The TOM complex guides precursor proteins from the cell's cytosolic fluid into organelles called mitochondria. Biochemical analyses reveal the architecture of this complex and show how precursor proteins pass through its narrow pores.
Intracellular organelles called mitochondria are bounded by an inner and an outer membrane and house their own genome — two reminders that they arose from a bacterium that was engulfed by a host cell. As mitochondria evolved, most of their genome was transferred to the nucleus, and mitochondrial precursor proteins are now imported into mitochondria from the cell's cytosolic fluid. Writing in Science, Shiota et al.1 describe the molecular architecture of the main entry gate for these proteins, the TOM complex, in its native environment, achieving a milestone in the study of mitochondrial-import pathways.
The pathways that import mitochondrial precursor proteins are amazingly intricate2, involving some 40 proteins organized into 10 complexes called mitochondrial protein translocases. Together, the translocases ensure that more than 1,000 proteins, of diverse types, are correctly recognized, imported and sorted to the right place inside mitochondria. Much progress has been made in identifying the components of each translocase2, but atomic-resolution structures, which are needed for a deeper mechanistic understanding, are available for only a few soluble complexes and, with some translocases, for individual protein domains. The structural characterization of intact translocases is far from trivial, because they are complex, of low abundance, unstable and contain many membrane-spanning components.
The TOM complex recognizes and guides mitochondrial proteins across the organelle's outer membrane. Many experiments2 have demonstrated that the complex forms pores in the outer membrane that are each made up of a channel-forming β-barrel protein called Tom40 and six proteins containing single α-helical membrane-spanning segments — three receptors (Tom22, Tom20 and Tom70) that recognize mitochondrial precursor proteins for import and three subunits (Tom5, Tom6 and Tom7) that are functionally less well defined. Structural analyses have suggested that the TOM complex has two or three pores3,4. But although these studies made it relatively clear that Tom40 is the major pore-forming component, they were not conducted at high-enough resolution to identify the positions of the other subunits, nor to define whether precursor proteins traverse the membrane through the Tom40 β-barrels or through the spaces formed between the β-barrel pores.
Shiota et al. incorporated an unnatural amino acid into various sites in Tom40 in vivo. On activation by light5, this amino acid forms crosslinks with neighbouring proteins, revealing the positions of each subunit in the complex relative to Tom40. The authors overlaid these crosslinking data on a computational model of the structure of the Tom40 β-barrel, which was generated using previously obtained data6,7,8. On the basis of the observed crosslinks, the researchers proposed a three-pore model for the TOM complex, in which three Tom40 β-barrels (each consisting of 19 β-strands that span the outer membrane) are bridged by three Tom22 molecules, such that each Tom22 binds to two β-barrels. Tom5, Tom6 and Tom7 are positioned at the outer face of the β-barrel (Fig. 1). Importantly, Shiota and colleagues also analysed interactions between Tom40 and examples of two major types of precursor protein. They observed that, during import, only amino-acid residues facing the interior of the β-barrel are crosslinked to the precursors. These data strongly support the idea that precursor proteins cross the outer membrane through β-barrels.
The two types of precursor analysed — one containing positively charged targeting signals, one hydrophobic — seemed to take different routes through Tom40. Whether all precursor proteins are imported through the β-barrel, or whether some take alternative routes, remains to be clarified. It will be interesting to investigate the import routes of proteins that were previously suggested2 not to use Tom40. It will also now be possible to analyse the folding state of precursors as they pass through the narrow pore of the β-barrel.
The functional significance of the three pores in the complex remains unclear. Previous biochemical work9,10 indicates that the TOM complex undergoes structural rearrangement during protein import. It could be that all the pores are identical and simultaneously active, or that two inactive pores maintain the conformation of one active pore. It could also be that several pores are needed to attract the protein-synthesis machinery, thus ensuring that precursors are swiftly imported after their synthesis. It will be interesting to see whether molecular modification11, such as phosphorylation, of the TOM complex influences its multimeric state or its conformation.
Shiota and colleagues' model provides clues to a possible molecular mechanism underlying the structural rearrangement of the TOM complex. Tom22 contacts Tom40 where the first and last β-strand meet, an optimal site for Tom22 to regulate the conformation of the β-barrel. Notably, the first and last β-strands of Tom40 are parallel to one another. Lateral opening (in which these β-strands part, opening the barrel lengthways across the membrane) has been reported for β-barrels sealed by antiparallel strands12. Because the stability of parallel β-strands is lower than that of antiparallel strands, the conformational flexibility of Tom40 and thereby of the entire complex might be higher than previously assumed. This could reflect the role of the complex in transporting diverse types of precursor protein.
Finally, the authors' model suggests that the amino-terminal segment of Tom40 can traverse the β-barrel from the cytosolic side to the inter-membrane space. Perhaps this segment blocks the pore, retracting before or during protein import. Alternatively, it might have an active role in moving precursor proteins across the barrel. The loops connecting individual β-strands are also likely to be involved in protein movement, as in related bacterial transporters13.
Future experiments should address how the interplay between precursor proteins and TOM-complex receptors influences the architecture of the entire complex. Does it undergo conformational changes as precursors are moved? Evidence suggests that some precursor proteins can also exit mitochondria through the TOM complex14, raising questions about how directionality is achieved. Because different types of precursor protein preferentially bind to different receptors, it is likely that the TOM complex responds to each type in a different manner.
The TOM complex cooperates with different downstream translocases to ensure that precursor proteins reach the correct location. Perhaps binding to downstream translocases changes the conformation of the TOM complex to optimize the efficiency with which specific types of precursor protein are imported, even before they bind to the complex15. The new model will allow a directed analysis of the interplay between the TOM complex and downstream translocases, building on previous studies1,10,15.
Shiota and colleagues' pioneering model will certainly inspire many hypothesis-driven analyses. Moreover, it will provide impetus for further structural and functional studies of the molecular mechanisms that underlie mitochondrial-protein import.Footnote 1
Shiota, T. et al. Science 349, 1544–1548 (2015).
Neupert, W. J. Mol. Biol. 427, 1135–1158 (2015).
Künkele, K. P. et al. Cell 93, 1009–1019 (1998).
Model, K., Meisinger, C. & Kühlbrandt, W. J. Mol. Biol. 383, 1049–1057 (2008).
Liu, C. C. & Schultz, P. G. Annu. Rev. Biochem. 79, 413–444 (2010).
Ujwal, R. et al. Proc. Natl Acad. Sci. USA 105, 17742–17747 (2008).
Qiu, J. et al. Cell 154, 596–608 (2013).
Lackey, S. W. K. et al. J. Biol. Chem. 289, 21640–21650 (2014).
Rapaport, D. et al. Mol. Cell. Biol. 18, 5256–5262 (1998).
Shiota, T., Mabuchi, H., Tanaka-Yamano, S., Yamano, K. & Endo, T. Proc. Natl Acad. Sci. USA 108, 15179–15183 (2011).
Schmidt, O. et al. Cell 144, 227–239 (2011).
Noinaj, N., Kuszak, A. J., Balusek, C., Gumbart, J. C. & Buchanan, S. K. Structure 22, 1055–1062 (2014).
Baud, C. et al. Nature Commun. 5, 5271 (2014).
Bragoszewski, P. et al. Proc. Natl Acad. Sci. USA 112, 7713–7718 (2015).
Waegemann, K., Popov-Čeleketić, D., Neupert, W., Azem, A. & Mokranjac, D. J. Mol. Biol. 427, 1075–1084 (2015).
About this article
Biochimica et Biophysica Acta (BBA) - Biomembranes (2017)
Annual Review of Biochemistry (2017)