranslocator domainHenrietta Venter, Richard A. Shilling, Saroj Velamakanni,Lekshmy Balakrishnan & Hendrik W. van Veen

Department of Pharmacology, University of Cambridge, Tennis Court Road,Cambridge CB2 1PD, UK.............................................................................................................................................................................

Multidrug resistance, by which cells become resistant to multipleunrelated pharmaceuticals, is due to the extrusion of drugs fromthe cell’s interior by active transporters such as the humanmultidrug resistance P-glycoprotein1. Two major classes of trans-porters mediate this extrusion2,3. Primary-active transporters aredependent on ATP hydrolysis, whereas secondary-active trans-

letters to nature

NATURE | VOL 426 | 18/25 DECEMBER 2003 | www.nature.com/nature866 © 2003 Nature Publishing Group