FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2807655717> ?p ?o ?g. }

• W2807655717 endingPage "986.e6" @default.
• W2807655717 startingPage "972" @default.
• W2807655717 abstract "•The ABC transporter Rv1747 has two cytoplasmic phosphothreonine-binding FHA domains•FHA-2 is circularly permuted and more dynamic than the canonical β-sandwich FHA-1•Both FHA domains bind either of two linker pThr via intra/intermolecular pathways•Complex binding equilibria may enable tunable association and regulation of Rv1747 The Mycobacterium tuberculosis ATP-binding cassette transporter Rv1747 is a putative exporter of cell wall biosynthesis intermediates. Rv1747 has a cytoplasmic regulatory module consisting of two pThr-interacting Forkhead-associated (FHA) domains connected by a conformationally disordered linker with two phospho-acceptor threonines (pThr). The structures of FHA-1 and FHA-2 were determined by X-ray crystallography and nuclear magnetic resonance (NMR) spectroscopy, respectively. Relative to the canonical 11-strand β-sandwich FHA domain fold of FHA-1, FHA-2 is circularly permuted and lacking one β-strand. Nevertheless, the two share a conserved pThr-binding cleft. FHA-2 is less stable and more dynamic than FHA-1, yet binds model pThr peptides with moderately higher affinity (∼50 μM versus 500 μM equilibrium dissociation constants). Based on NMR relaxation and chemical shift perturbation measurements, when joined within a polypeptide chain, either FHA domain can bind either linker pThr to form intra- and intermolecular complexes. We hypothesize that this enables tunable phosphorylation-dependent multimerization to regulate Rv1747 transporter activity. The Mycobacterium tuberculosis ATP-binding cassette transporter Rv1747 is a putative exporter of cell wall biosynthesis intermediates. Rv1747 has a cytoplasmic regulatory module consisting of two pThr-interacting Forkhead-associated (FHA) domains connected by a conformationally disordered linker with two phospho-acceptor threonines (pThr). The structures of FHA-1 and FHA-2 were determined by X-ray crystallography and nuclear magnetic resonance (NMR) spectroscopy, respectively. Relative to the canonical 11-strand β-sandwich FHA domain fold of FHA-1, FHA-2 is circularly permuted and lacking one β-strand. Nevertheless, the two share a conserved pThr-binding cleft. FHA-2 is less stable and more dynamic than FHA-1, yet binds model pThr peptides with moderately higher affinity (∼50 μM versus 500 μM equilibrium dissociation constants). Based on NMR relaxation and chemical shift perturbation measurements, when joined within a polypeptide chain, either FHA domain can bind either linker pThr to form intra- and intermolecular complexes. We hypothesize that this enables tunable phosphorylation-dependent multimerization to regulate Rv1747 transporter activity. Mycobacterium tuberculosis (Mtb), the etiological agent of tuberculosis, has elaborate strategies to circumvent host defense mechanisms. These strategies rely in part on an intricate stimulus-response system that has a set of 11 serine/threonine protein kinases (STPKs) as its central players (Chao et al., 2010Chao J. Wong D. Zheng X. Poirier V. Bach H. Hmama Z. Av-Gay Y. Protein kinase and phosphatase signaling in Mycobacterium tuberculosis physiology and pathogenesis.Biochim. Biophys. Acta. 2010; 1804: 620-627Crossref PubMed Scopus (102) Google Scholar, Wu et al., 2017Wu F.L. Liu Y. Jiang H.W. Luan Y.Z. Zhang H.N. He X. Xu Z.W. Hou J.L. Ji L.Y. Xie Z. et al.The Ser/Thr Protein kinase protein-protein interaction map of M. tuberculosis.Mol. Cell. Proteomics. 2017; 16: 1491-1506Crossref PubMed Scopus (26) Google Scholar). Several substrates of these STPKs contain a Forkhead-associated (FHA) domain (Grundner et al., 2005Grundner C. Gay L.M. Alber T. Mycobacterium tuberculosis serine/threonine kinases PknB, PknD, PknE, and PknF phosphorylate multiple FHA domains.Protein Sci. 2005; 14: 1918-1921Crossref PubMed Scopus (76) Google Scholar). This is a widespread phosphothreonine (pThr)-recognition module, found in eubacteria and eukaryotes, that mediates phosphorylation-dependent protein-protein interactions (Liang and Van Doren, 2008Liang X. Van Doren S.R. Mechanistic insights into phosphoprotein-binding FHA domains.Acc. Chem. Res. 2008; 41: 991-999Crossref PubMed Scopus (42) Google Scholar, Mahajan et al., 2008Mahajan A. Yuan C. Lee H. Chen E.S. Wu P.Y. Tsai M.D. Structure and function of the phosphothreonine-specific FHA domain.Sci. Signal. 2008; 1: re12Crossref PubMed Scopus (110) Google Scholar). One of the FHA domain-containing substrates of the Mtb STPKs is an ATP-binding cassette (ABC) transporter encoded by the open reading frame Rv1747 (Figure 1). Rv1747 has the single polypeptide chain topology expected for a homodimeric exporter with an N-terminal cytoplasmic nucleotide binding domain (NBD), the location of ATP hydrolysis, followed by a helical transmembrane domain through which substrate is transported (Braibant et al., 2000Braibant M. Gilot P. Content J. The ATP binding cassette (ABC) transport systems of Mycobacterium tuberculosis.FEMS Microbiol. Rev. 2000; 24: 449-467Crossref PubMed Google Scholar). Uniquely, the NBD of Rv1747 is preceded by a postulated regulatory module consisting of two FHA domains (FHA-1 and FHA-2) connected by an ∼100 residue intrinsically disordered (ID) linker. The FHA domains are required for specific interactions with Rv1747's “cognate” STPK PknF (Molle et al., 2004Molle V. Soulat D. Jault J.M. Grangeasse C. Cozzone A.J. Prost J.F. Two FHA domains on an ABC transporter, Rv1747, mediate its phosphorylation by PknF, a Ser/Thr protein kinase from Mycobacterium tuberculosis.FEMS Microbiol. Lett. 2004; 234: 215-223Crossref PubMed Google Scholar) and likely with other Mtb STPKs (Grundner et al., 2005Grundner C. Gay L.M. Alber T. Mycobacterium tuberculosis serine/threonine kinases PknB, PknD, PknE, and PknF phosphorylate multiple FHA domains.Protein Sci. 2005; 14: 1918-1921Crossref PubMed Scopus (76) Google Scholar). PknF is considered Rv1747's cognate kinase because the two are encoded by the same operon. PknF phosphorylates the ID linker on least at two confirmed acceptor sites (T152 and T210), and short synthetic peptides corresponding to phosphorylated pT152 or pT210 can bind either of the isolated FHA domains with micromolar affinities (Spivey et al., 2011Spivey V.L. Molle V. Whalan R.H. Rodgers A. Leiba J. Stach L. Walker K.B. Smerdon S.J. Buxton R.S. Forkhead-associated (FHA) domain containing ABC transporter Rv1747 is positively regulated by Ser/Thr phosphorylation in Mycobacterium tuberculosis.J. Biol. Chem. 2011; 286: 26198-26209Crossref PubMed Scopus (29) Google Scholar). The exact functions of Rv1747 have not yet been elucidated. However, four of the remaining five Mtb proteins with FHA domains are involved in processes connected to cell wall synthesis or remodeling (Pallen et al., 2002Pallen M. Chaudhuri R. Khan A. Bacterial FHA domains: neglected players in the phospho-threonine signalling game?.Trends Microbiol. 2002; 10: 556-563Abstract Full Text Full Text PDF PubMed Scopus (92) Google Scholar). Also, experiments on Rv1747 knockout strains revealed altered Mtb sedimentation phenotypes, decreased levels of phosphatidyl-myo-inositol mannosides in lipid extracts (Glass et al., 2017Glass L.N. Swapna G. Chavadi S.S. Tufariello J.M. Mi K. Drumm J.E. Lam T.T. Zhu G. Zhan C. Vilcheze C. et al.Mycobacterium tuberculosis universal stress protein Rv2623 interacts with the putative ATP binding cassette (ABC) transporter Rv1747 to regulate mycobacterial growth.PLoS Pathog. 2017; 13: e1006515Crossref PubMed Scopus (34) Google Scholar), and increased expression of the efflux pump-related iniBAC operon (Spivey et al., 2013Spivey V.L. Whalan R.H. Hirst E.M. Smerdon S.J. Buxton R.S. An attenuated mutant of the Rv1747 ATP-binding cassette transporter of Mycobacterium tuberculosis and a mutant of its cognate kinase, PknF, show increased expression of the efflux pump-related iniBAC operon.FEMS Microbiol. Lett. 2013; 347: 107-115PubMed Google Scholar). Collectively, these data link Rv1747 with the transport of cell wall biosynthesis intermediates. Most importantly, Rv1747 function is associated with Mtb infectivity. The ΔRv1747 mutant shows significantly attenuated growth in macrophages and mice (Curry et al., 2005Curry J.M. Whalan R. Hunt D.M. Gohil K. Strom M. Rickman L. Colston M.J. Smerdon S.J. Buxton R.S. An ABC transporter containing a forkhead-associated domain interacts with a serine-threonine protein kinase and is required for growth of Mycobacterium tuberculosis in mice.Infect. Immun. 2005; 73: 4471-4477Crossref PubMed Scopus (49) Google Scholar, Glass et al., 2017Glass L.N. Swapna G. Chavadi S.S. Tufariello J.M. Mi K. Drumm J.E. Lam T.T. Zhu G. Zhan C. Vilcheze C. et al.Mycobacterium tuberculosis universal stress protein Rv2623 interacts with the putative ATP binding cassette (ABC) transporter Rv1747 to regulate mycobacterial growth.PLoS Pathog. 2017; 13: e1006515Crossref PubMed Scopus (34) Google Scholar). Similarly, strains with the phosphor-ablative T152A/T210A mutations exhibit reduced growth in macrophages and mice, and those with a mutation in the canonical pThr-binding site of FHA-1 have impaired growth in macrophages (Spivey et al., 2011Spivey V.L. Molle V. Whalan R.H. Rodgers A. Leiba J. Stach L. Walker K.B. Smerdon S.J. Buxton R.S. Forkhead-associated (FHA) domain containing ABC transporter Rv1747 is positively regulated by Ser/Thr phosphorylation in Mycobacterium tuberculosis.J. Biol. Chem. 2011; 286: 26198-26209Crossref PubMed Scopus (29) Google Scholar). Thus the functions of Rv1747 are phosphorylation dependent, but very little is currently known about the roles played by the regulatory module. Here, we investigate the structural and functional features of the Rv1747 regulatory module using a combination of nuclear magnetic resonance (NMR) spectroscopy, X-ray crystallography, isothermal titration calorimetry (ITC), and molecular dynamics (MD) simulations. The two FHA domains are structurally independent and joined by a flexible linker (beads on a string). FHA-1 has the canonical 11-strand β-sandwich fold of an FHA domain. In contrast, FHA-2 possesses an unusual, circularly permutated topology lacking one β-strand. FHA-2 is also more dynamic and less stable than FHA-1, yet binds pThr-containing peptides with higher affinity. Constructs of Rv1747 containing either FHA domain and the ID linker with either or both phospho-acceptor threonines exhibit inter- and intramolecular modes of FHA-pThr binding. This enables a complex interaction network that leads to higher order oligomerization of the full regulatory module upon phosphorylation. We hypothesize that association of the regulatory module can be fine-tuned based on the degree and sites of phosphorylation, and that this in turn modulates assembly or activation of Rv1747. We initially focused on three non-phosphorylated constructs: residues 1–156 (Rv17471−156) corresponding to the FHA-1 domain and part of the following interdomain linker; residues 206–310 (Rv1747206−310) corresponding to the FHA-2 domain and a small part of the preceding linker; and residues 1–310 (Rv17471−310) spanning the entire regulatory module (Figure 2A). The assigned 15N heteronuclear single quantum correlation (HSQC) spectra of the first two constructs showed dispersed cross-peaks attributed to the well-ordered regions of the FHA domains and more intense, sharper peaks with random coil 1HN chemical shifts of ∼8–8.5 ppm originating from linker residues (Figures 2B and S1). Consistent with this interpretation, an analysis of the mainchain 1H, 13C, and 15N chemical shifts with the MICS algorithm (Shen and Bax, 2012Shen Y. Bax A. Identification of helix capping and β-turn motifs from NMR chemical shifts.J. Biomol. NMR. 2012; 52: 211-232Crossref PubMed Scopus (69) Google Scholar) indicated the presence of folded domains that are rich in β-strands (Figure 2C). In addition to lacking any predominant secondary structure, the linker residues exhibited MICS-calculated random coil index squared order parameters (RCI-S2) indicative of conformational disorder. The presence of ordered FHA domains and ID linker residues is corroborated by amide 15N relaxation (Figure S2) and hydrogen exchange (HX) experiments, discussed below. The dispersed, assigned signals arising from the ordered domains in the 15N-HSQC spectra of the FHA-1 and FHA-2 constructs overlap closely with those in the spectrum of the full regulatory module (Figure 2C). Based on the lack of amide chemical shift perturbations, it can be inferred that the two folded FHA domains do not interact when joined into a continuous non-phosphorylated polypeptide. Collectively, these data demonstrate that the Rv1747 regulatory module has a beads-on-a-string organization with two structurally independent FHA domains joined by an ID linker. We determined the structure of FHA-1 (Rv17473−116) by X-ray crystallography (Figure 3A and Table 1) and the tertiary structural ensemble of FHA-2 (Rv1747206−310) using NMR spectroscopy (Figure 3B and Table 2). FHA-1 has the canonical FHA domain topology with 11 β-strands in a β-sandwich fold and a shallow pThr-binding cleft extending across its apical surface. FHA-2 also has the expected two-sheet β-sandwich fold of an FHA domain. However, the canonical architecture exhibited by FHA-1 involves a six-stranded anti-parallel β-sheet (strands 2, 1, 11, 10, 7, 8) and five-stranded mixed β-sheet (4, 3, 5, 6, 9) (Liang and Van Doren, 2008Liang X. Van Doren S.R. Mechanistic insights into phosphoprotein-binding FHA domains.Acc. Chem. Res. 2008; 41: 991-999Crossref PubMed Scopus (42) Google Scholar). In contrast, FHA-2 is circularly permuted such that an additional C-terminal strand, denoted 1∗, adopts the position of strand 1 in the first sheet. Also, the equivalent of strand 2 is missing, and thus the anti-parallel β-sheet only has five strands (1∗, 11, 10, 7, 8). The permutation and the absence of strand 2 is particularly interesting as it places the T210 phospho-acceptor within the unstructured linker N-terminal to FHA-2. In a hypothetical homology model of this domain without the circular permutation, T210 would be part of strand 2 and thus likely inaccessible for modification by the Mtb kinases and for FHA domain binding.Table 1X-Ray Crystallography Data Collection and Refinement Statistics for FHA-1 (Rv17473−116)DerivativeNativeNaBrSpace groupP3221P3221Data CollectionUnit cell dimensions a = b, c (Å)58.92, 58.92, 68.6359.05, 59.05, 68.39 α = β, γ (°)90, 90, 12090, 90, 120Wavelength (Å)1.11590.9200.92021.1159Resolution (Å)68.6–1.8 (1.84–1.8)50–2.050–2.050–2.0Observations109,053289,614289,834164,976Unique reflections11,5089,7239,7539,697MultiplicityaParentheses denote values for the highest resolution shell.9.5 (4.9)<I/σI>aParentheses denote values for the highest resolution shell.26.0 (1.9)11 (2.5)12.4 (1.8)35.9 (8.2)Data coverage (%)87.0 (46.2)100100100RmergebRmerge = Σ|I − <I>|/ΣI; I, intensity.6.9 (82.2)15.1 (52.0)13.8 (67.0)6.0 (26.0)Mean figure of meritcMean figure of merit (after density modification) = <|ΣP(α)eiα/ΣP(α)|>; α, phase; P(α), phase probability distribution.0.72 (0.50)Refinement StatisticsResolution range (Å)51.0–1.80Reflections used21,394Mean B factor (Å2)24.1Rcryst (%)dRcryst = Σ|Fo − Fcalc|/ΣFo with observed (Fo) and calculated (Fcalc) structure-factor amplitudes.16.9Rfree (%)20.8Root-Mean-Square DeviationseRoot-mean-square deviations from ideal values.Bonds (Å)0.007Angles (°)1.13Ramachandran Most favored (%)97.0 Allowed (%)3.0a Parentheses denote values for the highest resolution shell.b Rmerge = Σ|I − <I>|/ΣI; I, intensity.c Mean figure of merit (after density modification) = <|ΣP(α)eiα/ΣP(α)|>; α, phase; P(α), phase probability distribution.d Rcryst = Σ|Fo − Fcalc|/ΣFo with observed (Fo) and calculated (Fcalc) structure-factor amplitudes.e Root-mean-square deviations from ideal values. Open table in a new tab Table 2NMR Spectroscopy Data Collection and Refinement Statistics for FHA-2 (Rv1747206−310)NMR Distance and Dihedral RestraintsDistance restraints Total nuclear Overhauser effect1,302 Intraresidue347 Sequential (|i − j| = 1)360 Medium range (|i − j| < 4)135 Long range (|i − j| > 5)460 Hydrogen bonds (present in >6 models)63Dihedral angle restraints Φ68 Ψ68Structure StatisticsViolations (mean ± SD) Distance restraints (Å)0.005 ± 0.002 Dihedral angle restraints (°)4.1 ± 0.2 Max. distance restraint violation (Å)0.54 Max. dihedral angle violation (°)26Ramachandran plot (%) Most favored91.0 Additionally allowed7.6 Generously allowed0.3 Disallowed1.0Average pairwise root-mean-square deviationsaPairwise root-mean-square deviation calculated among the 20 refined structures for residues 28–104. (Å) Backbone atoms0.48 ± 0.1 Heavy atoms0.96 ± 0.1a Pairwise root-mean-square deviation calculated among the 20 refined structures for residues 28–104. Open table in a new tab Despite significant sequence differences between FHA-1 and FHA-2 (27% identity, 36% similarity), as well as the circular permutation of FHA-2 and the absence of one β-strand, the two FHA domains have very similar overall tertiary structures (backbone root-mean-square deviation = 1.4 Å; Figure S3A). Importantly, residues that have been shown to be involved in phosphothreonine peptide binding by canonical FHA domains are conserved in sequence and structure (R33, S47, and H50 in FHA-1; R234, S248, and H251 in FHA-2; Figures 3C, 3D, and S3B). Furthermore, a hydrogen bond network supporting the binding site that is distinctive for FHA domains is present in both FHA-1 and FHA-2. This is readily seen by the characteristic (Mahajan et al., 2008Mahajan A. Yuan C. Lee H. Chen E.S. Wu P.Y. Tsai M.D. Structure and function of the phosphothreonine-specific FHA domain.Sci. Signal. 2008; 1: re12Crossref PubMed Scopus (110) Google Scholar) downfield amide 1HN chemical shifts of S47/S248 (11.58 ppm and 12.15 ppm), which reflect the formation of a hydrogen bond to the deprotonated Nδ1 in the neutral H50/H251 sidechains of FHA-1/2, respectively. In turn, the nitrogen-bonded Hɛ2 on these imidazole rings hydrogen bond to the backbone carbonyl of G71/G271. This protects the labile 1Hɛ2 from rapid hydrogen exchange, yielding detectable signals at 11.61 ppm and 10.77 ppm, respectively, in the 15N-HSQC spectra of the two FHA domains (Figure 2B). The neutral Nɛ2H tautomeric state of H251 in FHA-2 at pH 6 is unambiguously shown by the highly diagnostic chemical shifts of its deprotonated Nδ1 (253 ppm) and protonated Nɛ2 (170 ppm) imidazole ring nitrogens (Platzer et al., 2014Platzer G. Okon M. McIntosh L.P. pH-dependent random coil 1H, 13C, and 15N chemical shifts of the ionizable amino acids: a guide for protein pK a measurements.J. Biomol. NMR. 2014; 60: 109-129Crossref PubMed Scopus (131) Google Scholar). In the case of H50, this conclusion is derived from a physicochemical consideration of the high-resolution crystal structure of FHA-1, combined with NMR spectral features closely resembling those of FHA-2. The interactions of the Rv1747 FHA domains with short phosphopeptides corresponding to the reported PknF acceptor sites pT152 and pT210 in the ID linker were characterized by NMR spectroscopy and ITC. In 15N-HSQC-monitored titrations, both FHA domains bound both phosphopeptides (Figures 4A , S4, and S5). Furthermore, based on the fit Kd values from these titrations (Table 3 and Figure S6), neither FHA domain showed any significant discrimination between the two peptides. This is consistent with the observation that, beyond the invariant pThr, FHA domain specificity is often set by the pT+3 residue (Liang and Van Doren, 2008Liang X. Van Doren S.R. Mechanistic insights into phosphoprotein-binding FHA domains.Acc. Chem. Res. 2008; 41: 991-999Crossref PubMed Scopus (42) Google Scholar, Mahajan et al., 2008Mahajan A. Yuan C. Lee H. Chen E.S. Wu P.Y. Tsai M.D. Structure and function of the phosphothreonine-specific FHA domain.Sci. Signal. 2008; 1: re12Crossref PubMed Scopus (110) Google Scholar). In the case of the two peptides (pT152TRI and pT219SMM), these both have long, hydrophobic sidechains. In contrast, the peptides bound the two FHA domains with substantially different, albeit modest, affinities. The Kd values for either pT152 or pT210 phosphopeptide with FHA-1 were in the near mM range, whereas those with FHA-2 were ∼20- to 30-fold lower.Table 3NMR Spectroscopic and ITC Studies of FHA-Phosphopeptide InteractionsProtein ConstructpHPhospho-peptideKd (NMR)aSee Figures S6 and S7 for binding isotherms. (μM)Kd (ITC)aSee Figures S6 and S7 for binding isotherms. (μM)ΔH (kcal/mol)ΔS (cal/mol K)FHA-1 (Rv17471−156)6pT210750 ± 170560 ± 406pT152680 ± 110350 ± 1908pT152100 ± 60FHA-2 (Rv1747206−310)6pT21023 ± 523 ± 3- 11.9 ± 0.5- 19 ± 26pT15238 ± 1572 ± 12- 31 ± 2- 84 ± 78pT15232 ± 7- 12 ± 2- 20 ± 7a See Figures S6 and S7 for binding isotherms. Open table in a new tab The same series of titrations were also measured using the complementary technique of ITC, and comparable Kd values were obtained (Table 3 and Figure S7). Due to the relatively weak interactions of FHA-1 with the phosphopeptides, reliable ΔH values could not be obtained. In the case of FHA-2, binding was accompanied by a favorable enthalpic change offset by an unfavorable net loss of entropy. For consistency, these calorimetric and spectroscopic studies were carried out at pH 6, with this condition chosen to reduce potential loss of NMR signals from amides in disordered regions of the proteins due to HX. Upon raising the sample pH to 8, the pT152 peptide bound both FHA-1 and FHA-2 with moderately higher affinity (Table 3). This may result in part from an increase in the net negative charge of the phosphate group (unperturbed second pKa ∼ 6.3 for pThr; Bienkiewicz and Lumb, 1999Bienkiewicz E.A. Lumb K.J. Random-coil chemical shifts of phosphorylated amino acids.J. Biomol. NMR. 1999; 15: 203-206Crossref PubMed Scopus (111) Google Scholar) and thus enhanced electrostatic interactions with the conserved R33/R234 in the binding clefts of the FHA domains (Figures 3C and 3D). Previously published ITC measurements of the same protein-phosphopeptide combinations at pH 8 revealed larger negative ΔH values with FHA-2 than FHA-1, yet relatively uniform Kd values of ∼2 μM (Spivey et al., 2011Spivey V.L. Molle V. Whalan R.H. Rodgers A. Leiba J. Stach L. Walker K.B. Smerdon S.J. Buxton R.S. Forkhead-associated (FHA) domain containing ABC transporter Rv1747 is positively regulated by Ser/Thr phosphorylation in Mycobacterium tuberculosis.J. Biol. Chem. 2011; 286: 26198-26209Crossref PubMed Scopus (29) Google Scholar). The differences in these measured affinities likely arise from differences in exact experimental conditions. The NMR-monitored titration experiments also provide insights into the mechanism of phosphopeptide binding. When mapped onto the structures of FHA-1 and FHA-2, residues showing pronounced amide chemical shift perturbations upon addition of either the pT152 or pT210 peptides cluster near the canonical FHA domain binding interface (Figures 4B–4D). However, a visual comparison of the 15N-HSQC titration spectra presented in Figures 4A, S4, and S5 shows clear differences in the effects of the peptides on the two FHA domains. In the case of FHA-1, only a relatively small number of amides had perturbed NMR signals and, of those, most exhibited progressive chemical shift changes upon titration with either phosphopeptide. This is diagnostic of binding in the fast exchange limit, whereby the exchange rate constant kex = kon[peptide] + koff is much greater than the resonance frequency difference Δω between free and bound states (Kleckner and Foster, 2011Kleckner I.R. Foster M.P. An introduction to NMR-based approaches for measuring protein dynamics.Biochim. Biophys. Acta. 2011; 1814: 942-968Crossref PubMed Scopus (359) Google Scholar). This is also consistent with the relatively weak Kd = koff/kon values for binding. A few amides, including S47 and N69, exhibited intermediate exchange broadening (kex ∼ Δω). This can be explained by their intimate involvement in the binding interface and hence a large Δω accompanying phosphopeptide association. Since only 50%–60% saturation was achieved in the titration experiments with FHA-1, sharpening and reappearance of these signals from the fully bound state was not observed. In contrast to FHA-1, numerous amides in FHA-2 showed spectral perturbations upon binding either phosphopeptide. In accordance with Kd values in the 10 μM range, amides with larger Δω values also exhibited line broadening at intermediate titration points (fast-intermediate exchange regime). Somewhat surprisingly, the signals from numerous amides disappeared in the presence of sub-stoichiometric amounts of peptide without the concomitant appearance of new signals from the bound state (90%–95% saturation), as would be expected for slow exchange binding. This is suggestive of conformational exchange broadening in the bound state. Most residues with complete peak attenuation in spectra of FHA-2 were located in or close to the canonical peptide-binding interface, including R234, S248, and N270. It is also noteworthy that the 1NH-15N signals of residues in the appended C-terminal strand 1∗, as well as strand 11, in FHA-2 were more perturbed than those in the equivalent strands 1 and 11 of FHA-1. Thus, the circularly permuted region of FHA-2 is also affected by phosphopeptide binding. Overall, our NMR titrations reveal that pT152 and pT210 phosphopeptide binding leads to significantly more amides showing larger chemical shift perturbations and exchange broadening in FHA-2 compared with FHA-1. This correlates with their relative affinities, and points to a larger binding interface and/or greater conformational perturbations upon association for FHA-2 than FHA-1. Such perturbations could, for example, involve changes in hydrogen bonding, upon which amide chemical shifts are highly sensitive. The bound state of FHA-2 also appears to undergo conformational exchange as indicated by the absence of many amide signals upon saturation. Furthermore, results from ITC experiments indicate that pThr peptide binding by FHA-2 is associated with an entropic loss, which is offset by a large enthalpy change. Taken together, these results suggest that the conformational adaptability of FHA-2 allows for a more favorable bound state. The direct correlation of phosphopeptide affinity with flexibility for the FHA domains is bolstered by denaturation and amide HX studies. To help understand their different structural and phosphopeptide-binding properties, we probed the stability and dynamics of FHA-1 and FHA-2 both experimentally and computationally. In chemical denaturation studies monitored with circular dichroism (CD) spectroscopy, the domains unfolded at distinctly different mid-point [GuHCl]1/2 of 1 M (FHA-1) and 0.5 M (FHA-2) (Figures 5A–5C ). Non-linear regression of these data yielded values of 3.2 ± 0.3 and 2.3 ± 0.8 kcal/mol for ΔGu,H2O, the extrapolated free energy change upon unfolding of FHA-1 and FHA-2, respectively, in the absence of denaturant. By both criteria, FHA-2 is substantially less stable than FHA-1. This conclusion is supported by protection factors (PFs) obtained from amide HX measurements (Figures 5D and 5E). Under EX2 conditions, PFs provide a measure of the residue-specific free energy changes, ΔGHX = 2.303RTlog(PF), governing local or global conformational equilibria leading to exchange. The most slowly exchanging amides in FHA-1 have PFs ∼100-fold greater than those in FHA-2 (log(PF)max ∼ 5.75 versus 3.5). Assuming these amides exchange via global unfolding, this corresponds to ΔGHX values of 7.8 kcal/mol and 4.8 kcal/mol for the FHA domains, respectively. Differences between these values and those obtained via chemical denaturation experiments are attributed to the different experimental conditions used for the two approaches and the extrapolations made to extract free energy changes. Nevertheless, the key conclusion is that FHA-1, with a canonical 11-strand FHA domain fold, is more stable than the circularly permuted 10-strand FHA-2. HX measurements also reveal that circular permutation has an effect on the dynamics of various parts of the FHA domain. Whereas the interacting N- and C-terminal β-strands 1 and 11 are highly protected in FHA-1, the terminal strand 3 and 1∗ of FHA-2 are not interacting and much less protected. Furthermore, amides in or close to the binding interface of FHA-2, including those of residues following strand 3 and in the loop connecting strands 6 and 7, have lower PFs than the corresponding amides in FHA-1. This finding indicates higher relative flexibility for these regions of FHA-2. Importantly, this aligns well with the segments of FHA-2 that displayed pronounced chemical shift changes upon phosphopeptide binding (Figure 4). The correlation of large peptide-dependent spectral perturbations with low unbound state PFs suggests that FHA-2 has a conformationally dynamic binding interface. In contrast, FHA-1 shows somewhat more uniform PFs, with strands 1 and 2 well protected from HX. The greater stability and lower flexibility of FHA-1 may explain the relatively small spectral perturbations upon phosphopeptide binding. To further investigate the dynamics of both FHA domains, we conducted 900 ns all-atom MD simulations in explicit water using the AMBER force field. Over the course of the runs, FHA-1 and FHA-2 remained stable with average Cα root-mean-square deviations from the corresponding experimental structures of only 0.75 Å and 1.35 Å, respectively (Figures S8A and S8B). AMBER back-calculated B-factors, which report on the magnitude of structure dynamics in the nano- to microsecond timescale, showed that, for both domains, the β-strands were rigid, wherea" @default.
• W2807655717 created "2018-06-13" @default.
• W2807655717 creator A5011423412 @default.
• W2807655717 creator A5015787601 @default.
• W2807655717 creator A5015894312 @default.
• W2807655717 creator A5026434669 @default.
• W2807655717 creator A5030761703 @default.
• W2807655717 creator A5038491893 @default.
• W2807655717 creator A5039217746 @default.
• W2807655717 creator A5050925089 @default.
• W2807655717 creator A5055591879 @default.
• W2807655717 creator A5070022547 @default.
• W2807655717 date "2018-07-01" @default.
• W2807655717 modified "2023-10-18" @default.
• W2807655717 title "Biophysical Characterization of the Tandem FHA Domain Regulatory Module from the Mycobacterium tuberculosis ABC Transporter Rv1747" @default.
• W2807655717 cites W1539796472 @default.
• W2807655717 cites W1969164033 @default.
• W2807655717 cites W1969672293 @default.
• W2807655717 cites W1979300827 @default.
• W2807655717 cites W1983732912 @default.
• W2807655717 cites W1986033032 @default.
• W2807655717 cites W1998499862 @default.
• W2807655717 cites W2010255232 @default.
• W2807655717 cites W2013180816 @default.
• W2807655717 cites W2017369100 @default.
• W2807655717 cites W2037575546 @default.
• W2807655717 cites W2038363827 @default.
• W2807655717 cites W2039345443 @default.
• W2807655717 cites W2053202752 @default.
• W2807655717 cites W2055354487 @default.
• W2807655717 cites W2072737923 @default.
• W2807655717 cites W2075008928 @default.
• W2807655717 cites W2096335011 @default.
• W2807655717 cites W2106246537 @default.
• W2807655717 cites W2119320884 @default.
• W2807655717 cites W2120271321 @default.
• W2807655717 cites W2124026197 @default.
• W2807655717 cites W2126818937 @default.
• W2807655717 cites W2127469284 @default.
• W2807655717 cites W2127886370 @default.
• W2807655717 cites W2129510187 @default.
• W2807655717 cites W2136332525 @default.
• W2807655717 cites W2141218040 @default.
• W2807655717 cites W2141433038 @default.
• W2807655717 cites W2141980994 @default.
• W2807655717 cites W2145066627 @default.
• W2807655717 cites W2146387970 @default.
• W2807655717 cites W2152326664 @default.
• W2807655717 cites W2160505835 @default.
• W2807655717 cites W2160519676 @default.
• W2807655717 cites W2163341755 @default.
• W2807655717 cites W2169656841 @default.
• W2807655717 cites W2169821755 @default.
• W2807655717 cites W2180229411 @default.
• W2807655717 cites W2275091356 @default.
• W2807655717 cites W2279433007 @default.
• W2807655717 cites W2330799739 @default.
• W2807655717 cites W23648889 @default.
• W2807655717 cites W2403264193 @default.
• W2807655717 cites W2468070683 @default.
• W2807655717 cites W2620859306 @default.
• W2807655717 cites W2740459087 @default.
• W2807655717 cites W322900245 @default.
• W2807655717 cites W4234722199 @default.
• W2807655717 cites W4245017398 @default.
• W2807655717 doi "https://doi.org/10.1016/j.str.2018.04.018" @default.
• W2807655717 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/29861345" @default.
• W2807655717 hasPublicationYear "2018" @default.
• W2807655717 type Work @default.
• W2807655717 sameAs 2807655717 @default.
• W2807655717 citedByCount "9" @default.
• W2807655717 countsByYear W28076557172018 @default.
• W2807655717 countsByYear W28076557172019 @default.
• W2807655717 countsByYear W28076557172020 @default.
• W2807655717 countsByYear W28076557172021 @default.
• W2807655717 countsByYear W28076557172022 @default.
• W2807655717 countsByYear W28076557172023 @default.
• W2807655717 crossrefType "journal-article" @default.
• W2807655717 hasAuthorship W2807655717A5011423412 @default.
• W2807655717 hasAuthorship W2807655717A5015787601 @default.
• W2807655717 hasAuthorship W2807655717A5015894312 @default.
• W2807655717 hasAuthorship W2807655717A5026434669 @default.
• W2807655717 hasAuthorship W2807655717A5030761703 @default.
• W2807655717 hasAuthorship W2807655717A5038491893 @default.
• W2807655717 hasAuthorship W2807655717A5039217746 @default.
• W2807655717 hasAuthorship W2807655717A5050925089 @default.
• W2807655717 hasAuthorship W2807655717A5055591879 @default.
• W2807655717 hasAuthorship W2807655717A5070022547 @default.
• W2807655717 hasBestOaLocation W28076557171 @default.
• W2807655717 hasConcept C104317684 @default.
• W2807655717 hasConcept C134306372 @default.
• W2807655717 hasConcept C142724271 @default.
• W2807655717 hasConcept C149011108 @default.
• W2807655717 hasConcept C159985019 @default.
• W2807655717 hasConcept C185592680 @default.
• W2807655717 hasConcept C192562407 @default.
• W2807655717 hasConcept C2777814067 @default.
• W2807655717 hasConcept C2777975735 @default.

• next