- Email: [email protected]
Membrane Insertases Are Present in All Three Domains of Life
Structure
Previews Membrane Insertases Are Present in All Three Domains of Life Ross E. Dalbey1,* and Andreas Kuhn2,* 1Department
of Chemistry and Biochemistry, The Ohio State University, Columbus, OH 43210, USA of Microbiology and Molecular Biology, University of Hohenheim, 70599 Stuttgart, Germany *Correspondence: [email protected] (R.E.D.), [email protected] (A.K.) http://dx.doi.org/10.1016/j.str.2015.08.002 2Institute
In this issue of Structure, Borowska et al. (2015) report the crystal structure and provide experimental evidence on an archaeal membrane insertase, the DUF106 protein from Methanocaldococcus jannaschii, demonstrating that YidC/Oxa1/Alb3-like insertases exist in the archaeal plasma membrane. Ubiquitous in all cells are the integral membrane proteins that are responsible for many essential cellular functions and are important as ion channels, transporters and ATPases. In addition, they play critical roles in signal transduction as cell surface receptors. These membrane proteins are critical for maintaining the pH and the ion composition in the cytoplasm and other aspects of cellular homeostasis. In the three domains of life, most of these proteins are targeted to the membrane by the signal recognition particle (SRP) and co-translationally inserted into the membrane by the Sec61 translocon (SecYEG and YidC in bacteria). The Sec machinery forms a translocation channel and is composed of the three subunits Sec61a, Sec61b, and Sec61g. The Sec translocation channel can open laterally and release partially translocated proteins into the membrane bilayer for their integration. In the case of tail anchored proteins, the Get machinery is employed for membrane insertion into the ER in eukaryotes and operates by an unknown mechanism (Wang et al., 2014). Another widely used system for membrane insertion in mitochondria (the inner membrane), chloroplasts (the thylakoid membrane), and bacteria (the cytoplasmic membrane) is the YidC/Oxa1/ Alb3 pathway (Dalbey et al., 2014). The membrane insertases of the YidC/Oxa-1/ Alb3 family were first detected in mitochondria as an essential component for the assembly of the cytochrome oxidase complex (Bauer et al., 1994). Soon after, homologs in bacteria and chloroplasts were detected and found to support membrane insertion of proteins into the cytoplasmic membrane and the thylakoid bilayer (Samuelson et al., 2000; Moore et al., 2000). Besides membrane insertion, YidC family members are involved in lateral
membrane integration, as well as in folding and assembly of membrane proteins. YidC/Alb3 can operate independently or cooperatively with the Sec machinery in bacteria and chloroplasts (Scotti et al., 2000; Klostermann et al., 2002). Despite an extensive search of archaeal genomes, only weak homology was found for some members of euryarchaeota, but no experimental data provided proof that they function as a membrane insertase. Borowska et al. (2015) now show that DUF106 (Mj0480) from M. jannaschii not only possesses structural similarities to YidC but also binds to nascent chains of an E. coli substrate as occurs for the bacterial YidC. These results now establish that members of the YidC/Oxa-1/Alb3 family are present in all three domains of life. Whereas the E. coli YidC is a 60 kDa protein and spans the membrane six times, Mj0480 has a size of only 23 kDa and three predicted transmembrane segments (TMS). These three TMS form an astonishingly similar structure to parts of the YidC hydrophilic groove of the Bacillus halodurans YidC2 and the E. coli YidC (Kumazaki et al., 2014a, 2014b). All of these structures are closed to the periplasmic/extracellular side (Figure 1). In addition, the first cytoplasmic loop of Mj0480 is predicted to form a coiled-coil structure as is found for the C1 loop of YidC. Further, an amphiphilic helix is present at the periplasmic/ extracellular side in both proteins. In summary, Mj0480 resembles a minimal version of the bacterial YidC. Borowska et al. (2015) make an interesting structural comparison of their minimal three TM YidC module found within the archaeal YidC with the structural features of the insertases of the eukaryotic Get machinery. Both Get1/ Get2, which function as the insertase for
the tail anchored proteins (Wang et al., 2014), have three TM segments; whether the Get insertases also have a YidC-like hydrophilic cavity that is open to the cytoplasm and to the lipid bilayer will have to wait until their structures are solved. The intriguing structural features of the membrane insertases give us some hints as to how they operate mechanistically. Distinct from the mechanism of the Sec translocases, which have a transmembrane channel that is controlled by a hydrophobic pore ring and a periplasmic lid, YidC does not have such a channel (Van den Berg et al., 2004); rather, it provides a platform to bind small proteins in an amphiphilic groove that is located in the inner leaflet of the membrane bilayer. By binding a substrate protein into this groove, the translocation energy that is required to cross the membrane is reduced (Dalbey and Kuhn, 2014). Using photocrosslinking techniques, Borowska et al. (2015) found that, indeed, residues inside the groove of Mj0480 interact with residues of the C-subunit of the FoF1 ATPase from E. coli that was used as substrate. Another feature of the archaeal YidC, which is conserved in all family members, is that it has a number of methionine residues on the cytoplasmic membrane surface region of the protein. The authors suggest that these methionines might play a role in binding the inserting transmembrane segments of the substrates. Previously, methionine residues were discovered in the cleft region of SRP, and these methionine bristles were proposed to bind to the signal peptide of exported proteins (Bernstein et al., 1989). When comparing the now available structures of different YidC family members, it is fascinating to see how the basic structural element is conserved
Structure 23, September 1, 2015 ª2015 Elsevier Ltd All rights reserved 1559
Structure
Previews
Figure 1. Membrane Insertases from M. jannaschii, B. halodurans, and E. coli Membrane insertases from the archaeon M. jannaschii, the Gram-positive B. halodurans, and the Gram-negative E. coli show an increasing complexity. Note the structure of the DUF106 protein is a model of a monomer based on the domain-swapped dimer that was seen in the crystal (Borowska et al., 2015). The coiled-coil region of the M. jannaschii YidC protein, as well as the first TM segment and the N-terminal part of the P1 domain of the E. coli YidC are not seen in the crystal structure because they were disordered in the crystal.
and that additional features were added on or were lost during evolution (Figure 1). Certainly, the archaeal version of YidC describes the minimal core. In Gram-positive bacteria and in the organellar homologs, two additional transmembrane segments exist that most likely support the substrate binding and stability of the protein (Luirink et al., 2001). A large periplasmic domain is present that supports the interaction with the Sec translocase, but could have an additional, yet unknown function in Gram-negative bacteria. In summary, the new member in the YidC/Oxa-1/Alb3 family has a simpler core motif forming the insertase domain than that used by other members of the family. A subsequent important step in characterizing its role in M. jannaschii is to find an endogenous substrate that justifies its presence and to show that the reconstituted insertase can insert substrates independently. Moreover, it will be interesting to discover whether the archaeal protein
has the capability to cooperate with the Sec translocase as is the case for the bacterial and chloroplast members. ACKNOWLEDGMENTS We thank Dirk Spann and Yuanyuan Chen for help preparing Figure 1. This work was supported by a National Science Foundation grant MCB1052033 to R.E.D. and a DFG grant Ku749/6 to A.K.
Klostermann, E., Droste Gen Helling, I., Carde, J.P., and Schu¨nemann, D. (2002). Biochem. J. 368, 777–781. Kumazaki, K., Chiba, S., Takemoto, M., Furukawa, A., Nishiyama, K., Sugano, Y., Mori, T., Dohmae, N., Hirata, K., Nakada-Nakura, Y., et al. (2014a). Nature 509, 516–520. Kumazaki, K., Kishimoto, T., Furukawa, A., Mori, H., Tanaka, Y., Dohmae, N., Ishitani, R., Tsukazaki, T., and Nureki, O. (2014b). Sci. Rep. 4, 7299. Luirink, J., Samuelsson, T., and de Gier, J.W. (2001). FEBS Lett. 501, 1–5.
REFERENCES Bauer, M., Behrens, M., Esser, K., Michaelis, G., and Pratje, E. (1994). Mol. Gen. Genet. 245, 272–278. Bernstein, H.D., Poritz, M.A., Strub, K., Hoben, P.J., Brenner, S., and Walter, P. (1989). Nature 340, 482–486. Borowska, M.T., Dominik, P.K., Anghel, S.A., Kossiakoff, A.A., and Keenan, R.J. (2015). Structure 23, this issue, 1715–1724.
Moore, M., Harrison, M.S., Peterson, E.C., and Henry, R. (2000). J. Biol. Chem. 275, 1529–1532. Samuelson, J.C., Chen, M., Jiang, F., Mo¨ller, I., Wiedmann, M., Kuhn, A., Phillips, G.J., and Dalbey, R.E. (2000). Nature 406, 637–641. Scotti, P.A., Urbanus, M.L., Brunner, J., de Gier, J.W., von Heijne, G., van der Does, C., Driessen, A.J., Oudega, B., and Luirink, J. (2000). EMBO J. 19, 542–549.
Dalbey, R.E., and Kuhn, A. (2014). Nat. Struct. Mol. Biol. 21, 435–436.
Van den Berg, B., Clemons, W.M., Jr., Collinson, I., Modis, Y., Hartmann, E., Harrison, S.C., and Rapoport, T.A. (2004). Nature 427, 36–44.
Dalbey, R.E., Kuhn, A., Zhu, L., and Kiefer, D. (2014). Biochim. Biophys. Acta 1843, 1489–1496.
Wang, F., Chan, C., Weir, N.R., and Denic, V. (2014). Nature 512, 441–444.
1560 Structure 23, September 1, 2015 ª2015 Elsevier Ltd All rights reserved
Previews Membrane Insertases Are Present in All Three Domains of Life Ross E. Dalbey1,* and Andreas Kuhn2,* 1Department
of Chemistry and Biochemistry, The Ohio State University, Columbus, OH 43210, USA of Microbiology and Molecular Biology, University of Hohenheim, 70599 Stuttgart, Germany *Correspondence: [email protected] (R.E.D.), [email protected] (A.K.) http://dx.doi.org/10.1016/j.str.2015.08.002 2Institute
In this issue of Structure, Borowska et al. (2015) report the crystal structure and provide experimental evidence on an archaeal membrane insertase, the DUF106 protein from Methanocaldococcus jannaschii, demonstrating that YidC/Oxa1/Alb3-like insertases exist in the archaeal plasma membrane. Ubiquitous in all cells are the integral membrane proteins that are responsible for many essential cellular functions and are important as ion channels, transporters and ATPases. In addition, they play critical roles in signal transduction as cell surface receptors. These membrane proteins are critical for maintaining the pH and the ion composition in the cytoplasm and other aspects of cellular homeostasis. In the three domains of life, most of these proteins are targeted to the membrane by the signal recognition particle (SRP) and co-translationally inserted into the membrane by the Sec61 translocon (SecYEG and YidC in bacteria). The Sec machinery forms a translocation channel and is composed of the three subunits Sec61a, Sec61b, and Sec61g. The Sec translocation channel can open laterally and release partially translocated proteins into the membrane bilayer for their integration. In the case of tail anchored proteins, the Get machinery is employed for membrane insertion into the ER in eukaryotes and operates by an unknown mechanism (Wang et al., 2014). Another widely used system for membrane insertion in mitochondria (the inner membrane), chloroplasts (the thylakoid membrane), and bacteria (the cytoplasmic membrane) is the YidC/Oxa1/ Alb3 pathway (Dalbey et al., 2014). The membrane insertases of the YidC/Oxa-1/ Alb3 family were first detected in mitochondria as an essential component for the assembly of the cytochrome oxidase complex (Bauer et al., 1994). Soon after, homologs in bacteria and chloroplasts were detected and found to support membrane insertion of proteins into the cytoplasmic membrane and the thylakoid bilayer (Samuelson et al., 2000; Moore et al., 2000). Besides membrane insertion, YidC family members are involved in lateral
membrane integration, as well as in folding and assembly of membrane proteins. YidC/Alb3 can operate independently or cooperatively with the Sec machinery in bacteria and chloroplasts (Scotti et al., 2000; Klostermann et al., 2002). Despite an extensive search of archaeal genomes, only weak homology was found for some members of euryarchaeota, but no experimental data provided proof that they function as a membrane insertase. Borowska et al. (2015) now show that DUF106 (Mj0480) from M. jannaschii not only possesses structural similarities to YidC but also binds to nascent chains of an E. coli substrate as occurs for the bacterial YidC. These results now establish that members of the YidC/Oxa-1/Alb3 family are present in all three domains of life. Whereas the E. coli YidC is a 60 kDa protein and spans the membrane six times, Mj0480 has a size of only 23 kDa and three predicted transmembrane segments (TMS). These three TMS form an astonishingly similar structure to parts of the YidC hydrophilic groove of the Bacillus halodurans YidC2 and the E. coli YidC (Kumazaki et al., 2014a, 2014b). All of these structures are closed to the periplasmic/extracellular side (Figure 1). In addition, the first cytoplasmic loop of Mj0480 is predicted to form a coiled-coil structure as is found for the C1 loop of YidC. Further, an amphiphilic helix is present at the periplasmic/ extracellular side in both proteins. In summary, Mj0480 resembles a minimal version of the bacterial YidC. Borowska et al. (2015) make an interesting structural comparison of their minimal three TM YidC module found within the archaeal YidC with the structural features of the insertases of the eukaryotic Get machinery. Both Get1/ Get2, which function as the insertase for
the tail anchored proteins (Wang et al., 2014), have three TM segments; whether the Get insertases also have a YidC-like hydrophilic cavity that is open to the cytoplasm and to the lipid bilayer will have to wait until their structures are solved. The intriguing structural features of the membrane insertases give us some hints as to how they operate mechanistically. Distinct from the mechanism of the Sec translocases, which have a transmembrane channel that is controlled by a hydrophobic pore ring and a periplasmic lid, YidC does not have such a channel (Van den Berg et al., 2004); rather, it provides a platform to bind small proteins in an amphiphilic groove that is located in the inner leaflet of the membrane bilayer. By binding a substrate protein into this groove, the translocation energy that is required to cross the membrane is reduced (Dalbey and Kuhn, 2014). Using photocrosslinking techniques, Borowska et al. (2015) found that, indeed, residues inside the groove of Mj0480 interact with residues of the C-subunit of the FoF1 ATPase from E. coli that was used as substrate. Another feature of the archaeal YidC, which is conserved in all family members, is that it has a number of methionine residues on the cytoplasmic membrane surface region of the protein. The authors suggest that these methionines might play a role in binding the inserting transmembrane segments of the substrates. Previously, methionine residues were discovered in the cleft region of SRP, and these methionine bristles were proposed to bind to the signal peptide of exported proteins (Bernstein et al., 1989). When comparing the now available structures of different YidC family members, it is fascinating to see how the basic structural element is conserved
Structure 23, September 1, 2015 ª2015 Elsevier Ltd All rights reserved 1559
Structure
Previews
Figure 1. Membrane Insertases from M. jannaschii, B. halodurans, and E. coli Membrane insertases from the archaeon M. jannaschii, the Gram-positive B. halodurans, and the Gram-negative E. coli show an increasing complexity. Note the structure of the DUF106 protein is a model of a monomer based on the domain-swapped dimer that was seen in the crystal (Borowska et al., 2015). The coiled-coil region of the M. jannaschii YidC protein, as well as the first TM segment and the N-terminal part of the P1 domain of the E. coli YidC are not seen in the crystal structure because they were disordered in the crystal.
and that additional features were added on or were lost during evolution (Figure 1). Certainly, the archaeal version of YidC describes the minimal core. In Gram-positive bacteria and in the organellar homologs, two additional transmembrane segments exist that most likely support the substrate binding and stability of the protein (Luirink et al., 2001). A large periplasmic domain is present that supports the interaction with the Sec translocase, but could have an additional, yet unknown function in Gram-negative bacteria. In summary, the new member in the YidC/Oxa-1/Alb3 family has a simpler core motif forming the insertase domain than that used by other members of the family. A subsequent important step in characterizing its role in M. jannaschii is to find an endogenous substrate that justifies its presence and to show that the reconstituted insertase can insert substrates independently. Moreover, it will be interesting to discover whether the archaeal protein
has the capability to cooperate with the Sec translocase as is the case for the bacterial and chloroplast members. ACKNOWLEDGMENTS We thank Dirk Spann and Yuanyuan Chen for help preparing Figure 1. This work was supported by a National Science Foundation grant MCB1052033 to R.E.D. and a DFG grant Ku749/6 to A.K.
Klostermann, E., Droste Gen Helling, I., Carde, J.P., and Schu¨nemann, D. (2002). Biochem. J. 368, 777–781. Kumazaki, K., Chiba, S., Takemoto, M., Furukawa, A., Nishiyama, K., Sugano, Y., Mori, T., Dohmae, N., Hirata, K., Nakada-Nakura, Y., et al. (2014a). Nature 509, 516–520. Kumazaki, K., Kishimoto, T., Furukawa, A., Mori, H., Tanaka, Y., Dohmae, N., Ishitani, R., Tsukazaki, T., and Nureki, O. (2014b). Sci. Rep. 4, 7299. Luirink, J., Samuelsson, T., and de Gier, J.W. (2001). FEBS Lett. 501, 1–5.
REFERENCES Bauer, M., Behrens, M., Esser, K., Michaelis, G., and Pratje, E. (1994). Mol. Gen. Genet. 245, 272–278. Bernstein, H.D., Poritz, M.A., Strub, K., Hoben, P.J., Brenner, S., and Walter, P. (1989). Nature 340, 482–486. Borowska, M.T., Dominik, P.K., Anghel, S.A., Kossiakoff, A.A., and Keenan, R.J. (2015). Structure 23, this issue, 1715–1724.
Moore, M., Harrison, M.S., Peterson, E.C., and Henry, R. (2000). J. Biol. Chem. 275, 1529–1532. Samuelson, J.C., Chen, M., Jiang, F., Mo¨ller, I., Wiedmann, M., Kuhn, A., Phillips, G.J., and Dalbey, R.E. (2000). Nature 406, 637–641. Scotti, P.A., Urbanus, M.L., Brunner, J., de Gier, J.W., von Heijne, G., van der Does, C., Driessen, A.J., Oudega, B., and Luirink, J. (2000). EMBO J. 19, 542–549.
Dalbey, R.E., and Kuhn, A. (2014). Nat. Struct. Mol. Biol. 21, 435–436.
Van den Berg, B., Clemons, W.M., Jr., Collinson, I., Modis, Y., Hartmann, E., Harrison, S.C., and Rapoport, T.A. (2004). Nature 427, 36–44.
Dalbey, R.E., Kuhn, A., Zhu, L., and Kiefer, D. (2014). Biochim. Biophys. Acta 1843, 1489–1496.
Wang, F., Chan, C., Weir, N.R., and Denic, V. (2014). Nature 512, 441–444.
1560 Structure 23, September 1, 2015 ª2015 Elsevier Ltd All rights reserved