Nature Structural Biology
9, 81 - 83 (2002)
doi:10.1038/nsb0202-81
A GTPase gate for protein import into chloroplastsFelix Kessler1
& Danny J. Schnell21 Felix Kessler is in the Institute of Plant Sciences, Plant Physiology and Biochemistry Group, ETH Zürich Universitätstrasse 2, 8092 Zürich, Switzerland. 2 Danny J. Schnell is in the Department of Biochemistry and Molecular Biology, University of Massachusetts, Amherst, Massachusetts 01003, USA.
Correspondence should be addressed to Felix Kessler felix.kessler@ipw.biol.ethz.ch or Danny J. Schnell dschnell@biochem.umass.eduProtein import into chloroplasts is regulated by the binding and hydrolysis of GTP at two homologous GTPases, Toc34 and Toc159. The crystal structure of the Toc34 GTP-binding domain suggests that GTP-regulated dimerization of the Toc GTPase domains controls the targeting and translocation of preproteins at the chloroplast envelope.The intracellular trafficking of newly synthesized polypeptides is governed by a set of organelle-specific soluble and/or membrane-bound receptor systems that recognize targeting signals within passenger proteins and direct them to protein translocation machines (translocons) at the boundary membrane of the appropriate organelle. Although the major signal receptors and translocon components for most organelles have been identified, the communication mechanisms assuring the fidelity of targeting within these systems remain largely mysterious.
Chloroplast biogenesis requires the import of hundreds of nuclear-encoded preproteins across the double membrane envelope from the cytosol. These preproteins are targeted to the organelle by cleavable N-terminal transit sequences. A multimeric translocon at the outer membrane (Toc complex) functions in the recognition of the transit sequence. The Toc complex associates with the translocon at the inner membrane (Tic complex) to provide a direct conduit for preproteins from the cytosol to the internal stromal compartment1,
2,
3 (Fig. 1). Two homologous GTPases, Toc34 and Toc159, mediate the recognition of transit sequences at the chloroplast surface4,
5 and regulate the presentation of preproteins to the translocation channel6,
7. On page 95 of this issue of Nature Structural Biology, Chwan-Deng Hsiao and coworkers8 offer a tantalizing glimpse into the role of GTP in regulating this novel class of GTPases by providing the crystal structure of the GTP binding domain of Toc34. The structure indicates that GTP binding and hydrolysis at Toc34 are regulated by novel homodimeric interactions. These data suggest how GTP-regulated dimerization may regulate access of preproteins to the protein translocation machinery of the chloroplast envelope.
 | | |
Toc complex
The core of the Toc complex consists of Toc159 and Toc34 (refs 9,
10,
11) in association with Toc75, the major component of the translocation channel12,
13,
14. The G-domains of Toc159 and Toc34 are homologous and define a unique subgroup of the GTPase superfamily. Toc159 exists in both membrane-bound and cytosolic forms15 and is the major receptor for transit sequences of newly synthesized preproteins4. Toc34 is an integral membrane protein whose GTPase activity is required for translocation of bound preproteins across the outer envelope membrane7. On the basis of these observations, Toc159 is proposed to target preproteins to the outer envelope membrane and Toc34 is proposed to act as a GTP-regulated gate to the translocation channel. Therefore, the structure of the G-domain of Toc34 provides an entry point to defining the role of GTP in regulating the activities of both Toc GTPases.
Novel Toc34 G-domain structure
Sun et al.8 determined the structure of the GDP-bound form of the 28 kDa Toc34 cytosolic G-domain. The construct lacks only the 52-amino acid C-terminal membrane anchor. The first intriguing feature of the structure is the unique arrangement of motifs that define the GTP/GDP binding pocket. The overall fold resembles that of the canonical GTP-binding protein H-ras-p21 (Ras)16. However, unlike Ras, six insertions considerably enlarge the structure of the Toc34 G-domain. With one exception, all of these insertions are of relevance for dimerization and GTPase activity.
Five motifs (G1−G5) within Ras are involved in GTP binding and hydrolysis. With the exception of G1 (P loop), the corresponding motifs either do not exist in Toc34 or diverge significantly from Ras−GDP, suggesting that Toc34 may have evolved novel mechanisms of GTP binding and hydrolysis. For example, novel G4 and G5 motifs within Toc34 coordinate GTP via the guanine ring. Although Toc34 has an apparent G4 motif (216NKxD219) similar to that of Ras (117NKxD120), it does not appear to be involved in GDP binding. Rather, Thr 162 and His 163 of Toc34 form contacts with the ribose and guanine ring of GDP in a manner analogous to the canonical G4 motif. These two residues may therefore form the functional G4 motif of Toc34. Asp 119 of the Ras G5 motif, which forms two hydrogen bonds to the guanine ring, is absent from Toc34. Instead, Glu 210 and Asn 211 make strong and weak interactions with the ring, thereby defining a unique G5 motif.
Additionally, Toc34 does not have a classic G2 motif as defined by the invariant Thr 35 of Ras; rather, alternative residues of Toc34 (Ser 72 or Ser 68) are predicted to interact with the  -phosphate of GTP. Finally, the G3 motif that contributes a catalytic glutamine (Gln 61) to the active site of Ras contains a leucine (Leu 97) in the corresponding position in Toc34, suggesting a novel mechanism of hydrolysis. These observations not only define Toc34 as a model for the unique Toc GTPase family, but also will allow in depth functional analysis of nucleotide binding and hydrolysis in the import reaction through effective mutagenesis.
A GTPase switch?
The most exciting implications of the structure relate to the activity of Toc34 within the Toc complex. A particularly striking feature is that the G-domain forms a dimer with residues of one monomer contributing to the formation of the GDP binding pocket of the other monomer. Remarkably, the dimer interface resembles the interface between small GTPases and their GTPase-activating proteins (GAPs)17 (Fig. 2). Such GAPs contribute a catalytic 'arginine finger' to the active site of small GTPases upon binding. Arg 133 of Toc34 is deeply inserted into the GDP binding region of the other monomer, suggesting that it may function as an arginine finger to stimulate hydrolysis (Fig. 2). The authors further show that mutation of a second arginine, Arg 128, which forms critical hydrogen bonds at the dimer interface, disrupts dimer formation in solution and greatly reduces GTPase activity relative to wild type. These findings suggest a novel reciprocal activation of the Toc34 GTPase subunits in the homodimer.
| | |
Previous biochemical analyses indicate that GTP binding alters the interaction of Toc34 with preproteins5,
18. In addition, GTP hydrolysis at Toc34 is required to initiate preprotein translocation through the Toc complex7. These two observations in conjunction with the structural data suggest a possible model in which GTP-regulated dimerization of Toc34 functions as part of the molecular switch controlling the commitment of bound preproteins to translocation. In the GDP-bound dimer of Toc34, Glu 73 of the G2 motif occupies the predicted position of the -phosphate of a bound GTP. Furthermore, there is no apparent entrance/exit site for nucleotide in the dimer. Thus, a significant conformational shift or perhaps dimer dissociation may be required for GDP/GTP exchange. As a consequence, the conformational state of Toc34 may be directly regulated by bound nucleotide. It is tempting to speculate that this shift coupled with subsequent dimer-activated GTP hydrolysis could constitute at least part of the GTP-dependent switch required for the initiation of membrane translocation.
One possible mechanism for coupling the GTP binding and hydrolytic events at Toc34 to other components of the Toc complex is also suggested by the crystal structure. The single cysteine at position 215 in Toc34 is located in the longest loop region in the structure. Cys 215 can be oxidatively crosslinked directly to other Toc components19, in particular Toc75, indicating close contact to the channel component. The newly assigned G5 is located on the longest loop as well, suggesting that this loop may transmit events at the GTP binding site to other translocon subunits. Conversely, the contacts at the Cys 215 loop provide a mechanism for other Toc components to participate in nucleotide exchange, possibly by acting as GTP/GDP exchange factors (GEFs).
The Toc34 G-domain structure suggests a second possible interaction with implications for preprotein targeting to the Toc complex. The Toc159 receptor contains a G-domain with significant sequence similarity to that of Toc34 (ref. 9). The most conserved regions between Toc34 and Toc159 lie in their GTP-binding motifs and the regions responsible for dimer formation. The putative arginine finger is conserved in all known Toc159 sequences8. This raises the possibility that Toc159 also forms homodimers with reciprocal GTPase activating properties. Another, more intriguing possibility is heterodimer formation between Toc34 and Toc159. This proposal is supported by the observation that the G-domain of the Arabidopsis ortholog of Toc34 competes for the insertion of the soluble form of Toc159 in the chloroplast outer membrane15. The new structural data therefore suggest the possibility that Toc34 may serve as a GTP-regulated docking site for the cytosolic form of the Toc159 receptor as it targets to the Toc complex in the outer membrane. In fact, Sun et al.8 report that the GDP-bound form of the Toc34 G-domain exists in solution at a dimer:monomer ratio of 1:4. Although this observation could be an artifact of measuring association external to the two dimensional lipid bilayer of the outer membrane, it provides additional indirect evidence that homodimerization is dynamic and opens the door for Toc159−Toc34 heterodimer formation.
The scenario of a preprotein targeting cycle involving cognate GTPases bears an uncanny resemblance to the targeting of the ribosome−nascent chain complexes (RNC) to the endoplasmic reticulum (ER) membrane by the signal recognition particle (SRP) and SRP receptor (SR)20. SRP associates with RNCs in the cytoplasm and the SRP−RNC is targeted to the ER protein translocon by docking at SR. Upon binding, SRP and the SR  -subunit act as reciprocal GTPase activators in a GTP-dependent reaction that ensures unidirectional transfer of the RNC to the translocation channel of the translocon21. The reciprocal GAP activities are proposed to involve conserved arginine finger domains similar to those observed in the Toc34 dimer. Although it remains to be demonstrated that the Toc34−Toc159 interaction represents a step in preprotein targeting to chloroplasts, the potential conservation of GTP-dependent mechanisms in these two targeting reactions is very compelling. As such, the structure of Toc34 provides a bounty of new testable hypotheses to account for the GTP-dependent protein targeting mechanisms at the outer chloroplast membrane and other intracellular protein targeting pathways.
REFERENCES
- Bauer, J., Hiltbrunner, A. & Kessler, F. Cell Mol. Life Sci. 58, 420−433 (2001). | PubMed | ISI | ChemPort |
- Keegstra, K. & Froehlich, J.E. Curr. Opin. Plant Biol. 2, 471−476 (1999). | Article | PubMed | ISI | ChemPort |
- Schnell, D.J. et al. Trends Cell Biol. 7, 303−304 (1997). | Article | ISI |
- Ma, Y., Kouranov, A., LaSala, S. & Schnell, D.J. J. Cell Biol. 134, 315−327 (1996). | Article | PubMed | ISI | ChemPort |
- Sveshnikova, N., Soll, J. & Schleiff, E. Proc. Natl. Acad. Sci. USA 97, 4973−4978 (2000). | Article | PubMed | ChemPort |
- Young, M.E., Keegstra, K. & Froehlich, J.E. Plant Physiol. 121, 237−244 (1999). | Article | PubMed | ISI | ChemPort |
- Chen, K., Chen, X. & Schnell, D.J. Plant Physiol. 122, 813−822 (2000). | Article | PubMed | ISI | ChemPort |
- Sun, Y.-J. et al. Nature Struct. Biol. 9, 95−100 (2002). | Article | PubMed | ISI | ChemPort |
- Kessler, F., Blobel, G., Patel, H. A. & Schnell, D. J. Science 266, 1035−1039 (1994). | PubMed | ISI | ChemPort |
- Seedorf, M., Waegemann, K. & Soll, J. Plant J. 7, 401−411 (1995). | Article | PubMed | ISI | ChemPort |
- Hirsch, S., Muckel, E., Heemeyer, F., von Heijne, G. & Soll, J. Science 266, 1989−1992 (1994). | PubMed | ISI | ChemPort |
- Schnell, D.J., Kessler, F. & Blobel, G. Science 266, 1007−1012 (1994). | PubMed | ISI | ChemPort |
- Hinnah, S. C. et al. EMBO J. 16, 7351−7360 (1997). | Article | PubMed | ChemPort |
- Tranel, P.J., Froehlich, J., Goyal, A. & Keegstra, K. EMBO J. 14, 2436−46 (1995). | PubMed | ISI | ChemPort |
- Hiltbrunner, A. et al. J. Cell Biol. 154, 309−316 (2001). | Article | PubMed | ISI | ChemPort |
- Pai, E.F. et al. Nature 341, 209−214 (1989). | Article | PubMed | ISI | ChemPort |
- Vetter, I.R. & Wittinghofer, A. Science 294, 1299−1304 (2001). | Article | PubMed | ISI | ChemPort |
- Kouranov, A. & Schnell, D.J. J. Cell Biol. 139, 1677−1685 (1997). | Article | PubMed | ISI | ChemPort |
- Seedorf, M. & Soll, J. FEBS Lett. 367, 19−22 (1995). | Article | PubMed | ISI | ChemPort |
- Keenan, R.J., Freymann, D.M., Stroud, R.M. & Walter, P. Annu. Rev. Biochem. 70, 755−775 (2001). | Article | PubMed | ISI | ChemPort |
- Powers, T. & Walter, P. Science 269, 1422−1424 (1995). | PubMed | ISI | ChemPort |
|
