- Email: [email protected]
Conformational variability of the GTPase domain of the signal recognition particle receptor FtsY
Journal of Structural Biology 153 (2006) 85–96 www.elsevier.com/locate/yjsbi
Conformational variability of the GTPase domain of the signal recognition particle receptor FtsY Talal Gariani a,1, Tore Samuelsson b, A. Elisabeth Sauer-Eriksson a,¤ a
b
Umeå Centre for Molecular Pathogenesis, Umeå University, SE-901 87 Umeå, Sweden Department of Medical Biochemistry, University of Göteborg, Medicinareg. 9A, SE-405 30 Göteborg, Sweden Received 3 August 2005; received in revised form 20 October 2005; accepted 24 October 2005 Available online 30 November 2005
Abstract The prokaryotic signal recognition particle Ffh and its receptor FtsY allow targeting of proteins into or across the plasma membrane. The targeting process is GTP dependent and the two proteins constitute a distinct GTPase family. The receptor FtsY is composed of A and NG domains where the NG’s GTPase domain plays a critical role in the targeting process. In this study, we describe two X-ray structures determined independently of each other of the NG domain of FtsY from Mycoplasma mycoides (MmFtsY). The two structures are markedly diVerent in three of the nucleotide-binding segments, GI (P-loop), GII, and GIII, making only one of the structures compatible with nucleotide binding. Interestingly, the two distinct conformations of the nucleotide-binding segments of MmFtsY are similar to the apo- and ADP-loaded forms of certain ATPases. The structure of the extended interface between the A and NG domains of MmFtsY provides new insights into the role of the A domain for phospholipid interaction. 2005 Elsevier Inc. All rights reserved. Keywords: Signal recognition particle; SRP; SRP receptor; FtsY; Mycoplasma mycoides
1. Introduction The signal recognition particle (SRP) is one of the key components of the cellular machinery for membrane targeting. Many of its structural and functional features have been highly conserved during evolution; the SRP and its receptor have been identiWed in all three kingdoms of life (for reviews see Doudna and Batey, 2004; Keenan et al., 2001; Luirink and Sinning, 2004; Wild et al., 2004). In eukaryotic cells, the SRP targets proteins into or across the membrane of the endoplasmic reticulum (ER). Eukaryotic SRP comprises one 7S RNA molecule (»300 nucleotides) and six protein components named SRP9, SRP14, SRP19, SRP54, SRP68, and SRP72 according to *
Corresponding author. Fax: +46 90 778007. E-mail address: [email protected] (A.E. Sauer-Eriksson). 1 Present address: SanoW-Aventis, Chemical Science, Centre de Recherche Paris, Friedel, B 122.2, 13 Quai Jules Guesde, 94403 Vitry-sur-Seine Cedex, France. 1047-8477/$ - see front matter 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.jsb.2005.10.003
their mass in kilodalton. The SRP pathway in prokaryotes is similar to that in eukaryotes. The major protein targets for bacterial SRP are integral membrane proteins (Luirink and Dobberstein, 1994; Ulbrandt et al., 1997). Eubacterial SRP has only two components: a 4.5S RNA and the Ffh protein. The Ffh protein is homologous to SRP54, which has a key role in recognizing the signal sequence of the nascent polypeptide chain as it emerges from the ribosomes at the peptide exit site (Halic et al., 2004; Pool et al., 2002; Rinke-Appel et al., 2002), and in binding to the SRP receptor (SR). The eukaryotic SR is a heterodimer made up of the GTPases SR and SR (Gilmore et al., 1982; Meyer et al., 1982; Rapiejko and Gilmore, 1997), whereas the eubacterial SR consists of only one protein, FtsY, homologous to SR (Luirink et al., 1994). Mutual GTP hydrolysis of SRP and its receptor regulate the SRP pathway such that SRP is released from the ribosome complex (Miller et al., 1993, 1994; Powers and Walter, 1995), and translocation is allowed (Connolly and Gilmore, 1989; Fulga et al., 2001).
86
T. Gariani et al. / Journal of Structural Biology 153 (2006) 85–96
Ffh and FtsY are evolutionarily related and share N and G (GTPase) domains, where the G domains have GTP-consensus elements GI–GV (Bourne et al., 1991). In addition, Ffh includes a methionine-rich M domain, which binds both to the signal sequence and to RNA (Lütcke et al., 1992; Zopf et al., 1993). FtsY includes the highly acidic A domain whose function might be to interact with the lipid membrane (de Leeuw et al., 2000; Zelazny et al., 1997). The high resolution structures of the NG domains of Ffh from Thermus aquaticus (TaqFfh) (Freymann et al., 1997) and FtsY from Escherichia coli (EcFtsY) (Montoya et al., 1997a) provided us with the Wrst detailed description of these domains at the molecular level. The domain structures of Ffh and FtsY are very similar. The G domain has a fold similar to that of the Ras-like GTPases, but with an - insertion box domain (IBD) unique to the SRPGTPase family (Freymann et al., 1997; Montoya et al., 1997a). The N domain forms a four-helix bundle that can be considered as an extension of the G domain. This N domain has a role in GTP hydrolysis and in membrane binding in association with the A domain (Millman and Andrews, 1999). The recent crystal structures of the Ffh– FtsY complex from T. aquaticus revealed a new mechanism in which the association of the two proteins activates reciprocal GTP hydrolysis (Egea et al., 2004; Focia et al., 2004). SRP components have been identiWed in other bacteria, including mycoplasmas (Samuelsson, 1992). Mycoplasmas are simple eubacteria and have a small genome that codes for a highly limited set of proteins. They are insensitive to penicillin, and cause diseases including arthritis and pneumonia. The fact that they have retained an SRP pathway possibly reXects the essential nature of this machinery. We have previously shown that the GTPase activity of mycoplasma FtsY is stimulated when it interacts with Ffh. In contrast to E. coli, however, this stimulation is possible without SRP RNA and the M domain of Ffh (Macao et al., 1997). This indicates that the interaction of the NG domains of Ffh and FtsY plays an important part in the activation mechanism, and the mycoplasma proteins constitute an interesting model system for SRP function. In this study, we present structures of the NG domain of FtsY in two diVerent forms from Mycoplasma mycoides (MmFtsY): an apo and a sulfate-loaded form. Whereas the sulfate-loaded form is similar to the previously determined structures of both Ffh and FtsY, large structural changes in three of the GTP-binding loops were observed in the apo form. This implies that the loops in FtsY possess a higher intrinsic mobility than previously believed. Interestingly, the two conformations of the nucleotide-binding segments of MmFtsY are similar to the apo- and ADP-loaded forms of certain ATPases. 2. Materials and methods 2.1. Structure determination The NG domain of MmFtsY was engineered and puriWed as previously described (Gariani and Sauer-Eriksson,
Table 1 Data collection and reWnement statistics F1 Quality of reXections used in reWnement Space group P21212 Unit cell parameters (Å) a D 68.74, b D 101.13 c D 42.53 Resolution range (Å) 20–1.90 (1.97–1.9) Number of observation 497,368 Number of unique 22,061 reXection Completeness (%) 95.7 (79.5) I/(I) 17.45 (4.78) 6.9 (32.1) Rsyma,b (%) I/(I) > 2 83.2 (51.5) ReWnement statistics Resolution ReXections work set ReXections test set Rcryst Rfreec Number of protein atoms Chain A Chain B Number of water molecules Rms deviations Bonds (Å) Angles (o) Average B factor (Å2) Chain A Chain B Sulfate molecule Water molecules
F2 R32 a D b D 148.44 c D 223.93 20–2.4 (2.46–2.40) 630,367 35,370 94.8 (87.3) 21.68 (2.49) 5.6 (60.9) 82.2 (49.4)
20–1.95 18,640 2062 20.7 26.4
20–2.4 33,620 1750 20.9 24.6
2405 — 252
2439 2453 139
0.017 1.62
0.018 1.67
26.0 — 39.0 34.5
52.1 53.7 — 45.2
a
Value in parentheses is in the high resolution bin. Rsym D hkli冷Ii(hkl)-具I(hkl)典冷/hkliIi(hkl). c Rcryst/Rfree D 储Fo冷 ¡ 冷Fc储/ 冷Fo冷. Rfree was calculated using 10% (model F1) and 5% (model F2) of data excluded from reWnement. b
2000). At the N-terminal region, sequence 92-KEKDKKV98 was substituted with the His-tag 92-HHHHHPM-98. Crystals were obtained with two diVerent crystallization precipitants: ammonium sulfate (model F1) and sodium citrate (model F2) (Gariani and Sauer-Eriksson, 2000). DiVraction data to 1.95 and 2.4 Å resolution were collected at 100 K on the F1 and F2 crystals, respectively, at beam line X-11 in DESY, Hamburg. The data were processed with DENZO (Otwinowski and Minor, 1997), and merged using the programs TRUNCATE and SCALA from the CCP4 suite (Collaborative Computational Project, 1994). A summary of the data collection statistics is given in Table 1. Although the sequential and structural similarities between the NG domains of EcFtsY and MmFtsY are quite high (34% identity for the whole molecule, 44% identity for the G domain), solving the structure by molecular replacement (MR) was not straightforward. The NG domain of EcFtsY (PDB accession code 1FTS, Montoya et al., 1997a) was used as a search model for CNS (Brünger et al., 1998). The model was progressively truncated to 680 atoms before the correct solution appeared among the top ten solutions. This corresponded
T. Gariani et al. / Journal of Structural Biology 153 (2006) 85–96
to 30% of the protein and included only a polyalanine model of secondary structural elements from the G domain. After initial rigid-body reWnement, the Rwork and Rfree for the correct solution decreased from 55.0 and 54.0% to 54.7 and 53.0%, respectively, whereas the incorrect solutions remained constant or slightly increased their R values. Only by comparing the quality of the electron-density maps could the correct solution be identiWed. Subsequent rounds of model building and reWnement using the programs O (Jones et al., 1991) and CNS (Brünger et al., 1998) reduced the Rwork and Rfree to 20.7 and 26.4%, respectively. The F2 structure was determined by MR methods using AMoRe (Navaza, 1994) with the entire F1 structure as search model. The structure was reWned using a similar procedure as for F1 to a Wnal Rwork and Rfree of 20.9% and 24.6%, respectively. Figures were made with the programs MOSLCRIPT (Kraulis, 1991), PYMOL (DeLano W.L., http://www.pymol.org), and MOLRAY–POVRAY (Harris and Jones, 2001). 2.2. Protein DataBank accession number Coordinates and structure factors have been deposited in the Protein Data Bank [accession codes 1ZU4 (model F1) and 1ZU5 (model F2)].
87
(Montoya et al., 1997a), TaqFfh and TaqFtsY (Egea et al., 2004; Focia et al., 2004; Freymann et al., 1997, 1999; Keenan et al., 1998; Padmanabhan and Freymann, 2001), and Acidianus ambivalens, AaFfh (Montoya et al., 2000) (Fig. 1A). The sequence identities between MmFtsY and EcFtsY, and between MmFtsY and TaqFtsY, are 34% and 35%, respectively (Fig. 1B). The N domain is made of a four-helix bundle (helices A2, B–D), which in MmFtsY comprises 78 residues at the N-terminal end (Met-102 to Asp-179). The G domain of MmFtsY is formed by residues Leu-196 to Asp-391, which are structurally homologous to other SRP GTPases and include the Wve conserved nucleotide-consensus elements required for GTP binding and hydrolysis: GI or P loop (Gly-203 to Ser-212), GII (Asp233 to Ala-239), GIII (Asp-289 to Gln-295), GIV (Thr-353 to Asp-356), and GV or ‘closing loop’ (Gly-379 to Glu382). Helix 7 (also named G5), comprising eleven residues at the C-terminal end (Asp-391 to Ser-401), seems more connected to the N domain than the G domain. We therefore refer to this helix as helix E. The –– motif (IBD, Ala-237 to Gln-281), which is unique for the SRP GTPases, is inserted between strands 2 and 4. At the Nterminal end of the N domain, helix A1 (Pro-97 to Met-102) is situated. This helix is not structurally deWned in the previously determined SRP GTPases. Structurally, we Wnd that the A1 helix seems more associated with the G domain than with the N domain.
3. Results 3.1. Quality of the F1 and F2 models The structures of the NG domain of MmFtsY were determined from two crystal forms. One sulfate ion was identiWed at the nucleotide-binding site in model F1, which therefore represents a sulfate-bound form of the protein. The F1 model includes 307 out of the 318 residues, which comprise the NG domain (Glu-99 to Gln-405), and two residues, Pro-97 and Met-98, which are part of the His-tag. Omissions from the model because of no visible electron density are: Wve residues of the His-tag (His-92 to His-96), the loop connecting the N and G domains (Thr-183 to Lys-186), and the last six residues of the C-terminal end (Gly-406 to Lys-411). Additional residues with weak or no density for their side chains include predominantly polar amino acids situated at surface-exposed regions of the protein, and the entire side chain of Phe-235, which is part of the GII-loop region. The F2 structure comprises two molecules, A and B, in the asymmetric unit. The two molecules are very similar with a root mean squared (rms) deviation of 0.4 Å for all C atoms. Residue Phe-235 in the GII loop has well deWned electron densities in both molecules of the F2 model. 3.2. General architecture of MmFtsY The NG domain of MmFtsY presented the same overall fold as the previously determined structures from EcFtsY
3.3. The F1 model of MmFtsY is similar to the previous structures of FtsY and Ffh The structure of the Wve nucleotide-binding segments of model F1 is similar to the previously published SRP GTPases, however, with a better structural Wt to the segments of FtsY than the Ffh structures. The GI loop (P-loop) comprises residues Gly-203 to Ser-212 and is situated on a Xexible loop bridging 1 and 1. The role of the P loop is to properly position the -phosphate of the GTP and to bind to the Mg2+ ion that is necessary for GTP hydrolysis. In the F1 model of MmFtsY, there is one sulfate ion situated at the GI loop that is not present in EcFtsY (Fig. 2). The sulfate ions form hydrogen bonds to Wve consecutive main chain nitrogen atoms of the switch I regions (residues Gly-206 to Thr-210), to the N 2 atom of Arg-236, and to two water molecules, W1 and W2. The position of the sulfate ion is almost superimposable onto the position of the -phosphate position of the GMP-PCP analogue in the TaqFtsY–TaqFfh complex structure (Egea et al., 2004; Focia et al., 2004). An extensive hydrogen-bonding network with the sulfate ion at a pivotal position connects the GI, GII, and GIII regions in the F1 model (Fig. 2). The GII region of small GTPases is involved in Mg2+ binding. The role of the GII loop in SRP GTPases is less well deWned; studies on TaqFfh in complex with GTP analogues showed no interaction between residues situated at the GII region to the bound Mg2+ ion (Egea et al., 2004; Focia et al., 2004; Freymann et al., 1999). Residues Phe-235
88
T. Gariani et al. / Journal of Structural Biology 153 (2006) 85–96
T. Gariani et al. / Journal of Structural Biology 153 (2006) 85–96
89
Fig. 1. (A) Ribbon trace of the F1 model of MmFtsY. The N domain, G domain, and IBD box are colored yellow, blue, and red, respectively. The -helix A1 is shown in turquoise and the sulfate ion in green. (B) Structural-based alignment of SRP GTPases from Mycoplasma mycoides (1ZU4), Thermus aquaticus (1OKK), Acidianus ambivalens (1J8M), and Thermotoga martima (1VMA). Sequences from TqFfh and AFfh are also included for comparison. Identical sequences are boxed in dark grey and the Wve conserved GTP-binding regions are underlined in pink. The LIQTDM and TSKGG motifs are also highlighted. The extra helical turn present only in the P1 motif of model F2 is highlighted in magenta. Numbering is based on the M. mycoides FtsY sequence. The Wrst and last amino acids observed in the X-ray structures are highlighted in yellow. The two Wrst amino acids observed in MmFtsY (marked with *) constitute proline and methionine in our construct (see materials and methods). Secondary structures are colored as in (A). Helix 7 (also named G5), comprising eleven residues of the C-terminal end (Asp-391 to Ser-401), is more coupled to the N domain than the G domain. We therefore refer to this helix as helix E and show it in yellow. (For interpretation of the references to colour in this Wgure legend, the reader is referred to the web version of this paper.)
to Ala-237 situated in the GII region of the F1 model are more mobile compared to the rest of the F1 structure. In particular, the side chain of Phe-235 displays high Xexibility as evidenced by the lack of visible electron density for it. Residue Thr-234, which interacts with the Mg2+ ion and the -phosphate of the GTP nucleotide in Ras p21 (Schlichting et al., 1990; Wittinghofer et al., 1991), is oriented away from the GI loop. Its O1 atom forms a direct hydrogen bond to the main-chain carbonyl oxygen of Ala-265 that is situated in the IBD. This same interaction is found in the EcFtsY and TaqFtsY structures, and seems to be speciWc for the FtsY proteins. Residue Asp-233 points to the active site, and makes a salt link with the invariant Arg-236. In addition, Asp-233 is in contact with the GI and GIII regions through water-bridged interactions, W1, with the sulfate ion and the side chain of Asp-289. In EcFtsY, a similar pattern can be seen where the corresponding residue Asp-230 interacts with Asn-202 and Lys-207 from the GI region, and via two water molecules, W131 and W122, to the Asp-282 from the GIII region. The GIII-consensus element is well ordered in the F1 model of MmFtsY unlike what is seen in EcFtsY. In addition to the aforementioned interactions involving Asp-289 and Thr-290, Gly-292 is linked to residue Asn-205 from the GI region via three hydrogen bonds. These additional hydrogen bonds might
explain the stability of this consensus element with respect to the EcFtsY. The GIV (Thr-353 to Asp-356) and GV (i.e., the ‘closing loop’ Gly-379 to Glu-382) consensus elements interact with the nucleotide-ribose ring and are well conserved in the SRP-GTPase family. The GIV region is connected to the GI region through a hydrogen bond from the O atom of Thr-353 to the main chain carbonyl oxygen of Thr-207 via a water molecule W135 that is also found in the other SRP GTPase structures. The side chain of Lys-215 forms hydrogen bonds to the main chain oxygens of Gly-381 and Lys383. The GV region forms a -hairpin loop that is stabilized by polar interactions and similar in conformation to the GV loop of EcFtsY, TaqFtsY, TaqFfh, and AaFfh. Contacts between the GIV and GV regions are mediated through a direct hydrogen bond between the main-chain carbonyl oxygen of Gly-379 and the main-chain nitrogen of Met-355. In addition, three water molecules W9, W14, and W234, link Lys-354 and Asp-356 to Gly-381. 3.4. The F2 model of MmFtsY reveals a large conformational change at the nucleotide-binding site At the nucleotide-binding site, three major structural changes not found in model F1 have occurred. The Wrst is
Fig. 2. (A) Ribbon trace of model F1 at the GTP-binding site. Motifs I, II, and III are shown in red, and the sulfur atom is shown in green. (B) Detailed view of part of the extensive hydrogen-bonding network formed at the sulfate-binding site. (For interpretation of the references to colour in this Wgure legend, the reader is referred to the web version of this paper.)
90
T. Gariani et al. / Journal of Structural Biology 153 (2006) 85–96
Fig. 3. (A) Ribbon trace of model F2 at the GTP-binding site. In motif I, residues Asn-205 to Gly-208 are involved in a novel helical turn. (B) Detailed view of part of the hydrogen-bonding network. The orientation and coloring of the structure is as in Fig. 2. (For interpretation of the references to colour in this Wgure legend, the reader is referred to the web version of this paper.)
situated at the GI element. In the F1 model, water molecule W2 is at the N-terminal end of helix 1 that immediately follows the GI-loop region. In the F2 model, however, this water is missing and Asn-205, Gly-206, and Thr-207 have changed their positions such that helix 1 is extended by one turn (Fig. 3). This erases the active site known to accommodate the - and -phosphates of the substrate and the sulfate ion present in model F1. In its new position the O1 atom of Thr-207 is hydrogen-bonded to the N2 atom of Gln-242, the latter residue being situated at the C-terminal end of the GII region. The O1 atom of Asn-205 forms hydrogen bonds to the main-chain nitrogen atoms of Thr207 and Gly-208. At this position, two N2 atoms of A- and B-Asn-205 are situated 2.8 Å from each other in the asymmetric unit. Gly-206 is situated right at the sulfate-binding site in the F1 model. Due to the structural changes at the GI region, the two direct hydrogen bonds formed between Gly-292 and Asn205 are missing in the F2 model. Instead, a novel water molecule V3 in monomer A (U3 in monomer B) is introduced and connects the side chains of Lys-209, Asp-233, Arg-236, and Gly-292. The GI and GIII regions are situated further apart. For example, the distance between the C atoms of Gly-292 and Asn-207 is extended from 5.3 Å in the F1 model to 11.3 Å in the F2 model. This shift of the switch II region (residues Asp-289 to Arg-293) and the movement of helix 4 (residues Asn-296 to Lys-309) is reminiscent of the switch II conformation changes observed in small GTPases like EFTu, transducin, and Ran (Kjeldgaard and Nyborg, 1992; ScheVzek et al., 1995; Sprang, 1997). The third structural change is found at the GII region. As previously discussed, the residue Phe-235 is not stable in the sulfate-loaded form of MmFtsY. In the apo form a half turn of helix 2, which includes residues Arg-236 to Gly238, is lost. Residue Phe-235 is well deWned in the electron density and the side chain of Arg-236 is shifted by 2.5 Å towards the GI loop where it is situated at the position of
Asn-205 in the F1 model. Furthermore, hydrogen bonds between Arg-236 and Asp-233, and between Arg-236 and Val-204, lock the position of their side chains. In conclusion, the F2 model represents a substantial conformational change as compared to the F1 structure and this state of the protein seems to be incompatible with nucleotide binding. 3.5. The interface between the N and G domains The -helix bundle (helices A–D) is well deWned in the electron density of MmFtsY. Helix A is well-ordered, extended, and kinked in MmFtsY (Fig. 4), whereas it is absent in TaqFtsY and distorted in EcFtsY. The kink divides helix A into two sub-helices. Helix A1, residues Pro-97 to Met-102, is closely associated with the G domain and comprises what we believe is the C-terminal part of the A domain, whereas helix A2, residues Met-102 to Lys-118, is part of the N domain. Helix A1 packs against helix E from the N domain and two loops of the G domain located between strand 6 and helix E, and between strand 5 and helix 5. The N and G domains from the previously known structures of Ffh and FtsY are oriented diVerently with respect to each other. This Xexibility is intrinsic and the relative positions of the domains are regulated by the interaction of two highly conserved motifs. The Wrst of these, 135LIQTDM140, is homologous to the ‘ALLEADV’ motif in TaqFfh situated within the N domain, whereas the second motif, 358TSKGG362, is homologous to the ‘DARGG’ motif situated within the G domain (see Fig. 1B) (Egea et al., 2004; Focia et al., 2004). Movement of the N domain with respect to the G domain occurs during the formation of the Ffh–FtsY complex, and repositions several conserved hydrophobic residues in these motifs as well as residues situated in motif IV (Egea et al., 2004; Focia et al., 2004). An extensive interface between the N and G domains arises in MmFtsY due to the ordered structure of helices A1
T. Gariani et al. / Journal of Structural Biology 153 (2006) 85–96
91
Fig. 4. The Wgure visualizes the quality of the 2冷Fo冷¡冷Fc冷 electron-density map over helices A1 and A2. Residues are colored as in Fig. 1A. (For interpretation of the references to colour in this Wgure legend, the reader is referred to the web version of this paper.)
Fig. 5. Interactions formed between the N and G domains of MmFtsY due to the ordered structure of helix A. A more detailed view of part of the interactions is shown in the right panel. The contact interface made of residues from helices A1, A2, and E. The A1-helix is a novel structure so far only found in MmFtsY. This helix forms a water-mediated contact to the GIV domain via two water molecules, W8 and W201, linking Ala-101, Lys-104, and Ser-105 to the Met-355 and Asp-356 residues.
and A2. This interface was not observed in the previous structures as they are lacking helix A1. Two water molecules W8 and W201 bridge by hydrogen bonds residues Ala-201 to Ser-205 located in the short linker region connecting helices A1 and A2, to the main chain atoms of Met-355, Asp-356, and Thr-358 located in the GIV region (Fig. 5). Furthermore, Met-98, which is a valine in the native structure, is buried and
makes hydrophobic interactions with Ile-378, Val-380, Leu388, and Tyr-395. A second interface, which was also present in other structurally characterized SRP GTPases, consists of residues Ala-328, Gly-331, Ser-359, Lys-360, and Gly-362. These come in contact with residues Thr-138 and Asp-139, situated at the BC loop, either directly or via water mediated networks. Notably, the extensive interactions formed
92
T. Gariani et al. / Journal of Structural Biology 153 (2006) 85–96
between the N and G domains in MmFtsY make their overall conformation more similar to the one of FtsY present in the TaqFtsY–Ffh complex, than to the conformation of EcFtsY in its free form. 4. Discussion 4.1. A new apo state of FtsY GTP hydrolysis, catalyzed by the Ffh and FtsY proteins, is highly essential for the targeting process. Structures of nucleotide-loaded Ffh proteins alone (Freymann et al., 1999; Padmanabhan and Freymann, 2001), and nucleotideloaded Ffh–FtsY complexes (Egea et al., 2004; Focia et al., 2004) have allowed a better understanding of the hydrolysis step. Since the NG domains of Ffh and FtsY are very similar, one might suspect that the nucleotide uptake and hydrolysis would also be similar for the two proteins. However, biochemical studies show that FtsY, unlike Ffh, has poor speciWcity for GTP in solution (Shan and Walter, 2003). It has been proposed that speciWc binding of GTP to FtsY is correlated to two conformational states of the protein, referred to as the ‘open’ and ‘closed’ states (Shan and Walter, 2003). In the open state, structural Xexibility prevents proper nucleotide binding, which would explain the poor aYnity of FtsY for GTP. In the closed state, local structural rearrangements occur that trap the GTP nucleotide and allow Ffh-complex formation. We present here structural evidence for a new FtsY state that is present only when the protein is in its apo form. The new spatial arrangements seen in model F2 obliterate the nucleotide-binding pocket by removing the potential binding sites of the phosphate moiety of the nucleotide. The Ploop reshuZing is the most drastic diVerence between the F1 and F2 models. It is accompanied by a shift of the GIIIconsensus element and stabilization of the GII region. This new state could be likened to an ‘inert’ state, where the FtsY protein is prevented from binding to any nucleotide. The existence of such a nucleotide-free state, as a unique conformer, is not consistent with the basal GTP hydrolysis activity (Moser et al., 1997; Shan and Walter, 2003). FtsY might therefore shuttle between the ‘inert’ and ‘open’ states. This could explain, in part, the low aYnity and hydrolysis rate for its cognate substrate. The changes seen at the GII-consensus element, which result in its destabilization when shifting from the ‘inert’ state (F2 model) to an ‘open’ state (F1 model), are targeted in particular to residues Phe-235 and Arg-236. The same pattern is observed in the GII region of TaqFfh, where the apo structure displays signiWcantly lower B factors than the Mg2+ GDP-complex or the sulfate-loaded form of the protein. It has been suggested that the increased disorder of the GII loop is speciWc to the Mg2+ GDP-bound state (Freymann et al., 1999). Our observation, that binding of sulfate ions in TaqFfh and MmFtsY results in the same destabilization of the GII-consensus element, shows that the disorder is linked to the binding site at the GI loop. It seems
unlikely that the structural diVerences in the F1 and F2 models are due to crystal packing interactions. First, both of the independently reWned molecules in the asymmetric unit of the F2 models display the same structural change. Second, this change is not localized to one region of the protein but found at three of the consensus regions, GI, GII, and GIII. Finally, the two structural conformations are not novel and have been observed in apo- and ADPloaded forms of some ATPases, as will be discussed. 4.2. FtsY shares structural similarities to ATPases So far the conformational change at the nucleotidebinding site of MmFtsY has not been observed in other GTPases. We found, however, that these rearrangements are reminiscent of those occurring in some ATPases. This suggests that FtsY could be structurally more closely related to ATPases than previously believed. It also highlights an alternate nucleotide uptake. Interestingly, it has been reported that FtsY harbors a basal ATPase activity (Shan and Walter, 2003). The P-loop NTPases share a speciWc fold (Milner-White et al., 1991; Saraste et al., 1990; Walker et al., 1982). This fold is divided into seven major lineages belonging to two major groups (Bourne et al., 1991). One group includes the nucleotide kinases (e.g., adenosine kinases) and GTPases, where the strand leading to the P-loop and the GIII-element strand are direct neighbors. The other group, populated solely by ATPases (e.g., ABC ATPases such as MutS), is characterized by an additional strand inserted between those of the P-loop and the GIII-consensus element. Computerized searches of the SRP G domain against structures from the PDB identiWed predominantly ATPases as structurally related to the SRP GTPases (Montoya et al., 2000; Schlessman et al., 1998). This led to the division of GTPases into two classes. The SRP GTPases were grouped together with related ATPases in the SIMIBI class comprising the SRP, MinD/Mrp, and Bio-D related superfamilies, while translation factors and Ras GTPase were included in the TRAFAC class (Leipe et al., 2002). The distinction between these two new GTPase classes arises not only from an analysis of the three-dimensional structures but from the amino acid sequence as well. In addition, SIMIBI NTPases tend to form homodimers in, e.g., BioD, Nip, ArsA, or heterodimers as in the case of the SRP GTPases. This is in contrast to the TRAFAC GTPases, which typically function as monomers or in larger complexes with other proteins. For the ATPases MutS (Lamers et al., 2000) and adenosine kinase (Schumacher et al., 2000), the transition from an apo conformation to a nucleotide-loaded conformation is accompanied by a P-loop reorganization as seen in the F1 and F2 forms of MmFtsY (Fig. 6). Considering that some ATPases populate the SIMIBI class, this similarity seems less exceptional. However, MutS and adenosine kinase are not members of the same NTPase superfamily; they are both more remotely related to the SRP GTPases than the TRAFAC GTPases.
T. Gariani et al. / Journal of Structural Biology 153 (2006) 85–96
93
Fig. 6. Structural similarities between ATPase and MmFtsY in the GI region. (A) Ribbon C-trace over the homo-dimeric structure of MutS (PDB code 1E3M, (Lamers et al., 2000)). ADP is present in one monomer as indicated. The P1 loops in the apo- and the ADP-loaded monomers are shown in green and orange, respectively. (B) Superimposed structures of adenosine kinase. The apo structure is shown in magenta (PDB code 1LIO, Schumacher et al., 2000), and the complex with an ATP analogue is shown in yellow (PDB code 1LII, (Schumacher et al., 2000)). (C) Superposition of the P1 loops of model F1 (red) and model F2 (blue) of MmFtsY. (For interpretation of the references to colour in this Wgure legend, the reader is referred to the web version of this paper.)
There is evidence that a common ancestor of the members of the P-loop GTPase superclass was a protein with a NK£D motif (Leipe et al., 2002). The ATPases belonging to this superclass evolved either through the loss of the NK£D motif (kinesin–myosin) or through divergence in the region containing the NK£D motif (MinD and Bio-D-related family). It is therefore tempting to propose that the FtsY protein structurally bridges the SRP GTPases and the SIMIBIrelated ATPases. This could explain why FtsY has a basal ATPase activity while Ffh has not (Shan and Walter, 2003).
4.3. Signaling mechanism of the NG domain The mechanism by which the SRP GTPases regulate the SRP pathway is correlated to the architecture unique to this family of proteins. In addition to the two conserved N and G domains, the M domain in Ffh and the A domain in FtsY further complicate the picture. For proper function, the G domain must sense when FtsY is membrane-bound, which necessitates that a communication system between the G, N, and A domains has to be in place. Recent data
94
T. Gariani et al. / Journal of Structural Biology 153 (2006) 85–96
have shed light on the function of the various domains in the particle. The NG domain of SRP54 and Ffh is required for signal-sequence binding (Cleverley and Gierasch, 2002; Newitt and Bernstein, 1997) and increases the aYnity of SRP54 for SRP RNA (Diener and Wilson, 2000; Samuelsson and Olsson, 1993). Similarly, in FtsY the N domain is involved in binding to the membrane (Millman and Andrews, 1997, 1999), where it probably functions as a sensor of the GTP-dependent FtsY–Ffh complex formation (Egea et al., 2004; Focia et al., 2004; Ramirez et al., 2002). The A domain of FtsY is located immediately upstream of the N domain. Also the A domain is suggested to have a role in membrane anchoring and GTPase regulation (de Leeuw et al., 1997, 2000; Herskovits et al., 2001; Millman and Andrews, 1999; Millman et al., 2001; Powers and Walter, 1997; Zelazny et al., 1997). Intriguingly, the A domain varies considerably in size and sequence. In E. coli and M. mycoides the A domain comprises approximately 200 and 100 amino acids, respectively, whereas in other species it includes only a few amino acids (Rosenblad et al., 2003). The A domain seems to be particularly important for FtsY’s initial association with the membrane. Once it is bound to the membrane it can be proteolytically removed from the NG domain without eVects on FtsY function (Herskovits et al., 2001). The border between the A and N domain has so far not been structurally deWned. The structure of the NG domain of EcFtsY included residues 197–497; however, the Wrst four amino acids at the N-terminal end were not deWned in the electron density and consequently the A1 helix is missing in the EcFtsY structure (Montoya et al., 1997a,b). In a recent study, a number of deletion mutants in the N-terminal end of EcFtsY were tested for their ability to complement an FtsY-depleted E. coli strain (Eitan and Bibi, 2004). Interestingly, it was found that whereas the original NG domain construct (residues 197–497) lacked biological activity, the addition of only one more amino acid at its Nterminal end restored the protein’s activity. This suggests that only the uttermost C-terminal part of the A domain is essential for membrane assembly of the receptor (Eitan and Bibi, 2004). Thermotoga martima FtsY (TmFtY) represents a minimalistic FtsY protein with only a few amino acids constituting its A domain (Fig. 1B). Its structure has recently been determined (PDB accession code 1VMA, Joint Center for Structural Genomics, unpublished). In TmFtsY, the helices A1 and A2 have conformations that are identical to those in MmFtsY. Furthermore, two water molecules W5 and W13 in TmFtsY are identically located as compared to W8 and W201 in MmFtsY. In TmFtsY, the A1 helix is extended with one turn and comprises the entire N-terminal end of the protein starting with residue Met-1. The side chain of Phe-4 from TmFtsY forms extensive packing interactions with helix E situated at the C-terminal end of the protein. This amino acid corresponds to Phe-196 in EcFtsY. Phe-196 was not included in the fragment used for the structural study (Montoya et al., 1997a) but is known to be crucial for the biological activity of the protein (Eitan and Bibi, 2004). It
Fig. 7. Positively charged residues and hydrophobic residues located at the polar interface of the amphiphatic helix A2.
seems plausible that an extended construct of EcFtsY could stabilize this region of the protein as a deWned -helical structure A1, similar to what is seen in MmFtsY and TmFtsY. We suggest that there is a communication between the N and G domains at the interface between helices A1 and A2 and the G domain. With helix A1 properly folded, helix A2 is highly ordered in both MmFtsY and TmFtsY in such a way that a cluster of lysine and hydrophobic residues are located on the same surface-exposed side of helix A2 (Fig. 7). These residues seem to be in an ideal position for membrane anchoring, for example through lysine snorkeling (Mishra and Palgunachari, 1996; Segrest et al., 1990) to the negatively charged phospholipid head groups of the membrane (de Leeuw et al., 2000; Millman et al., 2001). Helix A1 and A2 could therefore constitute a linker between membrane association and the GIV-consensus element in the G domain, which senses the conformational change that takes place when the protein is anchored to the membrane and primes the protein for hydrolysis (Millman and Andrews, 1999). In particular, the conserved water-mediated hydrogen interactions formed between Asp-356 and residues from helix A1, could be crucial in mediating such a communication. Acknowledgments We thank Terese Bergfors for valuable discussions and critical reading of the manuscript. This study was
T. Gariani et al. / Journal of Structural Biology 153 (2006) 85–96
supported by the Swedish Research Council (K5104-1098) and European Union (CLK3-CT-2000-00082). References Bourne, H.R., Sanders, D.A., McCormick, F., 1991. The GTPase superfamily: conserved structure and molecular mechanism. Nature 349, 117–127. Brünger, A.T., Adams, P.D., Clore, G.M., DeLano, W.L., Gros, P., Grosse-Kunstleve, R.W., Jiang, J.S., Kuszewski, J., Nilges, M., Pannu, N.S., Read, R.J., Rice, L.M., Simonson, T., Warren, G.L., 1998. Crystallography & NMR system: a new software suite for macromolecular structure determination. Acta Crystallogr. D Biol. Crystallogr. 54, 905–921. Cleverley, R.M., Gierasch, L.M., 2002. Mapping the signal sequence-binding site on SRP reveals a signiWcant role for the NG domain. J. Biol. Chem. 277, 46763–46768. Collaborative Computational Project, No. 4, 1994. The CCP4 suite: programs for protein crystallography. Acta Crystallogr. D 50, 760–763. Connolly, T., Gilmore, R., 1989. The signal recognition particle receptor mediates the GTP-dependent displacement of SRP from the signal sequence of the nascent polypeptide. Cell 57, 599–610. de Leeuw, E., Poland, D., Mol, O., Sinning, I., ten Hagen-Jongman, C.M., Oudega, B., Luirink, J., 1997. Membrane association of FtsY, the E. coli SRP receptor. FEBS Lett. 416, 225–229. de Leeuw, E., te Kaat, K., Moser, C., Menestrina, G., Demel, R., de KruijV, B., Oudega, B., Luirink, J., Sinning, I., 2000. Anionic phospholipids are involved in membrane association of FtsY and stimulate its GTPase activity. EMBO J. 19, 531–541. Diener, J.L., Wilson, C., 2000. Role of SRP19 in assembly of the Archaeoglobus fulgidus signal recognition particle. Biochemistry 39, 12862– 12874. Doudna, J.A., Batey, R.T., 2004. Structural insights into the signal recognition particle. Annu. Rev. Biochem. 73, 539–557. Egea, P.F., Shan, S.O., Napetschnig, J., Savage, D.F., Walter, P., Stroud, R.M., 2004. Substrate twinning activates the signal recognition particle and its receptor. Nature 427, 215–221. Eitan, A., Bibi, E., 2004. The core Escherichia coli signal recognition particle receptor contains only the N and G domains of FtsY. J. Bacteriol. 186, 2492–2494. Focia, P.J., Shepotinovskaya, I.V., Seidler, J.A., Freymann, D.M., 2004. Heterodimeric GTPase core of the SRP targeting complex. Science 303, 373–377. Freymann, D.M., Keenan, R.J., Stroud, R.M., Walter, P., 1997. Structure of the conserved GTPase domain of the signal recognition particle. Nature 385, 361–364. Freymann, D.M., Keenan, R.J., Stroud, R.M., Walter, P., 1999. Functional changes in the structure of the SRP GTPase on binding GDP and Mg2+GDP. Nat. Struct. Biol. 6, 793–801. Fulga, T.A., Sinning, I., Dobberstein, B., Pool, M.R., 2001. SRbeta coordinates signal sequence release from SRP with ribosome binding to the translocon. EMBO J. 20, 2338–2347. Gariani, T., Sauer-Eriksson, E., 2000. Crystallization and preliminary Xray diVraction studies of the signal recognition particle receptor FtsY from Mycoplasma mycoides. Acta Crystallogr. D Biol. Crystallogr. 56, 1030–1032. Gilmore, R., Blobel, G., Walter, P., 1982. Protein translocation across the endoplasmic reticulum. I. Detection in the microsomal membrane of a receptor for the signal recognition particle. J. Cell Biol. 95, 463–469. Halic, M., Becker, T., Pool, M.R., Spahn, C.M., Grassucci, R.A., Frank, J., Beckmann, R., 2004. Structure of the signal recognition particle interacting with the elongation-arrested ribosome. Nature 427, 808–814. Harris, M., Jones, T.A., 2001. Molray—a web interface between O and the POV-Ray ray tracer. Acta Crystallogr. D Biol. Crystallogr. 57, 1201– 1203. Herskovits, A.A., Seluanov, A., Rajsbaum, R., ten Hagen-Jongman, C.M., Henrichs, T., Bochkareva, E.S., Phillips, G.J., Probst, F.J., Nakae, T.,
95
Ehrmann, M., Luirink, J., Bibi, E., 2001. Evidence for coupling of membrane targeting and function of the signal recognition particle (SRP) receptor FtsY. EMBO Rep. 2, 1040–1046. Jones, T.A., Zou, J.Y., Cowan, S.W., Kjeldgaard, M., 1991. Improved methods for binding protein models in electron density maps and the location of errors in these models. Acta Crystallogr. A 47, 110–119. Keenan, R.J., Freymann, D.M., Stroud, R.M., Walter, P., 2001. The signal recognition particle. Annu. Rev. Biochem. 70, 755–775. Keenan, R.J., Freymann, D.M., Walter, P., Stroud, R.M., 1998. Crystal structure of the signal sequence binding subunit of the signal recognition particle. Cell 94, 181–191. Kjeldgaard, M., Nyborg, J., 1992. ReWned structure of elongation factor EF–Tu from Escherichia coli. J. Mol. Biol. 223, 721–742. Kraulis, P.J., 1991. MOLSCRIPT: A program to produce both detailed and schematic plots of protein structures. J. Appl. Crystallogr. 24, 946–950. Lamers, M.H., Perrakis, A., Enzlin, J.H., Winterwerp, H.H., de Wind, N., Sixma, T.K., 2000. The crystal structure of DNA mismatch repair protein MutS binding to a G £ T mismatch. Nature 407, 711–717. Leipe, D.D., Wolf, Y.I., Koonin, E.V., Aravind, L., 2002. ClassiWcation and evolution of P-loop GTPases and related ATPases. J. Mol. Biol. 317, 41–72. Luirink, J., Dobberstein, B., 1994. Mammalian and Escherichia coli signal recognition particles. Mol. Microbiol. 11, 9–13. Luirink, J., Sinning, I., 2004. SRP-mediated protein targeting: structure and function revisited. Biochim. Biophys. Acta 1694, 17–35. Luirink, J., ten Hagen-Jongman, C.M., van der Weijden, C.C., Oudega, B., High, S., Dobberstein, B., Kusters, R., 1994. An alternative protein targeting pathway in Escherichia coli: studies on the role of FtsY. EMBO J. 13, 2289–2296. Lütcke, H., High, S., Romisch, K., Ashford, A.J., Dobberstein, B., 1992. The methionine-rich domain of the 54 kDa subunit of signal recognition particle is suYcient for the interaction with signal sequences. EMBO J. 11, 1543–1551. Macao, B., Luirink, J., Samuelsson, T., 1997. Ffh and FtsY in a Mycoplasma mycoides signal-recognition particle pathway: SRP RNA and M domain of Ffh are not required for stimulation of GTPase activity in vitro. Mol. Microbiol. 24 (3), 523–534. Meyer, D.I., Krause, E., Dobberstein, B., 1982. Secretory protein translocation across membranes-the role of the ‘docking protein’. Nature 297, 647–650. Miller, J.D., Bernstein, H.D., Walter, P., 1994. Interaction of E. coli Ffh/ 4.5S ribonucleoprotein and FtsY mimics that of mammalian signal recognition particle and its receptor. Nature 367, 657–659. Miller, J.D., Wilhelm, H., Gierasch, L., Gilmore, R., Walter, P., 1993. GTP binding and hydrolysis by the signal recognition particle during initiation of protein translocation. Nature 366, 351–354. Millman, J.S., Andrews, D.W., 1997. Switching the model: a concerted mechanism for GTPases in protein targeting. Cell 89, 673–676. Millman, J.S., Andrews, D.W., 1999. A site-speciWc, membrane-dependent cleavage event deWnes the membrane binding domain of FtsY. J. Biol. Chem. 274, 33227–33234. Millman, J.S., Qi, H.Y., Vulcu, F., Bernstein, H.D., Andrews, D.W., 2001. FtsY binds to the Escherichia coli inner membrane via interactions with phosphatidylethanolamine and membrane proteins. J. Biol. Chem. 276, 25982–25989. Milner-White, E.J., Coggins, J.R., Anton, I.A., 1991. Evidence for an ancestral core structure in nucleotide-binding proteins with the type A motif. J. Mol. Biol. 221, 751–754. Mishra, V.K., Palgunachari, M.N., 1996. Interaction of model class A1, class A2, and class Y amphipathic helical peptides with membranes. Biochemistry 35, 11210–11220. Montoya, G., Kaat, K., Moll, R., Schafer, G., Sinning, I., 2000. The crystal structure of the conserved GTPase of SRP54 from the archaeon Acidianus ambivalens and its comparison with related structures suggests a model for the SRP–SRP receptor complex. Struct. Fold. Des. 8, 515–525. Montoya, G., Svensson, C., Luirink, J., Sinning, I., 1997a. Crystal structure of the NG domain from the signal-recognition particle receptor FtsY. Nature 385, 365–368.
96
T. Gariani et al. / Journal of Structural Biology 153 (2006) 85–96
Montoya, G., Svensson, C., Luirink, J., Sinning, I., 1997b. Expression, crystallization, and preliminary X-ray diVraction study of FtsY, the docking protein of the signal recognition particle of E. coli. Proteins 28, 285–288. Moser, C., Mol, O., Goody, R.S., Sinning, I., 1997. The signal recognition particle receptor of Escherichia coli (FtsY) has a nucleotide exchange factor built into the GTPase domain. Proc. Natl. Acad. Sci. USA 94, 11339–13344. Navaza, J., 1994. AMoRe: an automated package for molecular replacement. Acta Crystallogr. A 50, 157–163. Newitt, J.A., Bernstein, H.D., 1997. The N-domain of the signal recognition particle 54-kDa subunit promotes eYcient signal sequence binding. Eur. J. Biochem. 245, 720–729. Otwinowski, Z., Minor, W., 1997. Processing of X-ray diVraction data collected in oscillation mode. Methods Enzymol. 276, 307–326. Padmanabhan, S., Freymann, D.M., 2001. The conformation of bound GMPPNP suggests a mechanism for gating the active site of the SRP GTPase. Structure (Camb) 9, 859–867. Pool, M.R., Stumm, J., Fulga, T.A., Sinning, I., Dobberstein, B., 2002. Distinct modes of signal recognition particle interaction with the ribosome. Science 297, 1345–1348. Powers, T., Walter, P., 1995. Reciprocal stimulation of GTP hydrolysis by two directly interacting GTPases. Science 269, 1422–1424. Powers, T., Walter, P., 1997. Co-translational protein targeting catalyzed by the Escherichia coli signal recognition particle and its receptor. EMBO J. 16, 4880–4886. Ramirez, U.D., Minasov, G., Focia, P.J., Stroud, R.M., Walter, P., Kuhn, P., Freymann, D.M., 2002. Structural basis for mobility in the 1.1 A crystal structure of the NG domain of Thermus aquaticus Ffh. J. Mol. Biol. 320, 783–799. Rapiejko, P.J., Gilmore, R., 1997. Empty site forms of the SRP54 and SR alpha GTPases mediate targeting of ribosome-nascent chain complexes to the endoplasmic reticulum. Cell 89, 703–713. Rinke-Appel, J., Osswald, M., von Knoblauch, K., Mueller, F., Brimacombe, R., Sergiev, P., Avdeeva, O., Bogdanov, A., Dontsova, O., 2002. Crosslinking of 4.5S RNA to the Escherichia coli ribosome in the presence or absence of the protein Ffh. RNA 8, 612–625. Rosenblad, M.A., Gorodkin, J., Knudsen, B., Zwieb, C., Samuelsson, T., 2003. SRPDB: signal recognition particle database. Nucleic Acids Res. 31, 363–364. Samuelsson, T., 1992. A Mycoplasma protein homologous to mammalian SRP54 recognizes a highly conserved domain of SRP RNA. Nucleic Acids Res. 20, 5763–5770. Samuelsson, T., Olsson, M., 1993. GTPase activity of a bacterial SRP-like complex. Nucleic Acids Res. 21, 847–853.
Saraste, M., Sibbald, P.R., Wittinghofer, A., 1990. The P-loop–a common motif in ATP- and GTP-binding proteins. Trends Biochem. Sci. 15, 430–434. ScheVzek, K., Klebe, C., Fritz-Wolf, K., Kabsch, W., Wittinghofer, A., 1995. Crystal structure of the nuclear Ras-related protein Ran in its GDP- bound form. Nature 374, 378–381. Schlessman, J.L., Woo, D., Joshua-Tor, L., Howard, J.B., Rees, D.C., 1998. Conformational variability in structures of the nitrogenase iron proteins from Azotobacter vinelandii and Clostridium pasteurianum. J. Mol. Biol. 280, 669–685. Schlichting, I., Almo, S.C., Rapp, G., Wilson, K., Petratos, K., Lentfer, A., Wittinghofer, A., Kabsch, W., Pai, E.F., Petsko, G.A., et al., 1990. Time-resolved X-ray crystallographic study of the conformational change in Ha-Ras p21 protein on GTP hydrolysis. Nature 345, 309– 315. Schumacher, M.A., Scott, D.M., Mathews, I.I., Ealick, S.E., Roos, D.S., Ullman, B., Brennan, R.G., 2000. Crystal structures of Toxoplasma gondii adenosine kinase reveal a novel catalytic mechanism and prodrug binding. J. Mol. Biol. 298, 875–893. Segrest, J.P., De Loof, H., Dohlman, J.G., Brouillette, C.G., Anantharamaiah, G.M., 1990. Amphipathic helix motif: classes and properties. Proteins 8, 103–117. Shan, S.O., Walter, P., 2003. Induced nucleotide speciWcity in a GTPase. Proc. Natl. Acad. Sci. USA 100, 4480–4485. Sprang, S.R., 1997. G protein mechanisms: insights from structural analysis. Annu. Rev. Biochem. 66, 639–678. Ulbrandt, N.D., Newitt, J.A., Bernstein, H.D., 1997. The E. coli signal recognition particle is required for the insertion of a subset of inner membrane proteins. Cell 88, 187–196. Walker, J.E., Saraste, M., Runswick, M.J., Gay, N.J., 1982. Distantly related sequences in the alpha- and beta-subunits of ATP synthase, myosin, kinases and other ATP-requiring enzymes and a common nucleotide binding fold. EMBO J. 1, 945–951. Wild, K., Halic, M., Sinning, I., Beckmann, R., 2004. SRP meets the ribosome. Nat. Struct. Mol. Biol. 11, 1049–1053. Wittinghofer, F., Krengel, U., John, J., Kabsch, W., Pai, E.F., 1991. Threedimensional structure of p21 in the active conformation and analysis of an oncogenic mutant. Environ. Health Perspect. 93, 11–15. Zelazny, A., Seluanov, A., Cooper, A., Bibi, E., 1997. The NG domain of the prokaryotic signal recognition particle receptor, FtsY, is fully functional when fused to an unrelated integral membrane polypeptide. Proc. Natl. Acad. Sci. USA 94, 6025–6029. Zopf, D., Bernstein, H.D., Walter, P., 1993. GTPase domain of the 54-kD subunit of the mammalian signal recognition particle is required for protein translocation but not for signal sequence binding. J. Cell Biol. 120, 1113–1121.
Conformational variability of the GTPase domain of the signal recognition particle receptor FtsY Talal Gariani a,1, Tore Samuelsson b, A. Elisabeth Sauer-Eriksson a,¤ a
b
Umeå Centre for Molecular Pathogenesis, Umeå University, SE-901 87 Umeå, Sweden Department of Medical Biochemistry, University of Göteborg, Medicinareg. 9A, SE-405 30 Göteborg, Sweden Received 3 August 2005; received in revised form 20 October 2005; accepted 24 October 2005 Available online 30 November 2005
Abstract The prokaryotic signal recognition particle Ffh and its receptor FtsY allow targeting of proteins into or across the plasma membrane. The targeting process is GTP dependent and the two proteins constitute a distinct GTPase family. The receptor FtsY is composed of A and NG domains where the NG’s GTPase domain plays a critical role in the targeting process. In this study, we describe two X-ray structures determined independently of each other of the NG domain of FtsY from Mycoplasma mycoides (MmFtsY). The two structures are markedly diVerent in three of the nucleotide-binding segments, GI (P-loop), GII, and GIII, making only one of the structures compatible with nucleotide binding. Interestingly, the two distinct conformations of the nucleotide-binding segments of MmFtsY are similar to the apo- and ADP-loaded forms of certain ATPases. The structure of the extended interface between the A and NG domains of MmFtsY provides new insights into the role of the A domain for phospholipid interaction. 2005 Elsevier Inc. All rights reserved. Keywords: Signal recognition particle; SRP; SRP receptor; FtsY; Mycoplasma mycoides
1. Introduction The signal recognition particle (SRP) is one of the key components of the cellular machinery for membrane targeting. Many of its structural and functional features have been highly conserved during evolution; the SRP and its receptor have been identiWed in all three kingdoms of life (for reviews see Doudna and Batey, 2004; Keenan et al., 2001; Luirink and Sinning, 2004; Wild et al., 2004). In eukaryotic cells, the SRP targets proteins into or across the membrane of the endoplasmic reticulum (ER). Eukaryotic SRP comprises one 7S RNA molecule (»300 nucleotides) and six protein components named SRP9, SRP14, SRP19, SRP54, SRP68, and SRP72 according to *
Corresponding author. Fax: +46 90 778007. E-mail address: [email protected] (A.E. Sauer-Eriksson). 1 Present address: SanoW-Aventis, Chemical Science, Centre de Recherche Paris, Friedel, B 122.2, 13 Quai Jules Guesde, 94403 Vitry-sur-Seine Cedex, France. 1047-8477/$ - see front matter 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.jsb.2005.10.003
their mass in kilodalton. The SRP pathway in prokaryotes is similar to that in eukaryotes. The major protein targets for bacterial SRP are integral membrane proteins (Luirink and Dobberstein, 1994; Ulbrandt et al., 1997). Eubacterial SRP has only two components: a 4.5S RNA and the Ffh protein. The Ffh protein is homologous to SRP54, which has a key role in recognizing the signal sequence of the nascent polypeptide chain as it emerges from the ribosomes at the peptide exit site (Halic et al., 2004; Pool et al., 2002; Rinke-Appel et al., 2002), and in binding to the SRP receptor (SR). The eukaryotic SR is a heterodimer made up of the GTPases SR and SR (Gilmore et al., 1982; Meyer et al., 1982; Rapiejko and Gilmore, 1997), whereas the eubacterial SR consists of only one protein, FtsY, homologous to SR (Luirink et al., 1994). Mutual GTP hydrolysis of SRP and its receptor regulate the SRP pathway such that SRP is released from the ribosome complex (Miller et al., 1993, 1994; Powers and Walter, 1995), and translocation is allowed (Connolly and Gilmore, 1989; Fulga et al., 2001).
86
T. Gariani et al. / Journal of Structural Biology 153 (2006) 85–96
Ffh and FtsY are evolutionarily related and share N and G (GTPase) domains, where the G domains have GTP-consensus elements GI–GV (Bourne et al., 1991). In addition, Ffh includes a methionine-rich M domain, which binds both to the signal sequence and to RNA (Lütcke et al., 1992; Zopf et al., 1993). FtsY includes the highly acidic A domain whose function might be to interact with the lipid membrane (de Leeuw et al., 2000; Zelazny et al., 1997). The high resolution structures of the NG domains of Ffh from Thermus aquaticus (TaqFfh) (Freymann et al., 1997) and FtsY from Escherichia coli (EcFtsY) (Montoya et al., 1997a) provided us with the Wrst detailed description of these domains at the molecular level. The domain structures of Ffh and FtsY are very similar. The G domain has a fold similar to that of the Ras-like GTPases, but with an - insertion box domain (IBD) unique to the SRPGTPase family (Freymann et al., 1997; Montoya et al., 1997a). The N domain forms a four-helix bundle that can be considered as an extension of the G domain. This N domain has a role in GTP hydrolysis and in membrane binding in association with the A domain (Millman and Andrews, 1999). The recent crystal structures of the Ffh– FtsY complex from T. aquaticus revealed a new mechanism in which the association of the two proteins activates reciprocal GTP hydrolysis (Egea et al., 2004; Focia et al., 2004). SRP components have been identiWed in other bacteria, including mycoplasmas (Samuelsson, 1992). Mycoplasmas are simple eubacteria and have a small genome that codes for a highly limited set of proteins. They are insensitive to penicillin, and cause diseases including arthritis and pneumonia. The fact that they have retained an SRP pathway possibly reXects the essential nature of this machinery. We have previously shown that the GTPase activity of mycoplasma FtsY is stimulated when it interacts with Ffh. In contrast to E. coli, however, this stimulation is possible without SRP RNA and the M domain of Ffh (Macao et al., 1997). This indicates that the interaction of the NG domains of Ffh and FtsY plays an important part in the activation mechanism, and the mycoplasma proteins constitute an interesting model system for SRP function. In this study, we present structures of the NG domain of FtsY in two diVerent forms from Mycoplasma mycoides (MmFtsY): an apo and a sulfate-loaded form. Whereas the sulfate-loaded form is similar to the previously determined structures of both Ffh and FtsY, large structural changes in three of the GTP-binding loops were observed in the apo form. This implies that the loops in FtsY possess a higher intrinsic mobility than previously believed. Interestingly, the two conformations of the nucleotide-binding segments of MmFtsY are similar to the apo- and ADP-loaded forms of certain ATPases. 2. Materials and methods 2.1. Structure determination The NG domain of MmFtsY was engineered and puriWed as previously described (Gariani and Sauer-Eriksson,
Table 1 Data collection and reWnement statistics F1 Quality of reXections used in reWnement Space group P21212 Unit cell parameters (Å) a D 68.74, b D 101.13 c D 42.53 Resolution range (Å) 20–1.90 (1.97–1.9) Number of observation 497,368 Number of unique 22,061 reXection Completeness (%) 95.7 (79.5) I/(I) 17.45 (4.78) 6.9 (32.1) Rsyma,b (%) I/(I) > 2 83.2 (51.5) ReWnement statistics Resolution ReXections work set ReXections test set Rcryst Rfreec Number of protein atoms Chain A Chain B Number of water molecules Rms deviations Bonds (Å) Angles (o) Average B factor (Å2) Chain A Chain B Sulfate molecule Water molecules
F2 R32 a D b D 148.44 c D 223.93 20–2.4 (2.46–2.40) 630,367 35,370 94.8 (87.3) 21.68 (2.49) 5.6 (60.9) 82.2 (49.4)
20–1.95 18,640 2062 20.7 26.4
20–2.4 33,620 1750 20.9 24.6
2405 — 252
2439 2453 139
0.017 1.62
0.018 1.67
26.0 — 39.0 34.5
52.1 53.7 — 45.2
a
Value in parentheses is in the high resolution bin. Rsym D hkli冷Ii(hkl)-具I(hkl)典冷/hkliIi(hkl). c Rcryst/Rfree D 储Fo冷 ¡ 冷Fc储/ 冷Fo冷. Rfree was calculated using 10% (model F1) and 5% (model F2) of data excluded from reWnement. b
2000). At the N-terminal region, sequence 92-KEKDKKV98 was substituted with the His-tag 92-HHHHHPM-98. Crystals were obtained with two diVerent crystallization precipitants: ammonium sulfate (model F1) and sodium citrate (model F2) (Gariani and Sauer-Eriksson, 2000). DiVraction data to 1.95 and 2.4 Å resolution were collected at 100 K on the F1 and F2 crystals, respectively, at beam line X-11 in DESY, Hamburg. The data were processed with DENZO (Otwinowski and Minor, 1997), and merged using the programs TRUNCATE and SCALA from the CCP4 suite (Collaborative Computational Project, 1994). A summary of the data collection statistics is given in Table 1. Although the sequential and structural similarities between the NG domains of EcFtsY and MmFtsY are quite high (34% identity for the whole molecule, 44% identity for the G domain), solving the structure by molecular replacement (MR) was not straightforward. The NG domain of EcFtsY (PDB accession code 1FTS, Montoya et al., 1997a) was used as a search model for CNS (Brünger et al., 1998). The model was progressively truncated to 680 atoms before the correct solution appeared among the top ten solutions. This corresponded
T. Gariani et al. / Journal of Structural Biology 153 (2006) 85–96
to 30% of the protein and included only a polyalanine model of secondary structural elements from the G domain. After initial rigid-body reWnement, the Rwork and Rfree for the correct solution decreased from 55.0 and 54.0% to 54.7 and 53.0%, respectively, whereas the incorrect solutions remained constant or slightly increased their R values. Only by comparing the quality of the electron-density maps could the correct solution be identiWed. Subsequent rounds of model building and reWnement using the programs O (Jones et al., 1991) and CNS (Brünger et al., 1998) reduced the Rwork and Rfree to 20.7 and 26.4%, respectively. The F2 structure was determined by MR methods using AMoRe (Navaza, 1994) with the entire F1 structure as search model. The structure was reWned using a similar procedure as for F1 to a Wnal Rwork and Rfree of 20.9% and 24.6%, respectively. Figures were made with the programs MOSLCRIPT (Kraulis, 1991), PYMOL (DeLano W.L., http://www.pymol.org), and MOLRAY–POVRAY (Harris and Jones, 2001). 2.2. Protein DataBank accession number Coordinates and structure factors have been deposited in the Protein Data Bank [accession codes 1ZU4 (model F1) and 1ZU5 (model F2)].
87
(Montoya et al., 1997a), TaqFfh and TaqFtsY (Egea et al., 2004; Focia et al., 2004; Freymann et al., 1997, 1999; Keenan et al., 1998; Padmanabhan and Freymann, 2001), and Acidianus ambivalens, AaFfh (Montoya et al., 2000) (Fig. 1A). The sequence identities between MmFtsY and EcFtsY, and between MmFtsY and TaqFtsY, are 34% and 35%, respectively (Fig. 1B). The N domain is made of a four-helix bundle (helices A2, B–D), which in MmFtsY comprises 78 residues at the N-terminal end (Met-102 to Asp-179). The G domain of MmFtsY is formed by residues Leu-196 to Asp-391, which are structurally homologous to other SRP GTPases and include the Wve conserved nucleotide-consensus elements required for GTP binding and hydrolysis: GI or P loop (Gly-203 to Ser-212), GII (Asp233 to Ala-239), GIII (Asp-289 to Gln-295), GIV (Thr-353 to Asp-356), and GV or ‘closing loop’ (Gly-379 to Glu382). Helix 7 (also named G5), comprising eleven residues at the C-terminal end (Asp-391 to Ser-401), seems more connected to the N domain than the G domain. We therefore refer to this helix as helix E. The –– motif (IBD, Ala-237 to Gln-281), which is unique for the SRP GTPases, is inserted between strands 2 and 4. At the Nterminal end of the N domain, helix A1 (Pro-97 to Met-102) is situated. This helix is not structurally deWned in the previously determined SRP GTPases. Structurally, we Wnd that the A1 helix seems more associated with the G domain than with the N domain.
3. Results 3.1. Quality of the F1 and F2 models The structures of the NG domain of MmFtsY were determined from two crystal forms. One sulfate ion was identiWed at the nucleotide-binding site in model F1, which therefore represents a sulfate-bound form of the protein. The F1 model includes 307 out of the 318 residues, which comprise the NG domain (Glu-99 to Gln-405), and two residues, Pro-97 and Met-98, which are part of the His-tag. Omissions from the model because of no visible electron density are: Wve residues of the His-tag (His-92 to His-96), the loop connecting the N and G domains (Thr-183 to Lys-186), and the last six residues of the C-terminal end (Gly-406 to Lys-411). Additional residues with weak or no density for their side chains include predominantly polar amino acids situated at surface-exposed regions of the protein, and the entire side chain of Phe-235, which is part of the GII-loop region. The F2 structure comprises two molecules, A and B, in the asymmetric unit. The two molecules are very similar with a root mean squared (rms) deviation of 0.4 Å for all C atoms. Residue Phe-235 in the GII loop has well deWned electron densities in both molecules of the F2 model. 3.2. General architecture of MmFtsY The NG domain of MmFtsY presented the same overall fold as the previously determined structures from EcFtsY
3.3. The F1 model of MmFtsY is similar to the previous structures of FtsY and Ffh The structure of the Wve nucleotide-binding segments of model F1 is similar to the previously published SRP GTPases, however, with a better structural Wt to the segments of FtsY than the Ffh structures. The GI loop (P-loop) comprises residues Gly-203 to Ser-212 and is situated on a Xexible loop bridging 1 and 1. The role of the P loop is to properly position the -phosphate of the GTP and to bind to the Mg2+ ion that is necessary for GTP hydrolysis. In the F1 model of MmFtsY, there is one sulfate ion situated at the GI loop that is not present in EcFtsY (Fig. 2). The sulfate ions form hydrogen bonds to Wve consecutive main chain nitrogen atoms of the switch I regions (residues Gly-206 to Thr-210), to the N 2 atom of Arg-236, and to two water molecules, W1 and W2. The position of the sulfate ion is almost superimposable onto the position of the -phosphate position of the GMP-PCP analogue in the TaqFtsY–TaqFfh complex structure (Egea et al., 2004; Focia et al., 2004). An extensive hydrogen-bonding network with the sulfate ion at a pivotal position connects the GI, GII, and GIII regions in the F1 model (Fig. 2). The GII region of small GTPases is involved in Mg2+ binding. The role of the GII loop in SRP GTPases is less well deWned; studies on TaqFfh in complex with GTP analogues showed no interaction between residues situated at the GII region to the bound Mg2+ ion (Egea et al., 2004; Focia et al., 2004; Freymann et al., 1999). Residues Phe-235
88
T. Gariani et al. / Journal of Structural Biology 153 (2006) 85–96
T. Gariani et al. / Journal of Structural Biology 153 (2006) 85–96
89
Fig. 1. (A) Ribbon trace of the F1 model of MmFtsY. The N domain, G domain, and IBD box are colored yellow, blue, and red, respectively. The -helix A1 is shown in turquoise and the sulfate ion in green. (B) Structural-based alignment of SRP GTPases from Mycoplasma mycoides (1ZU4), Thermus aquaticus (1OKK), Acidianus ambivalens (1J8M), and Thermotoga martima (1VMA). Sequences from TqFfh and AFfh are also included for comparison. Identical sequences are boxed in dark grey and the Wve conserved GTP-binding regions are underlined in pink. The LIQTDM and TSKGG motifs are also highlighted. The extra helical turn present only in the P1 motif of model F2 is highlighted in magenta. Numbering is based on the M. mycoides FtsY sequence. The Wrst and last amino acids observed in the X-ray structures are highlighted in yellow. The two Wrst amino acids observed in MmFtsY (marked with *) constitute proline and methionine in our construct (see materials and methods). Secondary structures are colored as in (A). Helix 7 (also named G5), comprising eleven residues of the C-terminal end (Asp-391 to Ser-401), is more coupled to the N domain than the G domain. We therefore refer to this helix as helix E and show it in yellow. (For interpretation of the references to colour in this Wgure legend, the reader is referred to the web version of this paper.)
to Ala-237 situated in the GII region of the F1 model are more mobile compared to the rest of the F1 structure. In particular, the side chain of Phe-235 displays high Xexibility as evidenced by the lack of visible electron density for it. Residue Thr-234, which interacts with the Mg2+ ion and the -phosphate of the GTP nucleotide in Ras p21 (Schlichting et al., 1990; Wittinghofer et al., 1991), is oriented away from the GI loop. Its O1 atom forms a direct hydrogen bond to the main-chain carbonyl oxygen of Ala-265 that is situated in the IBD. This same interaction is found in the EcFtsY and TaqFtsY structures, and seems to be speciWc for the FtsY proteins. Residue Asp-233 points to the active site, and makes a salt link with the invariant Arg-236. In addition, Asp-233 is in contact with the GI and GIII regions through water-bridged interactions, W1, with the sulfate ion and the side chain of Asp-289. In EcFtsY, a similar pattern can be seen where the corresponding residue Asp-230 interacts with Asn-202 and Lys-207 from the GI region, and via two water molecules, W131 and W122, to the Asp-282 from the GIII region. The GIII-consensus element is well ordered in the F1 model of MmFtsY unlike what is seen in EcFtsY. In addition to the aforementioned interactions involving Asp-289 and Thr-290, Gly-292 is linked to residue Asn-205 from the GI region via three hydrogen bonds. These additional hydrogen bonds might
explain the stability of this consensus element with respect to the EcFtsY. The GIV (Thr-353 to Asp-356) and GV (i.e., the ‘closing loop’ Gly-379 to Glu-382) consensus elements interact with the nucleotide-ribose ring and are well conserved in the SRP-GTPase family. The GIV region is connected to the GI region through a hydrogen bond from the O atom of Thr-353 to the main chain carbonyl oxygen of Thr-207 via a water molecule W135 that is also found in the other SRP GTPase structures. The side chain of Lys-215 forms hydrogen bonds to the main chain oxygens of Gly-381 and Lys383. The GV region forms a -hairpin loop that is stabilized by polar interactions and similar in conformation to the GV loop of EcFtsY, TaqFtsY, TaqFfh, and AaFfh. Contacts between the GIV and GV regions are mediated through a direct hydrogen bond between the main-chain carbonyl oxygen of Gly-379 and the main-chain nitrogen of Met-355. In addition, three water molecules W9, W14, and W234, link Lys-354 and Asp-356 to Gly-381. 3.4. The F2 model of MmFtsY reveals a large conformational change at the nucleotide-binding site At the nucleotide-binding site, three major structural changes not found in model F1 have occurred. The Wrst is
Fig. 2. (A) Ribbon trace of model F1 at the GTP-binding site. Motifs I, II, and III are shown in red, and the sulfur atom is shown in green. (B) Detailed view of part of the extensive hydrogen-bonding network formed at the sulfate-binding site. (For interpretation of the references to colour in this Wgure legend, the reader is referred to the web version of this paper.)
90
T. Gariani et al. / Journal of Structural Biology 153 (2006) 85–96
Fig. 3. (A) Ribbon trace of model F2 at the GTP-binding site. In motif I, residues Asn-205 to Gly-208 are involved in a novel helical turn. (B) Detailed view of part of the hydrogen-bonding network. The orientation and coloring of the structure is as in Fig. 2. (For interpretation of the references to colour in this Wgure legend, the reader is referred to the web version of this paper.)
situated at the GI element. In the F1 model, water molecule W2 is at the N-terminal end of helix 1 that immediately follows the GI-loop region. In the F2 model, however, this water is missing and Asn-205, Gly-206, and Thr-207 have changed their positions such that helix 1 is extended by one turn (Fig. 3). This erases the active site known to accommodate the - and -phosphates of the substrate and the sulfate ion present in model F1. In its new position the O1 atom of Thr-207 is hydrogen-bonded to the N2 atom of Gln-242, the latter residue being situated at the C-terminal end of the GII region. The O1 atom of Asn-205 forms hydrogen bonds to the main-chain nitrogen atoms of Thr207 and Gly-208. At this position, two N2 atoms of A- and B-Asn-205 are situated 2.8 Å from each other in the asymmetric unit. Gly-206 is situated right at the sulfate-binding site in the F1 model. Due to the structural changes at the GI region, the two direct hydrogen bonds formed between Gly-292 and Asn205 are missing in the F2 model. Instead, a novel water molecule V3 in monomer A (U3 in monomer B) is introduced and connects the side chains of Lys-209, Asp-233, Arg-236, and Gly-292. The GI and GIII regions are situated further apart. For example, the distance between the C atoms of Gly-292 and Asn-207 is extended from 5.3 Å in the F1 model to 11.3 Å in the F2 model. This shift of the switch II region (residues Asp-289 to Arg-293) and the movement of helix 4 (residues Asn-296 to Lys-309) is reminiscent of the switch II conformation changes observed in small GTPases like EFTu, transducin, and Ran (Kjeldgaard and Nyborg, 1992; ScheVzek et al., 1995; Sprang, 1997). The third structural change is found at the GII region. As previously discussed, the residue Phe-235 is not stable in the sulfate-loaded form of MmFtsY. In the apo form a half turn of helix 2, which includes residues Arg-236 to Gly238, is lost. Residue Phe-235 is well deWned in the electron density and the side chain of Arg-236 is shifted by 2.5 Å towards the GI loop where it is situated at the position of
Asn-205 in the F1 model. Furthermore, hydrogen bonds between Arg-236 and Asp-233, and between Arg-236 and Val-204, lock the position of their side chains. In conclusion, the F2 model represents a substantial conformational change as compared to the F1 structure and this state of the protein seems to be incompatible with nucleotide binding. 3.5. The interface between the N and G domains The -helix bundle (helices A–D) is well deWned in the electron density of MmFtsY. Helix A is well-ordered, extended, and kinked in MmFtsY (Fig. 4), whereas it is absent in TaqFtsY and distorted in EcFtsY. The kink divides helix A into two sub-helices. Helix A1, residues Pro-97 to Met-102, is closely associated with the G domain and comprises what we believe is the C-terminal part of the A domain, whereas helix A2, residues Met-102 to Lys-118, is part of the N domain. Helix A1 packs against helix E from the N domain and two loops of the G domain located between strand 6 and helix E, and between strand 5 and helix 5. The N and G domains from the previously known structures of Ffh and FtsY are oriented diVerently with respect to each other. This Xexibility is intrinsic and the relative positions of the domains are regulated by the interaction of two highly conserved motifs. The Wrst of these, 135LIQTDM140, is homologous to the ‘ALLEADV’ motif in TaqFfh situated within the N domain, whereas the second motif, 358TSKGG362, is homologous to the ‘DARGG’ motif situated within the G domain (see Fig. 1B) (Egea et al., 2004; Focia et al., 2004). Movement of the N domain with respect to the G domain occurs during the formation of the Ffh–FtsY complex, and repositions several conserved hydrophobic residues in these motifs as well as residues situated in motif IV (Egea et al., 2004; Focia et al., 2004). An extensive interface between the N and G domains arises in MmFtsY due to the ordered structure of helices A1
T. Gariani et al. / Journal of Structural Biology 153 (2006) 85–96
91
Fig. 4. The Wgure visualizes the quality of the 2冷Fo冷¡冷Fc冷 electron-density map over helices A1 and A2. Residues are colored as in Fig. 1A. (For interpretation of the references to colour in this Wgure legend, the reader is referred to the web version of this paper.)
Fig. 5. Interactions formed between the N and G domains of MmFtsY due to the ordered structure of helix A. A more detailed view of part of the interactions is shown in the right panel. The contact interface made of residues from helices A1, A2, and E. The A1-helix is a novel structure so far only found in MmFtsY. This helix forms a water-mediated contact to the GIV domain via two water molecules, W8 and W201, linking Ala-101, Lys-104, and Ser-105 to the Met-355 and Asp-356 residues.
and A2. This interface was not observed in the previous structures as they are lacking helix A1. Two water molecules W8 and W201 bridge by hydrogen bonds residues Ala-201 to Ser-205 located in the short linker region connecting helices A1 and A2, to the main chain atoms of Met-355, Asp-356, and Thr-358 located in the GIV region (Fig. 5). Furthermore, Met-98, which is a valine in the native structure, is buried and
makes hydrophobic interactions with Ile-378, Val-380, Leu388, and Tyr-395. A second interface, which was also present in other structurally characterized SRP GTPases, consists of residues Ala-328, Gly-331, Ser-359, Lys-360, and Gly-362. These come in contact with residues Thr-138 and Asp-139, situated at the BC loop, either directly or via water mediated networks. Notably, the extensive interactions formed
92
T. Gariani et al. / Journal of Structural Biology 153 (2006) 85–96
between the N and G domains in MmFtsY make their overall conformation more similar to the one of FtsY present in the TaqFtsY–Ffh complex, than to the conformation of EcFtsY in its free form. 4. Discussion 4.1. A new apo state of FtsY GTP hydrolysis, catalyzed by the Ffh and FtsY proteins, is highly essential for the targeting process. Structures of nucleotide-loaded Ffh proteins alone (Freymann et al., 1999; Padmanabhan and Freymann, 2001), and nucleotideloaded Ffh–FtsY complexes (Egea et al., 2004; Focia et al., 2004) have allowed a better understanding of the hydrolysis step. Since the NG domains of Ffh and FtsY are very similar, one might suspect that the nucleotide uptake and hydrolysis would also be similar for the two proteins. However, biochemical studies show that FtsY, unlike Ffh, has poor speciWcity for GTP in solution (Shan and Walter, 2003). It has been proposed that speciWc binding of GTP to FtsY is correlated to two conformational states of the protein, referred to as the ‘open’ and ‘closed’ states (Shan and Walter, 2003). In the open state, structural Xexibility prevents proper nucleotide binding, which would explain the poor aYnity of FtsY for GTP. In the closed state, local structural rearrangements occur that trap the GTP nucleotide and allow Ffh-complex formation. We present here structural evidence for a new FtsY state that is present only when the protein is in its apo form. The new spatial arrangements seen in model F2 obliterate the nucleotide-binding pocket by removing the potential binding sites of the phosphate moiety of the nucleotide. The Ploop reshuZing is the most drastic diVerence between the F1 and F2 models. It is accompanied by a shift of the GIIIconsensus element and stabilization of the GII region. This new state could be likened to an ‘inert’ state, where the FtsY protein is prevented from binding to any nucleotide. The existence of such a nucleotide-free state, as a unique conformer, is not consistent with the basal GTP hydrolysis activity (Moser et al., 1997; Shan and Walter, 2003). FtsY might therefore shuttle between the ‘inert’ and ‘open’ states. This could explain, in part, the low aYnity and hydrolysis rate for its cognate substrate. The changes seen at the GII-consensus element, which result in its destabilization when shifting from the ‘inert’ state (F2 model) to an ‘open’ state (F1 model), are targeted in particular to residues Phe-235 and Arg-236. The same pattern is observed in the GII region of TaqFfh, where the apo structure displays signiWcantly lower B factors than the Mg2+ GDP-complex or the sulfate-loaded form of the protein. It has been suggested that the increased disorder of the GII loop is speciWc to the Mg2+ GDP-bound state (Freymann et al., 1999). Our observation, that binding of sulfate ions in TaqFfh and MmFtsY results in the same destabilization of the GII-consensus element, shows that the disorder is linked to the binding site at the GI loop. It seems
unlikely that the structural diVerences in the F1 and F2 models are due to crystal packing interactions. First, both of the independently reWned molecules in the asymmetric unit of the F2 models display the same structural change. Second, this change is not localized to one region of the protein but found at three of the consensus regions, GI, GII, and GIII. Finally, the two structural conformations are not novel and have been observed in apo- and ADPloaded forms of some ATPases, as will be discussed. 4.2. FtsY shares structural similarities to ATPases So far the conformational change at the nucleotidebinding site of MmFtsY has not been observed in other GTPases. We found, however, that these rearrangements are reminiscent of those occurring in some ATPases. This suggests that FtsY could be structurally more closely related to ATPases than previously believed. It also highlights an alternate nucleotide uptake. Interestingly, it has been reported that FtsY harbors a basal ATPase activity (Shan and Walter, 2003). The P-loop NTPases share a speciWc fold (Milner-White et al., 1991; Saraste et al., 1990; Walker et al., 1982). This fold is divided into seven major lineages belonging to two major groups (Bourne et al., 1991). One group includes the nucleotide kinases (e.g., adenosine kinases) and GTPases, where the strand leading to the P-loop and the GIII-element strand are direct neighbors. The other group, populated solely by ATPases (e.g., ABC ATPases such as MutS), is characterized by an additional strand inserted between those of the P-loop and the GIII-consensus element. Computerized searches of the SRP G domain against structures from the PDB identiWed predominantly ATPases as structurally related to the SRP GTPases (Montoya et al., 2000; Schlessman et al., 1998). This led to the division of GTPases into two classes. The SRP GTPases were grouped together with related ATPases in the SIMIBI class comprising the SRP, MinD/Mrp, and Bio-D related superfamilies, while translation factors and Ras GTPase were included in the TRAFAC class (Leipe et al., 2002). The distinction between these two new GTPase classes arises not only from an analysis of the three-dimensional structures but from the amino acid sequence as well. In addition, SIMIBI NTPases tend to form homodimers in, e.g., BioD, Nip, ArsA, or heterodimers as in the case of the SRP GTPases. This is in contrast to the TRAFAC GTPases, which typically function as monomers or in larger complexes with other proteins. For the ATPases MutS (Lamers et al., 2000) and adenosine kinase (Schumacher et al., 2000), the transition from an apo conformation to a nucleotide-loaded conformation is accompanied by a P-loop reorganization as seen in the F1 and F2 forms of MmFtsY (Fig. 6). Considering that some ATPases populate the SIMIBI class, this similarity seems less exceptional. However, MutS and adenosine kinase are not members of the same NTPase superfamily; they are both more remotely related to the SRP GTPases than the TRAFAC GTPases.
T. Gariani et al. / Journal of Structural Biology 153 (2006) 85–96
93
Fig. 6. Structural similarities between ATPase and MmFtsY in the GI region. (A) Ribbon C-trace over the homo-dimeric structure of MutS (PDB code 1E3M, (Lamers et al., 2000)). ADP is present in one monomer as indicated. The P1 loops in the apo- and the ADP-loaded monomers are shown in green and orange, respectively. (B) Superimposed structures of adenosine kinase. The apo structure is shown in magenta (PDB code 1LIO, Schumacher et al., 2000), and the complex with an ATP analogue is shown in yellow (PDB code 1LII, (Schumacher et al., 2000)). (C) Superposition of the P1 loops of model F1 (red) and model F2 (blue) of MmFtsY. (For interpretation of the references to colour in this Wgure legend, the reader is referred to the web version of this paper.)
There is evidence that a common ancestor of the members of the P-loop GTPase superclass was a protein with a NK£D motif (Leipe et al., 2002). The ATPases belonging to this superclass evolved either through the loss of the NK£D motif (kinesin–myosin) or through divergence in the region containing the NK£D motif (MinD and Bio-D-related family). It is therefore tempting to propose that the FtsY protein structurally bridges the SRP GTPases and the SIMIBIrelated ATPases. This could explain why FtsY has a basal ATPase activity while Ffh has not (Shan and Walter, 2003).
4.3. Signaling mechanism of the NG domain The mechanism by which the SRP GTPases regulate the SRP pathway is correlated to the architecture unique to this family of proteins. In addition to the two conserved N and G domains, the M domain in Ffh and the A domain in FtsY further complicate the picture. For proper function, the G domain must sense when FtsY is membrane-bound, which necessitates that a communication system between the G, N, and A domains has to be in place. Recent data
94
T. Gariani et al. / Journal of Structural Biology 153 (2006) 85–96
have shed light on the function of the various domains in the particle. The NG domain of SRP54 and Ffh is required for signal-sequence binding (Cleverley and Gierasch, 2002; Newitt and Bernstein, 1997) and increases the aYnity of SRP54 for SRP RNA (Diener and Wilson, 2000; Samuelsson and Olsson, 1993). Similarly, in FtsY the N domain is involved in binding to the membrane (Millman and Andrews, 1997, 1999), where it probably functions as a sensor of the GTP-dependent FtsY–Ffh complex formation (Egea et al., 2004; Focia et al., 2004; Ramirez et al., 2002). The A domain of FtsY is located immediately upstream of the N domain. Also the A domain is suggested to have a role in membrane anchoring and GTPase regulation (de Leeuw et al., 1997, 2000; Herskovits et al., 2001; Millman and Andrews, 1999; Millman et al., 2001; Powers and Walter, 1997; Zelazny et al., 1997). Intriguingly, the A domain varies considerably in size and sequence. In E. coli and M. mycoides the A domain comprises approximately 200 and 100 amino acids, respectively, whereas in other species it includes only a few amino acids (Rosenblad et al., 2003). The A domain seems to be particularly important for FtsY’s initial association with the membrane. Once it is bound to the membrane it can be proteolytically removed from the NG domain without eVects on FtsY function (Herskovits et al., 2001). The border between the A and N domain has so far not been structurally deWned. The structure of the NG domain of EcFtsY included residues 197–497; however, the Wrst four amino acids at the N-terminal end were not deWned in the electron density and consequently the A1 helix is missing in the EcFtsY structure (Montoya et al., 1997a,b). In a recent study, a number of deletion mutants in the N-terminal end of EcFtsY were tested for their ability to complement an FtsY-depleted E. coli strain (Eitan and Bibi, 2004). Interestingly, it was found that whereas the original NG domain construct (residues 197–497) lacked biological activity, the addition of only one more amino acid at its Nterminal end restored the protein’s activity. This suggests that only the uttermost C-terminal part of the A domain is essential for membrane assembly of the receptor (Eitan and Bibi, 2004). Thermotoga martima FtsY (TmFtY) represents a minimalistic FtsY protein with only a few amino acids constituting its A domain (Fig. 1B). Its structure has recently been determined (PDB accession code 1VMA, Joint Center for Structural Genomics, unpublished). In TmFtsY, the helices A1 and A2 have conformations that are identical to those in MmFtsY. Furthermore, two water molecules W5 and W13 in TmFtsY are identically located as compared to W8 and W201 in MmFtsY. In TmFtsY, the A1 helix is extended with one turn and comprises the entire N-terminal end of the protein starting with residue Met-1. The side chain of Phe-4 from TmFtsY forms extensive packing interactions with helix E situated at the C-terminal end of the protein. This amino acid corresponds to Phe-196 in EcFtsY. Phe-196 was not included in the fragment used for the structural study (Montoya et al., 1997a) but is known to be crucial for the biological activity of the protein (Eitan and Bibi, 2004). It
Fig. 7. Positively charged residues and hydrophobic residues located at the polar interface of the amphiphatic helix A2.
seems plausible that an extended construct of EcFtsY could stabilize this region of the protein as a deWned -helical structure A1, similar to what is seen in MmFtsY and TmFtsY. We suggest that there is a communication between the N and G domains at the interface between helices A1 and A2 and the G domain. With helix A1 properly folded, helix A2 is highly ordered in both MmFtsY and TmFtsY in such a way that a cluster of lysine and hydrophobic residues are located on the same surface-exposed side of helix A2 (Fig. 7). These residues seem to be in an ideal position for membrane anchoring, for example through lysine snorkeling (Mishra and Palgunachari, 1996; Segrest et al., 1990) to the negatively charged phospholipid head groups of the membrane (de Leeuw et al., 2000; Millman et al., 2001). Helix A1 and A2 could therefore constitute a linker between membrane association and the GIV-consensus element in the G domain, which senses the conformational change that takes place when the protein is anchored to the membrane and primes the protein for hydrolysis (Millman and Andrews, 1999). In particular, the conserved water-mediated hydrogen interactions formed between Asp-356 and residues from helix A1, could be crucial in mediating such a communication. Acknowledgments We thank Terese Bergfors for valuable discussions and critical reading of the manuscript. This study was
T. Gariani et al. / Journal of Structural Biology 153 (2006) 85–96
supported by the Swedish Research Council (K5104-1098) and European Union (CLK3-CT-2000-00082). References Bourne, H.R., Sanders, D.A., McCormick, F., 1991. The GTPase superfamily: conserved structure and molecular mechanism. Nature 349, 117–127. Brünger, A.T., Adams, P.D., Clore, G.M., DeLano, W.L., Gros, P., Grosse-Kunstleve, R.W., Jiang, J.S., Kuszewski, J., Nilges, M., Pannu, N.S., Read, R.J., Rice, L.M., Simonson, T., Warren, G.L., 1998. Crystallography & NMR system: a new software suite for macromolecular structure determination. Acta Crystallogr. D Biol. Crystallogr. 54, 905–921. Cleverley, R.M., Gierasch, L.M., 2002. Mapping the signal sequence-binding site on SRP reveals a signiWcant role for the NG domain. J. Biol. Chem. 277, 46763–46768. Collaborative Computational Project, No. 4, 1994. The CCP4 suite: programs for protein crystallography. Acta Crystallogr. D 50, 760–763. Connolly, T., Gilmore, R., 1989. The signal recognition particle receptor mediates the GTP-dependent displacement of SRP from the signal sequence of the nascent polypeptide. Cell 57, 599–610. de Leeuw, E., Poland, D., Mol, O., Sinning, I., ten Hagen-Jongman, C.M., Oudega, B., Luirink, J., 1997. Membrane association of FtsY, the E. coli SRP receptor. FEBS Lett. 416, 225–229. de Leeuw, E., te Kaat, K., Moser, C., Menestrina, G., Demel, R., de KruijV, B., Oudega, B., Luirink, J., Sinning, I., 2000. Anionic phospholipids are involved in membrane association of FtsY and stimulate its GTPase activity. EMBO J. 19, 531–541. Diener, J.L., Wilson, C., 2000. Role of SRP19 in assembly of the Archaeoglobus fulgidus signal recognition particle. Biochemistry 39, 12862– 12874. Doudna, J.A., Batey, R.T., 2004. Structural insights into the signal recognition particle. Annu. Rev. Biochem. 73, 539–557. Egea, P.F., Shan, S.O., Napetschnig, J., Savage, D.F., Walter, P., Stroud, R.M., 2004. Substrate twinning activates the signal recognition particle and its receptor. Nature 427, 215–221. Eitan, A., Bibi, E., 2004. The core Escherichia coli signal recognition particle receptor contains only the N and G domains of FtsY. J. Bacteriol. 186, 2492–2494. Focia, P.J., Shepotinovskaya, I.V., Seidler, J.A., Freymann, D.M., 2004. Heterodimeric GTPase core of the SRP targeting complex. Science 303, 373–377. Freymann, D.M., Keenan, R.J., Stroud, R.M., Walter, P., 1997. Structure of the conserved GTPase domain of the signal recognition particle. Nature 385, 361–364. Freymann, D.M., Keenan, R.J., Stroud, R.M., Walter, P., 1999. Functional changes in the structure of the SRP GTPase on binding GDP and Mg2+GDP. Nat. Struct. Biol. 6, 793–801. Fulga, T.A., Sinning, I., Dobberstein, B., Pool, M.R., 2001. SRbeta coordinates signal sequence release from SRP with ribosome binding to the translocon. EMBO J. 20, 2338–2347. Gariani, T., Sauer-Eriksson, E., 2000. Crystallization and preliminary Xray diVraction studies of the signal recognition particle receptor FtsY from Mycoplasma mycoides. Acta Crystallogr. D Biol. Crystallogr. 56, 1030–1032. Gilmore, R., Blobel, G., Walter, P., 1982. Protein translocation across the endoplasmic reticulum. I. Detection in the microsomal membrane of a receptor for the signal recognition particle. J. Cell Biol. 95, 463–469. Halic, M., Becker, T., Pool, M.R., Spahn, C.M., Grassucci, R.A., Frank, J., Beckmann, R., 2004. Structure of the signal recognition particle interacting with the elongation-arrested ribosome. Nature 427, 808–814. Harris, M., Jones, T.A., 2001. Molray—a web interface between O and the POV-Ray ray tracer. Acta Crystallogr. D Biol. Crystallogr. 57, 1201– 1203. Herskovits, A.A., Seluanov, A., Rajsbaum, R., ten Hagen-Jongman, C.M., Henrichs, T., Bochkareva, E.S., Phillips, G.J., Probst, F.J., Nakae, T.,
95
Ehrmann, M., Luirink, J., Bibi, E., 2001. Evidence for coupling of membrane targeting and function of the signal recognition particle (SRP) receptor FtsY. EMBO Rep. 2, 1040–1046. Jones, T.A., Zou, J.Y., Cowan, S.W., Kjeldgaard, M., 1991. Improved methods for binding protein models in electron density maps and the location of errors in these models. Acta Crystallogr. A 47, 110–119. Keenan, R.J., Freymann, D.M., Stroud, R.M., Walter, P., 2001. The signal recognition particle. Annu. Rev. Biochem. 70, 755–775. Keenan, R.J., Freymann, D.M., Walter, P., Stroud, R.M., 1998. Crystal structure of the signal sequence binding subunit of the signal recognition particle. Cell 94, 181–191. Kjeldgaard, M., Nyborg, J., 1992. ReWned structure of elongation factor EF–Tu from Escherichia coli. J. Mol. Biol. 223, 721–742. Kraulis, P.J., 1991. MOLSCRIPT: A program to produce both detailed and schematic plots of protein structures. J. Appl. Crystallogr. 24, 946–950. Lamers, M.H., Perrakis, A., Enzlin, J.H., Winterwerp, H.H., de Wind, N., Sixma, T.K., 2000. The crystal structure of DNA mismatch repair protein MutS binding to a G £ T mismatch. Nature 407, 711–717. Leipe, D.D., Wolf, Y.I., Koonin, E.V., Aravind, L., 2002. ClassiWcation and evolution of P-loop GTPases and related ATPases. J. Mol. Biol. 317, 41–72. Luirink, J., Dobberstein, B., 1994. Mammalian and Escherichia coli signal recognition particles. Mol. Microbiol. 11, 9–13. Luirink, J., Sinning, I., 2004. SRP-mediated protein targeting: structure and function revisited. Biochim. Biophys. Acta 1694, 17–35. Luirink, J., ten Hagen-Jongman, C.M., van der Weijden, C.C., Oudega, B., High, S., Dobberstein, B., Kusters, R., 1994. An alternative protein targeting pathway in Escherichia coli: studies on the role of FtsY. EMBO J. 13, 2289–2296. Lütcke, H., High, S., Romisch, K., Ashford, A.J., Dobberstein, B., 1992. The methionine-rich domain of the 54 kDa subunit of signal recognition particle is suYcient for the interaction with signal sequences. EMBO J. 11, 1543–1551. Macao, B., Luirink, J., Samuelsson, T., 1997. Ffh and FtsY in a Mycoplasma mycoides signal-recognition particle pathway: SRP RNA and M domain of Ffh are not required for stimulation of GTPase activity in vitro. Mol. Microbiol. 24 (3), 523–534. Meyer, D.I., Krause, E., Dobberstein, B., 1982. Secretory protein translocation across membranes-the role of the ‘docking protein’. Nature 297, 647–650. Miller, J.D., Bernstein, H.D., Walter, P., 1994. Interaction of E. coli Ffh/ 4.5S ribonucleoprotein and FtsY mimics that of mammalian signal recognition particle and its receptor. Nature 367, 657–659. Miller, J.D., Wilhelm, H., Gierasch, L., Gilmore, R., Walter, P., 1993. GTP binding and hydrolysis by the signal recognition particle during initiation of protein translocation. Nature 366, 351–354. Millman, J.S., Andrews, D.W., 1997. Switching the model: a concerted mechanism for GTPases in protein targeting. Cell 89, 673–676. Millman, J.S., Andrews, D.W., 1999. A site-speciWc, membrane-dependent cleavage event deWnes the membrane binding domain of FtsY. J. Biol. Chem. 274, 33227–33234. Millman, J.S., Qi, H.Y., Vulcu, F., Bernstein, H.D., Andrews, D.W., 2001. FtsY binds to the Escherichia coli inner membrane via interactions with phosphatidylethanolamine and membrane proteins. J. Biol. Chem. 276, 25982–25989. Milner-White, E.J., Coggins, J.R., Anton, I.A., 1991. Evidence for an ancestral core structure in nucleotide-binding proteins with the type A motif. J. Mol. Biol. 221, 751–754. Mishra, V.K., Palgunachari, M.N., 1996. Interaction of model class A1, class A2, and class Y amphipathic helical peptides with membranes. Biochemistry 35, 11210–11220. Montoya, G., Kaat, K., Moll, R., Schafer, G., Sinning, I., 2000. The crystal structure of the conserved GTPase of SRP54 from the archaeon Acidianus ambivalens and its comparison with related structures suggests a model for the SRP–SRP receptor complex. Struct. Fold. Des. 8, 515–525. Montoya, G., Svensson, C., Luirink, J., Sinning, I., 1997a. Crystal structure of the NG domain from the signal-recognition particle receptor FtsY. Nature 385, 365–368.
96
T. Gariani et al. / Journal of Structural Biology 153 (2006) 85–96
Montoya, G., Svensson, C., Luirink, J., Sinning, I., 1997b. Expression, crystallization, and preliminary X-ray diVraction study of FtsY, the docking protein of the signal recognition particle of E. coli. Proteins 28, 285–288. Moser, C., Mol, O., Goody, R.S., Sinning, I., 1997. The signal recognition particle receptor of Escherichia coli (FtsY) has a nucleotide exchange factor built into the GTPase domain. Proc. Natl. Acad. Sci. USA 94, 11339–13344. Navaza, J., 1994. AMoRe: an automated package for molecular replacement. Acta Crystallogr. A 50, 157–163. Newitt, J.A., Bernstein, H.D., 1997. The N-domain of the signal recognition particle 54-kDa subunit promotes eYcient signal sequence binding. Eur. J. Biochem. 245, 720–729. Otwinowski, Z., Minor, W., 1997. Processing of X-ray diVraction data collected in oscillation mode. Methods Enzymol. 276, 307–326. Padmanabhan, S., Freymann, D.M., 2001. The conformation of bound GMPPNP suggests a mechanism for gating the active site of the SRP GTPase. Structure (Camb) 9, 859–867. Pool, M.R., Stumm, J., Fulga, T.A., Sinning, I., Dobberstein, B., 2002. Distinct modes of signal recognition particle interaction with the ribosome. Science 297, 1345–1348. Powers, T., Walter, P., 1995. Reciprocal stimulation of GTP hydrolysis by two directly interacting GTPases. Science 269, 1422–1424. Powers, T., Walter, P., 1997. Co-translational protein targeting catalyzed by the Escherichia coli signal recognition particle and its receptor. EMBO J. 16, 4880–4886. Ramirez, U.D., Minasov, G., Focia, P.J., Stroud, R.M., Walter, P., Kuhn, P., Freymann, D.M., 2002. Structural basis for mobility in the 1.1 A crystal structure of the NG domain of Thermus aquaticus Ffh. J. Mol. Biol. 320, 783–799. Rapiejko, P.J., Gilmore, R., 1997. Empty site forms of the SRP54 and SR alpha GTPases mediate targeting of ribosome-nascent chain complexes to the endoplasmic reticulum. Cell 89, 703–713. Rinke-Appel, J., Osswald, M., von Knoblauch, K., Mueller, F., Brimacombe, R., Sergiev, P., Avdeeva, O., Bogdanov, A., Dontsova, O., 2002. Crosslinking of 4.5S RNA to the Escherichia coli ribosome in the presence or absence of the protein Ffh. RNA 8, 612–625. Rosenblad, M.A., Gorodkin, J., Knudsen, B., Zwieb, C., Samuelsson, T., 2003. SRPDB: signal recognition particle database. Nucleic Acids Res. 31, 363–364. Samuelsson, T., 1992. A Mycoplasma protein homologous to mammalian SRP54 recognizes a highly conserved domain of SRP RNA. Nucleic Acids Res. 20, 5763–5770. Samuelsson, T., Olsson, M., 1993. GTPase activity of a bacterial SRP-like complex. Nucleic Acids Res. 21, 847–853.
Saraste, M., Sibbald, P.R., Wittinghofer, A., 1990. The P-loop–a common motif in ATP- and GTP-binding proteins. Trends Biochem. Sci. 15, 430–434. ScheVzek, K., Klebe, C., Fritz-Wolf, K., Kabsch, W., Wittinghofer, A., 1995. Crystal structure of the nuclear Ras-related protein Ran in its GDP- bound form. Nature 374, 378–381. Schlessman, J.L., Woo, D., Joshua-Tor, L., Howard, J.B., Rees, D.C., 1998. Conformational variability in structures of the nitrogenase iron proteins from Azotobacter vinelandii and Clostridium pasteurianum. J. Mol. Biol. 280, 669–685. Schlichting, I., Almo, S.C., Rapp, G., Wilson, K., Petratos, K., Lentfer, A., Wittinghofer, A., Kabsch, W., Pai, E.F., Petsko, G.A., et al., 1990. Time-resolved X-ray crystallographic study of the conformational change in Ha-Ras p21 protein on GTP hydrolysis. Nature 345, 309– 315. Schumacher, M.A., Scott, D.M., Mathews, I.I., Ealick, S.E., Roos, D.S., Ullman, B., Brennan, R.G., 2000. Crystal structures of Toxoplasma gondii adenosine kinase reveal a novel catalytic mechanism and prodrug binding. J. Mol. Biol. 298, 875–893. Segrest, J.P., De Loof, H., Dohlman, J.G., Brouillette, C.G., Anantharamaiah, G.M., 1990. Amphipathic helix motif: classes and properties. Proteins 8, 103–117. Shan, S.O., Walter, P., 2003. Induced nucleotide speciWcity in a GTPase. Proc. Natl. Acad. Sci. USA 100, 4480–4485. Sprang, S.R., 1997. G protein mechanisms: insights from structural analysis. Annu. Rev. Biochem. 66, 639–678. Ulbrandt, N.D., Newitt, J.A., Bernstein, H.D., 1997. The E. coli signal recognition particle is required for the insertion of a subset of inner membrane proteins. Cell 88, 187–196. Walker, J.E., Saraste, M., Runswick, M.J., Gay, N.J., 1982. Distantly related sequences in the alpha- and beta-subunits of ATP synthase, myosin, kinases and other ATP-requiring enzymes and a common nucleotide binding fold. EMBO J. 1, 945–951. Wild, K., Halic, M., Sinning, I., Beckmann, R., 2004. SRP meets the ribosome. Nat. Struct. Mol. Biol. 11, 1049–1053. Wittinghofer, F., Krengel, U., John, J., Kabsch, W., Pai, E.F., 1991. Threedimensional structure of p21 in the active conformation and analysis of an oncogenic mutant. Environ. Health Perspect. 93, 11–15. Zelazny, A., Seluanov, A., Cooper, A., Bibi, E., 1997. The NG domain of the prokaryotic signal recognition particle receptor, FtsY, is fully functional when fused to an unrelated integral membrane polypeptide. Proc. Natl. Acad. Sci. USA 94, 6025–6029. Zopf, D., Bernstein, H.D., Walter, P., 1993. GTPase domain of the 54-kD subunit of the mammalian signal recognition particle is required for protein translocation but not for signal sequence binding. J. Cell Biol. 120, 1113–1121.