Compilation of some literature data:

Results of Biochemical Mutational Studies

SRP SR

Both Ffh and FtsY are GTP-binding proteins, but neither has significant GTP binding activity by itself (Poritz, Science 1990). Bacterial Ffh and FtsY function as reciprocal GTP-ase activating proteins for each other.

SRP54 (human) hydrolyzes GTP upon interaction with the receptor and so the receptor acts as a GNAP and GTPase activating protein. The ribosome acts as a GTP-loading factor for SRP54. But binding of a functional signal peptide to SRP54 prevents GTP binding (Bacher, Nature, 1996). The signal sequence stabilizes the nucleotide-free form of SRP54. The SRP54 bound to the signal sequence binds GTP and then to the SR. The binding of the SRP to the receptor causes conformational changes both and stimulates GTP binding by both the SRP and the receptor. GTP binding to SRP54 reduces its affinity for the bound signal sequence. Upon GTP binding to SRP54, the SRP54-SR complex dissociates from the translocation apparatus as a stable complex. The SR receptor stimulates GTP hydrolysis by the SRP54 and the SRP dissociates from the receptor and the membrane. SRb is a transmembrane GTPase that anchors SRa to the ER membrane (Miller, JCB, 1995).

(Legate, JBC, 2000) In SRb, after gel-filtration chromatography, a D181N mutant no longer dimerized with SRa as this mutant has reduced GTP binding affinity. SRa mutants G118L and H119L reduce GTP binding ability.

SR? is most like the GTPase Sar1p, an ARF-like GTPase involved in ER to Golgi-trafficking.

(Ogg, J. Cell Biol. 1998) SRb homologue SRP102p in S. cerevisiae anchors SRP101p, the SRa homologue, to the ER membrane but a mutant form of SRP102p lacking the transmembrane domain is functional in yeast, indicating that SRP102p is not solely required to tether SRP101p to the membrane. Mutations that disrupt the binding GTP-site of SRP102p inactivate the signal receptor. Most of these mutants also disrupt the interaction between SRP101p and SRP102p, suggesting that an interaction of the 2 proteins may be required to activate SRP101p as a prerequisite for a productive interaction with SRP during protein targeting. Cells without SRP102p exhibit a growth defect. SRP101p and SRP102p form a stable complex in the ER. In ultracentrifugation experiments, SRP101p was found associated with membranes. Deletion of SRP102p resulted in the release of the majority of SRP101p into the cytosol, though a significant fraction of SRP101p still co-fractionated with membranes from cells in which SRP102p was deleted. SRP102p (majority) was always found in membrane fractions. SRP101p must interact either directly or indirectly with membranes. (Good table on pg. 350. ) Many site directed mutants have been made of S. cerevisiae SRb homologue SRP102p. S49A in G1 reduced GTP hydrolysis. K51I is a G1 temperature-sensitive mutant but the temperature was found not to destabilize it but is found to have reduced nucleotide affinity. It also severely alters the interactions between SRP101p and SRP102p. T52N in G1 reduced the affinity for GTP but increased the affinity for GEF. T66A in G2 prevents a GTP-dependent interaction with the GAP. G90L in G3 produces an unknown effect. H91L in G3 reduces GTPase activity. N154I in G4 impairs nucleotide exchange. E157Q in G4 reduces the affinity for GTP. S220A in G5 is not expressed well and bypasses the requirement for GEF. It was found that the deletion of the transmembrane domain of SRP102 does not perturb its function. Without the TMD portion, though, 75% of the protein fractionates with the cytosol and not the membrane.

Only the ribosome-bound SRP has affinity for the signal sequences of the nascent polypeptide chain.

(Bernstein, Nature, 1989; also reviewed in Rapoport, FASEB, 1991) Ffh M domain is rich in Met (11(SRP54)-13(Ffh)%) and basic amino acids such as Ile and Leu. Met residues with their flexible side chains are able to accommodate hydrophobic ligands with a different spatial structure. The M domain forms a small globular structure with 2 a-helices in Ffh and 3 in SRP54 with the Met residues clustered on one face of the helices. It features a large hydrophobic pocket flanked by a flexible loop lined with Met side chains. The helices are arranged in a helix-turn-helix motif. The M domain recognizes the signal sequence, which may bind by hydrophobic interactions. The signal sequences of the nascent proteins as characterized by a stretch of at least six consecutive hydrophobic amino acids at their N-terminus.

The M domain also interacts with the 4.5S RNA and a continuous surface is formed. In the symmetric loop of the RNA, the minor groove disappears and a distinct flat face on the RNA surface is created (both when bound or free of the protein). The asymmetric loop is disordered in solution and confers conformational flexibility to that region of the RNA. Binding the RNA does not induce much conformational change to the protein in comparison to the protein alone. In the RNA, Cytosine62 flips from an exposed position to a buried one in the protein-RNA interface of the complex (Walter, Science, 2000). The helix-loop-helix motif of Ffh contacts the RNA in a different manner than observed for the classical helix-turn-helix proteins that bind DNA. In the case of the SRP, the HTH fold interacts with the RNA surface through many contacts with the protein backbone. In the crystal structure of the Ffh-RNA complex, the proposed signal sequence binding pocket is disordered, yet the binding site can be inferred. The hydrophobic portion of the signal sequence binding pocket extends to a ledge of RNA backbone formed by the helical segment between the symmetric loop and the terminal tetraloop of domain IV. The nascent protein is hypothesized to bind both Ffh and the RNA with the hydrophobic portion of the signal sequence coming into contact with the Met rich region of the M domain and the positively charged residues (that usually precede the hydrophobics) contacting the RNA directly.

(Kurita, JBC, 1996) In Bacillus subtilis, residues 364-432 in the N-terminal region of Ffh were shown to bind RNA. This sequence contains two hydrophobic regions (h2, residues 364-391, and h3, 416-435)(but NOT h1, residues 307-331), separated by the positively charged amino acid motif, ³⁹⁸RRKRIAKGSG⁴⁰⁷. Among the basic amino acid residues in this region, Arg401 was essential for binding to the RNA, but Arg399 and Lys400 were not. The co-existence of Arg398 and Lys404 was necessary for the same affinity as wild type Ffh. The two glycine residues of the ⁴⁰⁵GSG⁴⁰⁷ were also essential. Arg401, Gly405, and Gly407 all probably bind directly to the RNA. MH23 peptide (91 amino acids) encompassing from 356 to 446, consisting of h2-RRKRIAKGSG-h3, bound RNA with the same affinity as wild type Ffh, whereas a 24-amino acid synthetic peptide ³⁹²DIINASRRKRIAKGSGTSVQEVNR⁴¹⁵ did not. The region containing two hydrophobic segments separated by the positively charged motif is the minimal requirement of Ffh for RNA binding. h2 and h3 probably do not bind to the RNA directly, but a secondary or tertiary structure formed by h2 and h3 is involved in RNA binding.

(Rapoport, FASEB, 1991) Membrane proteins characteristically have membrane-spanning segments of about 20 uncharged, mostly hydrophobic amino acid residues. The most amino-terminal segment is often an uncleaved sequence that simultaneously serves as a membrane anchor.

Not all proteins are transported by the SRP system, rather the bacterial SRP is a specialized targeting mechanism for only the more hydrophobic signal sequences of E. coli proteins, such as those present in membrane proteins. Only these efficiently cross-link to Ffh in vitro. The chloroplast SRP transports proteins into the thylakoid membranes.

(Wood, Nucleic Acids Research, 1992) Nucleotides in the 4.5S RNA have been mutated. The mutants include a deletion of 12 nucleotides between 49 and 60, C23U, U37A, A39C, C40G, A47C, G48U, G49C, C52G, G53A, G53C, G54A, A56U, G57U, A60U, G61U, C62G, and A63C. C52G DISRUPTS THE BASE PAIRING WITH G57. A double mutation of C52G and G57C restores base pairing ability. C23U is located in a region conserved in eubacterial 4.5S RNA homologues. A47C, G48U, G53A, G53C , A56U, G61U, and C62G are invariant across all known 7S-like RNAs. The ability of the mutant RNAs to bind Ffh in vivo was determined experimentally. Deletion of the 12 nucleotides in a tetranucleotide loop abolishes RNA binding to Ffh. However point mutations in this region (including at nucleotides 52, 53, 54, 56, 57, and 60) do not abolish Ffh binding. Point mutations at 39, 47, 48, 49, 61, and 62 also completely abolish binding. Ffh binding may be sequence specific. In vivo, the 12 nucleotide deletion and the point mutant C62G do not support bacterial growth under any conditions. The other mutations produce growth defects only when the bacteria were grown in very restrictive conditions. A39C and C40G have no effect on colony size, nor does C23U.

(Wild, Science, 2001) The crystal structure of human SRP with a helix 6 29-nt fragment of RNA (homologous to the E. coli helix 8/Ffh M domain complex) has led the Sinning group to propose a model for the assembly of the SRP core. SRP19 sits on the tip of helix 6 with the b1 edge of the b-sheet contacting the GGAG tetraloop. SRP19 contacts the tip of helix 6 directly or with one intervening water molecule to the backbone of nucleotides C140 to C143 and G147 to G151. A149 of the GGAG tetraloop is essential for helix 6 binding to SRP19 and is universally conserved in all SRP RNAs. G147 is locked into place by stacking with the guanidinium group of Arg101, which itself is locked between phosphate groups of G148 and A149. Three conserved Tyr residues, Tyr19, Tyr22, and Tyr68, form a triangular clamp and point toward the phosphate group of G150 contributing to several direct and water-mediated hydrogen bonds and reaching to the phosphate groups of A149 and C151. The guanidinium group of Arg33 is interlocked with phosphate groups of C141 and U142 and is H-bonded to main chain carbonyls of Lys27 and Ile35. Arg34 on the other side of the groove is involved with specific, but indirect, recognition of the first tetraloop nucleotide G147 and forms water-mediated H-bonds with the G150 and C151 phosphates.

(Herkovits, PNAS, 2000) Membrane targeting of E. coli ribosomes requires FtsY in vivo. Depletion of Ffh has no significant effect on binding of ribosomes to the membrane, although Ffh depletion is detrimental to growth. On depletion of FtsY, membrane-associated ribosomes are reduced by approximately 75%. Ffh depletion prevents proper membrane protein assembly, but has little or no effect on expression.

(Fulga, EMBO J, 2001) SRb (human) is essential for protein translocation across the ER membrane. SRb can be crosslinked to a 21 kDa ribosomal protein in its empty and GDP-bound state, but not when GTP is bound. GTP binding to SRb is required to induce signal sequence release from SRP. The presence of the translocon changes the interaction between the 21 kDa protein and SRb and allows SRb to bind GTP. SRb coordinates the release of the signal sequence from SRP with the presence of the translocon. At the membrane, SRP contacts the SR. Contact between SRP54 and SRa leads to the transfer of the nascent chain from SRP54 into the translocation channel formed by the Sec61p complex. In the presence of the ribosome, SRP54 has an increased affinity for GTP. Upon contact with SRa, both proteins bind GTP with a high affinity and forma stable complex. Mutual hydrolysis in SRa and SRP54 lead to the dissociation of SRP from SR. Binding of GTP by SRb is required for complex formation with SRa. GTP binding to SRb is modulated by an interaction with 21 kDa protein of the large ribosomal subunit, in a manner that is sensitive to the presence of the Sec61p complex.

(Groves, JBC, 2001) The chloroplast SRP (cpSRP) has no known RNA component and is present in two parts in the chloroplast: a co-translationally active SRP54 homologue (cpSRP54), which associates with the chloroplast ribosome/nascent chain complex and a pot-translationally active cpSRP, which has been shown to contain cpSRP54 and a novel cpSRP43. The chloroplast SRP receptor is cpFtsY. The C terminus of cpSRP54 was found to be essential for the formation of a stable cpSRP complex and cpSRP43 interacts with distinct regions of the M domain of cpSRP54.

(de Leeuw, EMBO, 2000) FtsY interacts directly with E. coli phospholipids, with a preference for anionic phospholipids. FtsY is known to associate with SRP-bound ribosome-nascent chain complexes in the absence of membranes. The interaction involves at least two lipid-binding sites, one of which is present in the NG-domain. Lipid association induces a conformational change in FtsY and greatly enhances its GTPase activity. No hydrophobic helix has been found which would indicate FtsY is a transmembrane protein, however, FtsY is believed to associate with the membrane. FtsY may associate with the membrane via a saturable protein-protein interaction, in which the A-domain would be primarily involved or it may associate via a direct protein-lipid association. FtsY is a strongly negatively charged protein with ~50 net negative charges at physiological pH so that an electrostatic repulsion between the negative membrane surface and the protein, especially in the A-domain, is anticipated. However, an enhanced interaction of FtsY with negatively charged lipids as compared with zwitterionic lipids in monolayer and aggregation studies. This suggests an interaction based on electrostatic attraction. Insertion of FtsY occurs independently of nucleotide. The A-domain may influence the conformation of the NG-domain, facilitating insertion in the GDP or GTP bound forms or the A-domain may interact with the DOPG monolayer without being influenced by nucleotide binding to the NG-domain. Liposomes were found to stabilize FtsY in the GTP-bound conformation. In solution, though, GTPase activity of FtsY is repressed by the acidic domain. Their data indicate that the A-domain both senses the membrane and releases the block in GTP hydrolysis and stimulates the hydrolysis over the level of membrane-bound FtsY. Though the A-domain is negatively-charged, there are clusters of R/K residues that may interact with the membrane anionic phospholipids. Studies with different negatively-charged lipids of 10-90% indicate that not only the charge, but the nature of the headgroup, is important to promote GTP hydrolysis. The most efficient insertion is with anionic lipids at moderate salt concentration. FtsY is monomeric. The N-terminus of the protein is rich in positively-charged residues. Lipid attachment is also accompanied by a partial unfolding of the protein and insertion into monolayers.

(Stroud, CurrOpinMolBio, 1999) Signal sequences are not conserved in sequence. They generally begin with 10 amino acids of the N-terminus, are between 20-30 amino acids in length, characterized by a hydrophobic core of approximately 10-15 (but no less than 6) residues, and have a preference for leucines and alanines. They are flanked on the N-terminal side by a positively charged stretch of polar residues and by a neutral, but polar, C-terminal region. They are very tolerant of amino acid substitutions as long as the central hydrophobic core is preserved. The M-domain is formed from 4 helices around a conserved hydrophobic core. There is a deep hydrophobic groove, 25 A long, 15 A wide and 12 A deep lined by 21 hydrophobic sidechains from helices aM1, aM2, and aM4. The hydrophobic surface area of this site is 1487A² and the dimensions of the hydrophobic groove are sufficient to accommodate 17 amino acids if they are in an a-helical conformation or 14 amino acids in a fully-extended b-hairpin conformation. The M-domain of the SRP methionines act as flexible bristles that could interact with a variety of hydrophobic sidechains. Of the 14 well-conserved Met sites in the signal binding domain, 11 line the hydrophobic groove, with the majority mapping onto the hydrophobic faces of a-helices. As FtsY and Ffh form a complex together, they activate both GTPases. The RNA joins the M-domain and the GTPase domain of Ffh and orders its interaction with FtsY. The M-domain interacts with the RNA through the conserved, positively charged motif ³⁸³SRRKRIAKGSG³⁹³.

(Peluso, ???Journal, 2001) The 4.5S RNA does not affect the basal GTPase activity of Ffh or FtsY but enhances the GTPase activity of the complex. The RNA also enhances complex formation of Ffh and FtsY when GTP is bound to both proteins 400-fold. GTP hydrolysis from GTP*Ffh-FtsY*GTP was found to be faster than the dissociation of the complex. 4.5S RNA binds to the M-domain of Ffh, also the binding region for signal sequences. Ffh-FtsY association is rate-limiting- some measured rates are given (not particularly fast).

(Herskovits, EMBO Reports, 2001) FtsY membrane targeting was investigated. The NG domain alone is not properly targeted when expressed alone without the A-domain, however, it is functional when fused to polypeptides derived from the protein LacY, the lactose permease of E. coli that is co-translationally targeted to the membrane. FtsY was altered to study co-translational membrane targeting by inserting polypeptide spacers between A and NG, co-expressing separated A and NG domains, and inserting a specific protease recognition site between A and NG and proteolyzing in vivo. If the NG domain of FtsY assembles co-translationally on its membrane target, NG may be translated immediately after the FtsY-RNC exposing the A domain has reached the membrane. Hydrophilic spacers between the A and NG domains may cause a delay in translation of NG , may localize NG a distance from the membrane, and thus prevent proper co-translational assembly. The long linkers do not abolish FtsY function, but the distance between NG and its targeting domain is also important. Complete detachment of NG from A does not abolish function. Covalent contact between NG and A is not required for proper function. Intact FtsY is only required during translation and targeting. FtsY is targeted to the membrane during translation and this targeting is required for its biological activity. The A domain is required for proper targeting.

(de Leeuw, FEBS Lett.,1997) FtsY is an unusual protein in that it is located both in the cytosol and in the inner membrane whereas it is highly charged and does not contain any obvious membrane spanning sequences. Both domains (A and NG) of FtsY associate with the membrane, but the nature of the association differs. The structural domains of FtsY are not functional when expressed as a separate entity. In vivo, FtsY-NG was found almost exclusively in the membrane fraction, FtsY-G was found a little in the membrane fraction, and FtsY-A was found mostly in the soluble fraction but a little in the membrane pellet. The nature of the association of FtsY-G and FtsY-NG differs from that of FtsY-A as there were more resistant to extraction with urea. Experiments showed that the native FtsY was able to associate efficiently with the inner membrane in vitro, but that only 1% of FtsY-A, 16% of FtsY-NG and 12% of FtsY-G were found associated with membranes. Apparently, the protein fragments may tend to aggregate in vitro but not in vivo. The A domain has pI of 3.9. The surface of the N domain is also negatively charged. In E. coli, the FtsY NG domain alone is not sufficient to function as a receptor for the SRP, however several other bacterial homologues consist of only the NG domain.

(Kusters, FEBS Lett., 1995) Mutations of E. coli FtsY were made at Thr446Asn and Asp449Ala. Both mutants have reduced GTP binding capacity, the D449A mutation reduced GTP binding by 83% while the T446N mutation reduced it by 45%. Both also showed a reduced GTPase activity. Limited proteolysis in vitro by proteinase K was used to study whether GTP affects the conformation of FtsY and the mutants. Wild-type FtsY was readily degraded by proteinase K. FtsY-GTP was not as degraded and an intermediate product of 33 kDa was formed. FtsY-T446N-GTP was mostly degraded, even the 33 kDa product. FtsY-D449A-GTP was totally degraded even in the presence of GTP. In the presence of GDP, FtsY is still mostly degraded, while FtsY-T446N and FtsY-D449A are not. The mutant proteins seem to bind GDP with high affinity of with a decreased ability to release GDP. GMP did not change the degradation pattern of any of the proteins, while that of the GTP analogue GMP-PNP resembled the effect of GTP. The intermediate degradation product was FtsY missing 187 N-terminal residues. The protected C-terminal fragment corresponded with the X- (IBD?) and G-domains. These domains seem to fold in a packed protease-resistant conformation upon GTP-binding with the less rigid N-terminal part protruding. The FtsY-T446N mutant in the presence of GTP formed a complex with Ffh and the RNA at reduced efficiency as compared to the wild-type FtsY, while the FtsY-D449A-GTP mutant was unable to interact with the SRP and form a complex. GTP binding by FtsY is essential for complex formation with the SRP.

(Jagath, JMB, 2000) The fluorescence at E. coli FtsY Trp343 (I box mutant where the other two Trp residues at positions 12 and 128 had been replaced by Phe residues) was monitored to study the biding of GDP/GTP to FtsY and the FtsY-SRP complex formation. Upon SRP-FtsY complex formation in the presence of GTP, the fluorescence of Try343 increased by 50% and was blue-shifted by 10nm. GTP-dependent SRP-FtsY complex formation thus leads to an extensive conformational change in the I box insertion in the effector region of FtsY. In vitro crosslinking studies have shown that GTP biding to FtsY, but not GTP hydrolysis, is required for the release of SRP from nascent chains. Using tryptophan fluorescence, it was found that the affinity for GDP binding is five-fold higher than that for GTP. FtsY-W343 showed a 15% (-20%) fluorescence increase upon binding GTP or GDP. Addition of GMPPNP to FtsY-W343 did not cause any fluorescence change, showing that the binding of the non-hydrolyzable analogue does not induce the same structural change around W343 as GTP or GDP. Addition of SRP in the presence of GTP increased the fluorescence to 35% and the emission maxima shifted 10 nm. When the order of mixing was reversed and the GTP was added to premixed FtsY and SRP, the spectral shift was the same while the fluorescence increase was 50%. Experiments showed that GTP binding is much faster than the formation of the FtsY-SRP complex and that the complex is not formed in the absence of GTP. When GTP was rapidly mixed with a solution containing SRP and FtsY-W343, a biphasic fluorescence change was observed. The initial fluorescence increase is attributed to the formation of the SRP-FtsY complex which follows the binding of GTP to FtsY and SRP. The slower fluorescence decrease, the authors say, takes place parallel with GTP hydrolysis and reflects the dissociation of the complex that follows the conversion of GTP to GDP. The reversibility of the fluorescence change reflects the formation of a particular conformation of the I box of FtsY that is present in the complex with the SRP, but not in free FtsY. In the case of the truncated protein FtsY-NG-W343, the fluorescence increased by 60 or 80% upon binding GTP or GDP, respectively. GTP binding was show to be biphasic, indicating a two-step binding mechanism. The conformational change in the NG domain alone upon GTP or GDP binding requires GTP or GMPPNP and SRP binding the native FtsY. GTP binds to a somewhat lower affinity to FtsY (13 uM) than to Ffh (1.3 uM) or SRP (2.3 uM), but qualitatively the binding characteristics are similar. The kinetics are similar with association constants around 1 uM^-1s^-1 and dissociation rate constants of ~10 s^-1. The rate of GTP binding to FtsY is expected to be faster that FtsY- SRP complex formation and FtsY is likely to bind in the GTP-bound form to the complex predominantly, although this may be affected in vivo by ribosomes and membranes. While there is no indication of SRP-FtsY complex formation in the absence of GTP, the interaction does not seem to require GTP in the presence of ribosomes carrying a nascent chain peptide. The blue-shift of the spectrum and 20% enhancement of FtsY-W343 in the presence of GTP or GDP suggests a conformational change in the I box region of FtsY which changes the environment of W343 toward higher hydrophobicity. GTP hydrolysis results in the dissociation of the SRP-FtsY complex here as in eukaryotes.

(Jovine, Structure, 2000) The SRP 4.5S RNA also interacts with EF-G in the ribosome for efficient translation. The 2.7 Angstrom crystal structure of the 4.5S RNA fragment containing binding sites for Ffh and EF-G consists of 3 helices connected by a symmetric and an asymmetric internal loop. The symmetric loop is entire constituted by non-canonical base pairs that continuously stack and project unusual sets of hydrogen-bond donors and acceptors into the shallow minor groove. The structure is two double helical rods hinged by the asymmetric loop that protrudes from one strand. Deletion of the 4.5S RNA is lethal to the E. coli, probably because of its role in protein synthesis. The RNA associates with the ribosome following translocation and prior to the release of uncharged tRNA from the E-site. There are 4 RNA molecules for every Ffh in E. coli so the extras can be involved in protein synthesis, though ribosomes exist in a 25-100-fold excess over the 4.5S RNA. The RNA is a dimer in the crystal. The biologically relevant structure is the monomer represented by nucleotides 31-54 of the first molecule paired with 55-75 of the second. The tetraloop nucleotides of the monomer G53-A56 are the central internal loop in the dimer. Ffh binds to monomers and dimers of the RNA with similar affinity. Mg²⁺ stabilizes the proximity of O3' of A59 and one of the phosphate oxygens of A60. Sequence analysis and mutagenesis studies showed that Ffh binds to the symmetric loop of domain IV. A39 in the asymmetric loop increases the complex affinity. The apical tetraloop is not required for binding. In vitro chemical protection assays show that the protein mainly protects the 5' side of domain IV, with highly conserved nucleotides A39, A47, G48 and G49 being most protected. The last three residues are on the minor groove side of the symmetric loop A, which is unusual. This minor groove of loop A is particularly flat and exposes its stack of non-canonical base pairs, displaying a larger and unique set of hydrogen bond donors and acceptors for specific protein recognition.

(Keenan, Cell, 1998) The authors solved the signal sequence binding subunit of the SRP from T. aquaticus (2FFH.pdb). It shows a deep groove bounded by a flexible loop and lined with side chains of conserved hydrophobic residues to bind the signal sequence and a helix-turn-helix motif containing an arginine rich a-helix required for binding to SRP RNA. The side chains of Leu322, Ile365, Met369, and Phe406 form part of the hydrophobic core of the M-domain. Leu320, Phe325, Leu326, Met329, Leu362, Phe402, and Met409 contribute to the hydrophobic groove. The 19-amino acid, flexible finger loop (residues 337-355) is likely to be responsible for signal sequence binding. Arg384, Arg387, and Lys390 are essential for high affinity binding to SRP RNA. Gly391 and Gly393 are also essential for binding to SRP RNA. Isolated signal sequences are known to form an a-helical conformation in non-polar environments. The dimensions of the hydrophobic groove are sufficient to accommodate ~20 amino acids in an a-helical conformation of ~16 amino acids in a fully-extended a-hairpin conformation. The arginine rich part of the M domain is Arg378-Arg401 in T. aquaticus.

(Montoya, Nature, 1997)The NG domain of E. coli apo-FtsY was solved by x-ray crystallography. The structure consists of the N domain (197-280), the GTPase domain (291-495) and the I box (333-377). b2 in the I box is in an analogous position to that in Ras-GTPases and makes a good candidate for site interaction with a regulatory protein. A comparison of the GTP binding site of FtsY and Ras shows that FtsY residues Asp449, Thr446, Gly385, Arg333, and Asp330 are probably involved in GTP binding analogous to Asn119, Asn116, Gly60, and Thr35, respectively. The P loop in FtsY is stabilized by Asn302 forming a hydrogen bond to Gly385, and Lys306 interacting with Asp382 and Thr383. Thr331 of G2 forms hydrogen bonds to the main-chain oxygen of Ala358, while the main chain oxygen forms hydrogen bonds to the main-chain nitrogen of Gly357. Its being 8 A from the binding site explain the low-affinity of FtsY for GTP. Phe332 may be functionally significant in GTP binding. In FtsY changing Thr446 into Asn decreases GTP binding while substituting Asp449 with Asn changes nucleotide specificity from GTP to XTP. Asp449 has to move into the binding site to bind GTP, however experimental evidence supports this as the N domain is susceptible to protease digestion in the absence of nucleotide but is protected with GTP or GDP bound.

(Peluso, Science, 2000) 4.5S RNA facilitates assembly and disassembly of Ffh-FtsY complexes as monitored by tryptophan fluorescence. Incubation of FtsY with Ffh-RNA in the presence of GppNHp shifted the Trp fluorescence emission by ~10 nm and double the fluorescence intensity. This is consistent with burial of one or both of the Trp in a more hydrophobic environment upon formation of the SRP-FtsY complex. The changes occur in the presence of GppNHp but not GDP. Kinetics of association was monitored with fluorescence studies also. The experiments show that FtsY association in the presence of Ffh-RNA was faster than only Ffh by a factor of 100. The rate constant for Ffh-FtsY complexation is 5.6x10² M^-1s^-1, while that for Ffh-RNA-FtsY complexation is 9.2x10⁴ M^-1s^-1. Typical k_on rate constants for protein-protein association are 10⁶-10⁸ M^-1s^-1. This suggests that the association of the SRP and SR requires conformational rearrangement. In dissociation experiments, Ffh/4.5S RNA-FtsY complex (k_off=3.3x10^-3s^-1) dissociates faster than the Ffh-FtsY (k_off=1.2x10^-5s^-1) only complex. The rates are slower than GppNHp release from the individual components. Addition of 4.5S RNA accelerates the Ffh-FtsY dissociation rate by 200 times. The RNA therefore carries out a catalytic function in the reaction and may serve as as a tether to the 2 proteins. The signal sequence binding to the Ffh could change the RNA conformation, which could lead to changes in kinetics controlling FtsY-Ffh complex formation.

(Moser, PNAS, 1997) The intrinsic guanine nucleotide dissociation rates of FtsY are about 10⁵ times higher than in Ras, but similar to those seen in GTPases in the presence of an exchange factor. The NG domain of FtsY resembles a nucleotide exchange factor complex in its structure and kinetically. The I box is proposed to be a built-in effector that stabilizes the nucleotide-free form of the protein and contributed to low affinity for GTP. The made in E. coli FtsY the XTP-specific mutant Asp449Asn. The fluorescence is increased upon the addition of GDP or GTP to the NG fragment of FtsY by 76% or 57%, respectively. For GDP, the K_d is 2.14 uM and for GTP 10.7 uM. GDP binds 5 fold more strongly than GTP therefore and it causes a larger change in the Trp343 environment also. Binding of both GDP and GTP is a multistep process. FtsY has a much lower affinity for GDP than other GTP-binding proteins. GTP dissociates at an even higher rate than GDP at equilibrium, coupled with the slow rate of binding, this leads to a very low affinity compared to other GTPaes. Thr331 fixed with interactions with the I box could cause the reduced affinity for GTP in comparison to GDP in the FtsY GTPase.

(Batey, 2000) This 1.8 Angstrom structure of the domain IV 4.5S RNA and the Ffh M domain recognizes non-canonical bases pairs in the minor groove of the symmetric loop in a distorted RNA minor groove. The signal sequence recognition surface is composed of both protein and RNA. A helix-turn-helix fold in the 5 a-helices of the M-domain in Ffh is what recognizes the RNA. In the complex, the four nt of the asymmetric loop of the RNA are extruded from the helix. Three of these bases are stacked and wrap around the outside of the helix to form a unique surface that positions invariant nt A39 for contacts with the M domain. It stacks with Arg398 and forms hydrogen bonds to Arg401. Arg398 and Arg401 form an invariant salt bridge with Glu386. Instead of extending into the RNA major or minor grooves, the arginines in the salt bridge present a surface to which the RNA binds. Hydrogen bonds are formed between Arg401 and the N3 of A39 in the asymmetric loop and the 2'-OH of C62 in the asymmetric loop. There is a A47-C62 base pair in the symmetric loop of the minor groove. The C62 has exposed carbonyl and exocyclic amine groups for recognition by the M domain. Although invariant within Ffh/SRP54 proteins, Arg398 is not required for stable association of the SRP RNA to Ffh. Another intriguing possibility is that the A39-Arg398 stack modulates GTPase activity of Ffh and/or FtsY, as the authors suggest by enabling the Arg398 residue to facilitate GTP hydrolysis as has been observed for other GAPs. A39 N3 contacts NH1 of Arg401 and A39 N1 contacts Ser397 Og. A39 2'-OH contacts a phosphate oxygen of A63. The minimal M domain for binding the RNA is residues 328-432. The potassium ion bound to the carbonyl oxygen of G48 and the N3 and 2'-OH groups of G61 in the G-G base pair also coordinates to the backbone carbonyl of Gly-405. The functional groups that define this binding pocket are universally conserved in the SRP and removal of the K⁺ is deleterious to protein binding, much like observations with the role of K⁺ in group I self-splicing introns. Helix 2 of the M domain makes further base-specific contacts between Ala377 O and N6 of G48 and Asn380 and N2 of A47. At the end of helix 3, the conserved GSG sequence participates in RNA recognition through an interaction between Ser406 O and O2' of G49 and the coordination of the K⁺ ion by Gly405 O. The close approach of the protein backbone to the RNA dictates the requirement of a Gly at position 405, whereas Ser406 and Gly407 are necessary for maintenance of the turn between helices 3 and 4. All other contacts between the protein and the RNA are mediated through H-bonding to 2'-OH groups. A mutation at C62G was found to be lethal as it disrupts the protein-RNA interface. Protein-RNA contacts are also observed at the 2'-OH groups of G48 and G49.

(Freymann, Nature, 1997) The authors report the crystal structure of the NG fragment of T. aquaticus Ffh in the apo-form. A ten-residue peptide that links the two domains of the protein is packed tightly against the protein surface, rendering the NG fragment stable from proteolysis. Lys 111 is hypothesized to interact with the b-phosphate, Arg138 is hypothesized to stabilize the g-phosphate leaving group, Asp187 may ligate the Mg^2+, and Arg191 may stabilize the nucleophilic water molecule. Asp348 may provide specificity for the guanine base. In this apo-form Arg191 and Asp135 form a salt bridge. Gln144 will presumably interact with the a and b-phosphate oxygens of bound nucleotide. As hydrophobic residues Leu106 and Leu192 are exposed to the solvent near the active site, they may contribute as protein interaction surfaces such as that that will be recognized by FtsY of the C-terminal M domain of Ffh. Other conserved, possibly important residues, are Asp42, Arg252, Gly253, and Leu257 (part of a hydrophobic core in the N-domain helical bundle). Gly271 -Gly278 cannot be seen in the electron density map. There is no Mg²⁺ in this structure.

(Via, JMB, 2000) The authors determine the nucleotide-binding ability associated with P-loop (nucleotide binding pocket, phosphate binding loop) binding determinants. Could be useful for determining the nucleotide binding to FtsY or Ffh as it was used for Ras/GTP binding. A three-dimensional profile method was used to compare P-loop protein surfaces from various GTP-binding representatives to p21Ras. Conserved P loop residues:

c-H-ras-p21 (5p21.pdb) (GNP) GKSA¹⁸....D⁵⁷....NK¹¹⁷...D¹¹⁹

SRP Ffh (2ng1.pdb) (GDP) GKTT¹¹³...D¹⁸⁷...NK²⁴⁶...D²⁴⁸

FtsY (1fts.pdb) (none) GKTT³⁰⁸...D³⁸²...NK⁴⁴⁷...D⁴⁴⁹

(Wild, Science, 2001) The crystal structure of the human SRP19 complex with its binding site on helix 6 of SRP RNA (29 nt fragments corresponding to nt 135-163) is reported. A model of the assembly of the SRP core comprising SRP19, SRP54, and SRP RNA based on crystallographic and biochemical data is proposed. SRP19 sits on the tip of helix 6 with the ?1 edge of the ??sheet contacting the GGAG tetraloop (GNRA tetraloop is most conserved part of the RNA, but is GGAG here). It binds either direct or through one intervening ordered water molecule, to the backbone of nucleotides C140 to C143 and G147 to G151. A149 of helix 6 is conserved and essential for SRP19 binding, however it does not contact the protein. G147 is tightly embedded into the protein/RNA interface and makes pseudo-Watson-Crick interactions with two water molecules and the phosphate of G150. The water molecule bound to N1 of G147 forms hydrogen bonds to Tyr22 and Arg34. The guanine base is also stacked with the guanidinium group of Arg101, which itself is interlocked between phosphate groups of G148 and A149. Tyr19, Tyr22, and Tyr68 form a triangular clamp pointing to the phosphate group of G150 contributing to several direct and water-mediated hydrogen bonds and reaching also to the adjacent phosphate groups of A149 and C151. Arg81 also contributes to the protein/RNA interface. Glu31 and Gly32 are not involved in RNA binding but are involved in internal loop organization. Glu31 forms a hydrogen bond to the main chain nitrogen of Thr28, Gly32 has such a geometry so that Arg33 and Arg34 span the major groove. The guanidinium group of Arg33 is interlocked between the phosphate groups of C141 and U142 and is H-bonded to the main-chain carbonyls of Lys27 and Ile35. Arg34 is involved in specific but indirect recognition of G147 and forms water-mediated H-bonds to the phosphates of G150 and C151. Both Arg residues are held into place by interactions to the protein backbone to the carbonyl of Gly32. C141 and U142 are H-bonded to the main chain nitrogens of Ile29 and Ile37. The side chain of Arg70 bridges between the carbonyl of Ala30 and the phosphate of G151. SRP54 contact with helix 8 is modeled.

(Freymann, Nature Structural Biology, 1999) This paper reports the crystal structures of T. aquaticus NG domain Ffh as an apo form (with SO₄), with GDP, and with GDP/Mg²⁺ bound. The N domain is hypothesized to regulate or signal the occupancy of the G domain. In the Mg²⁺/GDP complex, the Mg²⁺ is coordinated by the side chain hydroxyl of Thr112, a phosphate oxygen and 4 waters and by the O3 b-phosphate oxygen of the bound nucleotide. The oxygens of the a and b-phosphates, three waters, and Gln144 and Asp187 side chains also interact. The salt bridges in the apo- form are broken so Lys111 turns toward the bound phosphate and the side chain of Arg191 is disordered. The closing loop closes on the nucleotide so that the guanine base is sandwiched by van der Waals interactions between main chain atoms of residues Ser273 and Glu274 on one side and the extended side chain of Lys246 of motif IV on the other. In the GDP complex, the b-phosphate is turned away from the P-loop and interacts with Gln144 and the active site network of salt bridges is restored. Arg191 forms a salt bridge with Asp135. In the apo structure, the closing loop between residues 271-279 is disordered. The side chains of Thr114, Lys117, Lys246, and Asp248 each have roles in binding the guanine bases, but no specific interactions in its absence. In the GDP bound structure, the guanine O6 forms a H-bond with the backbone amide of Lys246. The buried guanine N7 is H-bonded to a solvent inaccessible water that interacts with Lys117 and Thr114. Lys117 also H-bonds with Ser273 and Gly278. LSQMAN can be used for least squares superimpositions. In the presence of Mg²⁺, GDP binds to the Ffh in a conformation similar to that seen in other GTPases, however in the absence of Mg²⁺, the b-phosphate of GDP is turned away from the active site to interact with the Gln144 side chain that is universally conserved in SRPs. These two conformations may reflect different states in the GDP exchange cycle. In this Ffh Mg²⁺/GDP structure, the closing loop wraps around Lys117 and forms van der Waals contacts with the guanine base. Lys117 and Thr114 are bridged by a buried water molecule that forms the floor of the binding site and provides a H-bond to guanine N7. Lys246 and Thr245 interact with carbonyl oxygens of the motif I backbone. Comparison of the apo- and Mg²⁺/GDP structures reveals that the b-phosphate of the GDP is accommodated in the P loop by backbone shifts and coordination of the Mg²⁺ by Thr112. Lys111 positions near the b-phosphate oxygen. Gln107 is H-bonded to the carbonyl oxygen of Arg191. Motif II is disordered upon GDP/Mg²⁺ binding, but the rest of the IBD is unaffected. Gly190 is critical for the positioning of the g-phosphate in the bound GTP in GTPases. In the apo structure here, Gly190 is H-bonded to Arg191. In the Mg²⁺/GDP structure, the H-bond is broken and Gly190 oxygen and amide nitrogen atoms orient to the active site. The motifs II and III are disordered in the Mg²⁺/GDP bound state. The IBD helices pack against the N-domain. A hydrophobic interface comprising highly conserved residues of the ALLEADV and DARGG motifs serve to glue the N and G domains. Water-mediated H-bonds between the N-terminal end of the a4-helix and Asp250 couple the a4-helix shift to a translation of the motif IV/DARGG loop (res. Gly249-Gly253) resulting in the movement of the side chain of motif IV Asp248 ~1.5 Angstroms into the active site. Asp42 H-bonds with Gln224. There are many more details of GDP and apo forms.

(Panmanabhan Freymann, Structure, 2001) The authors discuss the x-ray crystal structure of the NG domain of T. aquaticus Ffh bound GMPPNP. There are 2 structures. In both, the b-phosphate is kinked away from the binding site and Mg²⁺ is not bound (though it was present in the crystallization) and the g-phosphate is turned back toward the P loop. The P loop in this protein is constricted so the protein does not bind the GTP analog as other GTP binding proteins. The authors predict there must be a structural change to move the nucleotide from the inactive to active binding mode. Gln107 and Thr112 form the top and bottom of the P loop jaws in this structure. D248 also binds the GTP analogue. In the canonical mode (this is with GDP?), 8 H-bonds can be formed between the phosphate oxygens and the backbone and side chains of the P loop whereas in the NG complex an alternative set of 7 H-bonds can be formed to the phosphate groups of the bound GMPPNP. In the absence of bound Mg²⁺, the temperature factors of the GMPPNP ligand increase systematically toward the g-phosphate in each of the binding sites. Conserved residues of the P loop include Leu106, Gln107, and Leu192. The interaction of these restricts the P loop from opening. Gln107 is hypothesized to open the P loop to accommodate the g-phosphate as this loop is more open in Ras and other GTP binding proteins by ~1.5 Angstroms. The three residues Leu192, Leu106 and Gln107 with the neighboring Arg191 and the main chain atoms of motifs I and III form an interdigitating and largely hydrophobic contact surface over the P loop that surrounds a buried H-bond between the carbonyl of Arg191 and the amide nitrogen of Gln107. Arg191 also interacts with Asp135. If Gly190 takes a 180^o rotation of the psi angle, the motif III side chain takes on a conformation similar to that seen in other GTPases in the ground state. This rotation takes residues Leu106-112/Asp187-Leu192 and changes 107 from H-bonding with Leu192 carbonyl oxygen to Arg191 H-bonding to the Leu106. The Leu192 residue is solvent exposed and may be involved with contact with FtsY. Binding of GDP and GTP in E. coli Ffh can occur without Mg²⁺. In SRP54, GTP binding alone does not put the protein in the active state. In the absence of a receptor, SRP54 exchanges GDP and GTP readily, even though SRP GTPases are know to have a low affinity for the nucleotides. The crystal structure here then probably shows a structure of the binding mode of GTP in which the protein maintains the conformation of the empty state (hasn't undergone the hypothesized conformational change). In order to have the protein in the active form bound the GTP, as when the FtsY and Ffh contact and stimulate each other, the GTP must be in the extended form to undergo hydrolysis. The authors propose a peptide-flip that is responsible for this rearrangement of the active site. Asp187 is H-bonded to a water H-bonded to the Mg²⁺. The amine of Gly190 forms an H-bond with g-phosphate and may interact with a water to be catalytically important. However, in T. aquaticus Ffh, in contrast to other GTPases with bound GTP analogue, the orientation of this Gly190 peptide bond is reversed so that the position of the carbonyl oxygen is now directed toward the active site and is stabilized by H-bonding with the side chain of Arg191. If the Gly190/Arg191 peptide bond is reoriented, the amide nitrogen of Arg191 could point into the active site and both amide nitrogens could interact with the g-phosphate and waters in the active site. The salt bridge between motif II and III is released. Leu192 contact with the P loop ceases when the g-phosphate goes into the active site. The Arg191 would therefor be an internal arginine finger in the mechanism of the GTPase. Ffh and FtsY probably both dock GTP in the active sites in inactive conformations, then the proteins associate, causing GTP to go into an active conformation so it could be hydrolyzed in the SRP process.

(Montoya, Structure, 2000) The authors report crystal structures of the conserved NG domain GTPase of archeon Acidianus ambivalens and a T112A mutant (which is no longer able to hydrolyze GTP) thereof. The authors also present a heterodimer model for the SRP-SR interaction based on the structural homology of SRP-GTPases with other nucleotide-binding proteins. All of the structures from T. aquaticus and E. coli (FtsY) superimpose within 2 Angstroms. The I-box of A. ambivalens Ffh superimposes better with the I-box of E. coli FtsY than with that from T. aquaticus. The G1 (or P-loop) shows a network of small hydrogen bonds with the G3 region, as in small GTPases. The sidechain of Lys111 interacts with the mainchain oxygen of Thr188. The OH group of Thr112 interacts weakly with Asp187. Gln107 H-bonds Arg138 in the I-box. In the T112A mutant, the loss of the polar hydroxyl group of Thr112 abolishes the interaction with the Asp187 sidechain and the H-bond formed by Gln107 and Arg138. In the T. aquaticus Ffh bound GDP/Mg²⁺, the sidechain of Thr112 is involved with coordinating the Mg²⁺. It interacts with Asp187 in the apo form. The loss of GTPase activity is due to the loss of Mg²⁺ binding ability. Irmi says there is no salt link in her group's structures of E. coli FtsY and A. ambivalens Ffh between corresponding residues of that of Walter's T. aquaticus Asp135-Arg191 salt link. The invariant Asp135 in A. ambivalens is oriented toward the active site to the corresponding Asp330 in E. coli FtsY and in both structures no salt link forms with the invariant Arg191 (Arg386 in FtsY) in the G3 region in contrast to T. aquaticus Ffh. Instead, the aromatic ring of Tyr137 stacks with the guanidinium moiety of Arg191. A conserved Thr residue in G2, Thr35 in p21ras, has been shown to play a role in GTP hydrolysis by the interaction with the Mg²⁺ ion, the g-phosphate and Ser17 and to rotate away after hydrolysis. This Thr is Thr331 in FtsY; substitution by alanine results in a ten-fold decrease in GTP hydrolysis similar to results reported for the Thr35Ala mutant in p21ras. In A. ambivalens and E. coli Ffh, this Thr is a Val so one of the adjacent residues such as Tyr137 or Asp135 must be involved in Mg²⁺ binding for GTP hydrolysis. The same might be proposed for Asp330 in E. coli FtsY where Thr331 stabilizes local interaction in the I-box. T. aquaticus Thr136 interacts with the I-box in the same way. In A. ambivalens, the conserved Arg191 interacts with Tyr137. There is a two residue insertion, Gly193 and Try194. The OH group of Try194 is engaged in a network of H-bonds with the mainchain of Val106 and Gly108 and with the sidechain of Thr110 in the P-loop which is mediated by water molecules. Tyr194 is connected to Thr247 and Lys248 in G4 through waters. Glu197 interacts with Lys227. The G5, or closing-loop region, interacts with the ribose ring of the nucleotide (residues 272-282). Polar interactions, including the sidechain of Lys117 and the mainchains of Gly275, Glu280, and Leu278, stabilize this b-hairpin loop. Gly273 is H-bonded to the mainchains of Met249 and Lys248 in the G4 region, contributing to the b-sheet core in the G domain region. H-bonds important for the N-G interaction are those between the mainchain nitrogen of Gly257 and the mainchain oxygen of Ala39. Mainchain and sidechain atoms of Gln226 are involved in H-bonds with Asp40 and Asn42. In the NG domain is hydrophobic packing of residues Ile4, Phe289, Phe284, Phe84, Met249, Leu259, Ile246, and Ile272. A mutational study in the 'ALLEADV' motif in the N domain showed defects in signal sequence binding by SRP54. The magnitude of the defect was independent of the pre-protein substrate, which suggested that the mutations do not alter the specificity of signal sequence recognition. Mutations in the GTPase consensus had no effect on signal sequence binding, but severely impaired protein translocation activity. These results suggest that the event of signal sequence binding is transmitted to the G-binding domain via a conformational change in the N domain. This change might affect the conserved hydrophobic packing of the aforementioned region. Otherwise a conformation change may occur in the Switch II region. The authors have proposed a model for SRP-SR complex formation using the nitrogenase iron protein structure as a guide. In NIP, the Mg²⁺ ion interacts with Ser16 (Thr112 in A. ambivalens, Ser17 in p21ras) and Asp39 (Asp330 in FtsY, Asp135 in A. ambivalens, Thr35 in p21ras). Asp135 in A. ambivalens may be involved in Mg²⁺ binding. The switch I and II regions are involved in protein-protein interactions in the model. SRP GTPases contain two strictly conserved residues in G2 and G3 regions (Arg138 and Arg191 in Ffh) that might play the role as Arg fingers. They did not consider the conformational change that must occur upon binding Mg²⁺/GTP. Also absent in the model are the M domain of Ffh in complex with a signal peptide, the 4.5S RNA, and the acidic, membrane-associating, A domain of FtsY. It was also formed from structures of the SRP and SR of two different organisms. In the model, the NG domains of A. ambivalens Ffh and E. coli FtsY are at a relative angle of about 90^o.

GTP binding by p21^rasand generalizations for GTP-binding proteins

(Bourne 1991) The protein's hydrophobic core comprises 6 strands of b-sheets, which are connected by hydrophilic loops and a-helices. There are 5 regions critical to GDP/GTP exchange, G1-G5. In G1, the e-amino group of Lys16 (Lys112 in E. coli Ffh and Lys305 in E. coli FtsY) forms a bond with the a- and b-phosphates of GTP or GDP. GTP binding in the G2 region changes the conformation of the second b-strand and the preceding loop by changing the orientation of a critical Thr35 (not conserved in E. coli Ffh and Thr331 in E. coli FtsY). Crystals of p21^ras bound to a GTP analogue, Gpp(NH)p, contain a Mg2+ ion that is coordinated to oxygen of the b-and g-phosphates of GTP and to the side-chain hydroxyls of Thr35 (not conserved in E. coli Ffh and Thr331 in E. coli FtsY) and Ser17 (this is a Thr in Ffh and FtsY: Thr113 in E. coli Ffh and Thr306 in E. coli FtsY) from G1. The invariant Asp57 (Asp190 in E. coli Ffh and Asp382 in E. coli FtsY) in the G3 region bind the catalytic Mg2+ through and intervening water molecule and the amine proton of the invariant Gly60 (Gly193 in E. coli Ffh and Gly385 in E. coli FtsY) forms a hydrogen bond to the g-phosphate of GTP. In the G4 region, the carboxyl oxygens of Asp119 (Asp252 in E. coli Ffh and Asp449 in E. coli FtsY) form hydrogen bonds with groups on the guanine ring, and amide protons of Asn116 (this is a Thr in Ffh and FtsY: Thr249 in E. coli Ffh and Thr446 in E. coli FtsY) and Lys117 (Lys250 in E. coli Ffh and Lys447 in E. coli FtsY) stabilize the guanine nucleotide binding site through hydrogen bonds to residues 13 (109 in E. coli Ffh and 302 in E. coli FtsY) and 14 (110 in E. coli Ffh and 303 in E. coli FtsY) in the G1 region. Residues 144-146 in G5 (274-276 in E. coli Ffh and 572-574 in E. coli FtsY), in a loop between the sixth b-strand and the a5 helix, interact with the guanine nucleotide mostly indirectly through the hydrogen bonds that stabilize side chains of Asn116 (this is a Thr in Ffh and FtsY: Thr249 in E. coli Ffh and Thr446 in E. coli FtsY) and Asp119 (Asp252 in E. coli Ffh and Asp449 in E. coli FtsY) of G4. Ala146 (276 in E. coli Ffh and 435 in E. coli FtsY) in G5 is the only direct contact from this region to the GTP.

Guanine nucleotide release proteins (GNRPs) promote the release of GTP by interacting with the G4 and G5 regions (which bind the nucleotide) and G1 region (which interacts with the G4 and G5 regions) and also with a C-terminal a-helix.

GTP vs. GDP-binding in p21^ras reveals changes in the G-2 loop and the G-3 loop and the a2 helix. Thr35 reorients in G-2. Asp330 in FtsY could position the g-phosphate of GTP in a favorable position for a nucleophilic attack. In some GTPases, deletions in the G3 region prevents conformational changes or prevents it from being transmitted to the effector. The G2 region interacts with GAPs and may interact with an effector. In p21^ras, there is evidence that the Gln61 (Arg194 of E. coli Ffh and Arg386 of E. coli FtsY) is involved in GTP hydrolysis.

Unanswered questions include how the a chain propagates a conformational change from residues in regions G-2 and G-3 which contact the g-phosphate of GTP.

(Resat, PNAS, 2001) Molecular dynamics simulations of the p21ras-p120GAP-GTP complex were performed and the role of Gln61 described. The carbonyl oxygen atom on the backbone of the arginine finger supplied in trans by p120GAP (Arg789) interacts with a water molecule (W230) in the active site that forms a bridge between the NH2 group of the Gln61 and the g-phosphate of GTP. Arg789 may play a dual role in generating the nucleophile as well as stabilizing the transition state for P-O bond cleavage. Gln61 mutations are carcinogenic due to their reducing GAP-stimulated rates of GTP hydrolysis and their increasing the activation barrier for Ras hydrolysis of GTP. That the g-phosphate has a pH of around 3 has been confirmed by ³¹P-NMR experiments.

Human Arf6

(Pasqualato, EMBO Rep., 2001) Arf6 (1HFV.pdb)(ADP-ribosylation) is a GTP-binding protein that coordinates membrane traffic at the plasma membrane with respects of cytoskeleton organization. It resembles very much the structures of Arf1, Ras and Rap2. It localizes at the plasma membrane coordinates endocyclic membrane traffic with aspects of cytoskeleton organization. The authors report the crystal structure of full-length, non-myristolated human Arf6 bound to GTPgS. Whereas the GDP bound forms differ, the GTP-bound forms of Arf6 and Arf1 are very similar. The structural cycle of Arf proteins differs from that of other small GTP-binding proteins in that it couples the classical GDP/GTP nucleotide switch, mediated by switch I and II regions, to a membrane/cytosol switch mediated by the N-terminal helix and a region that connects the switch I and switch II (interswitch). The N-terminal helix in Arf-GDP blocks the interswitch in a retracted conformation which is released by the interaction of the helix with the membranes. The highest B-factors are in the interswitch loop. The overall GDP-GTP switch involves the disorganization of the N-terminal helix binding site and the release of the helix (residues 1-10) into the solvent, and the reorganization of a continuous 40 residue peptide (residues 36-76), which encompasses switch I, the interswitch and switch II. In the GTP-bound form, switch I Thr44 interacts with the GTP?S and Mg²⁺. Gly66 and Gln67 are near the g-phosphate of GTPgS. Also involved in GTP binding are Gly23, Lys 26, Gly66, a water to Thr44, and Thr41 which interacts with the a-phosphate. The paper also details the conformational changes (rotations, changes in H-bonding, etc.) between the GDP to GTP bound forms. GTP-binding proteins display greater structural variations when bound to GDP than to GTP. Switch I Gln37 and Ser38 may be important for interactions with certain effectors when the protein is in the GTP-bound state.

Methods

(Gohlke, JMB, 2000) G. Klebe's group has developed a new knowledge-based scoring function, Drug Score, to describe the binding geometry of ligands in proteins. The program is an improvement from the DOCK program. More often, it is able to dock a ligand to a protein with a lower RMSD and closer to the best geometry than DOCK.

(deGroot, Proteins, 1997) This is a method to generate random structures faster and in a less costly manner as this is without using MD simulations.

FT-IR Mechanistic Results for H-Ras p21 GTPase

(Cepus, Biochemistry, 1998) Using FTIR and 18-O labeled caged GTP isotopomers, the authors propose a dissociative mechanism of hydrolysis for H-Ras p21 hydrolysis of GTP. The results of their experiments indicate that the oxygen atoms of the b-phosphate (PO^2- group) interact more strongly with the protein environment than the g-oxygen atoms. Electrons are withdrawn from the b-phosphorus and thus from the b/g-bridging oxygen. This leads to partial bond breakage or at least weakening of the bond between the b-g-bridging oxygen and the g-phosphorus atom as a putative early step of the GTP hydrolysis. In a dissociative transition state, negative charge is accumulated at the b-phosphate and the binding order of the g-phosphate increases, leading to a metaphosphate-like intermediate. The strong interaction of the b-phosphate with the protein points to a dissociative mechanism.