The present work describes the dynamics of the apo form of cytochrome b562, a small soluble protein consisting of 106 amino acid residues [Itagaki, E., and Hager, L. P. (1966) J. Biol. Chem. 241, 3687−3695]. The presence of exchange in the millisecond time scale is demonstrated for the last part of helix IV (residues 95−105 in the holo form). The chemical shift index analysis [Wishart, D. S., and Sykes, B. D. (1994) J. Biomol. NMR 4, 171−180] based on Hα, Cα, Cβ, and C' chemical shifts suggests a larger helical content than shown in the NMR structure based on NOEs. These results indicate the presence of helical-like conformations participating in the exchange process. This hypothesis is consistent with amide deuterium exchange rates and the presence of some hydrogen bonds identified from amide chemical shift temperature coefficients [Baxter, N. J., and Williamson, M. P. (1997) J. Biomol. NMR 9, 359−369]. 15N relaxation indicates limited mobility for the amide protons of this part of the helix in the picosecond time scale. A 30 ns stochastic dynamics simulation shows small fluctuations around the helical conformation on this time scale. These fluctuations, however, do not result in a significant decrease of the calculated order parameters which are consistent with the experimental 15N relaxation data. These results resolve an apparent discrepancy in the NMR structures between the disorder observed in helix IV due to a lack of NOEs and the secondary structure predictions based on Hα chemical shifts [Feng, Y., Wand, A. J., and Sligar, S. G. (1994) Struct. Biol. 1, 30−35].
Mersacidin belongs to the type B lantibiotics (lanthionine-containing antibiotics) that contain post-translationally modified amino acids and cyclic ring structures. It targets the cell wall precursor lipid II and thereby inhibits cell wall synthesis. In light of the emerging antibiotics resistance problem, the understanding of the antibacterial activity on a structural basis provides a key to circumvent this issue. Here we present solution NMR studies of mersacidin-lipid II interaction in dodecylphosphocholine (DPC) micelles. Distinct solution structures of mersacidin were determined in three different states: in water/methanol solution and in DPC micelles with and without lipid II. The structures in various sample conditions reveal remarkable conformational changes in which the junction between Ala-12 and Abu-13 (where Abu is aminobutyric acid) effectively serves as the hinge for the opening and closure of the ring structures. The DPC micelle-bound form resembles the previously determined NMR and x-ray crystal structures of mersacidin in pure methanol but substantially deviates from the other two states in our current report. The structural changes delineate the large chemical shift perturbations observed during the course of a two-step15N-1H heteronuclear single quantum coherence titration. They also modulate the surface charge distribution of mersacidin suggesting that electrostatics play a central role in the mersacidin-lipid II interaction. The observed conformational adaptability of mersacidin might be a general feature of lipid II-interacting antibiotics/peptides. Mersacidin belongs to the type B lantibiotics (lanthionine-containing antibiotics) that contain post-translationally modified amino acids and cyclic ring structures. It targets the cell wall precursor lipid II and thereby inhibits cell wall synthesis. In light of the emerging antibiotics resistance problem, the understanding of the antibacterial activity on a structural basis provides a key to circumvent this issue. Here we present solution NMR studies of mersacidin-lipid II interaction in dodecylphosphocholine (DPC) micelles. Distinct solution structures of mersacidin were determined in three different states: in water/methanol solution and in DPC micelles with and without lipid II. The structures in various sample conditions reveal remarkable conformational changes in which the junction between Ala-12 and Abu-13 (where Abu is aminobutyric acid) effectively serves as the hinge for the opening and closure of the ring structures. The DPC micelle-bound form resembles the previously determined NMR and x-ray crystal structures of mersacidin in pure methanol but substantially deviates from the other two states in our current report. The structural changes delineate the large chemical shift perturbations observed during the course of a two-step15N-1H heteronuclear single quantum coherence titration. They also modulate the surface charge distribution of mersacidin suggesting that electrostatics play a central role in the mersacidin-lipid II interaction. The observed conformational adaptability of mersacidin might be a general feature of lipid II-interacting antibiotics/peptides. dodecylphosphocholine high pressure liquid chromatography heteronuclear single quantum coherence nuclear Overhauser effect nuclear Overhauser effect spectroscopy pulse field gradient dehydroalanine aminobutyric acid S-aminovinyl-methyl-cysteine solvent-accessible surface root mean square deviation Many antimicrobial peptides act against microorganisms through pore formation on the cell membrane. The permeability originates in principle from a nonspecific assembly that results in a pore-like structure where the amphipathic nature of the amino acid composition facilitates the clustering process (for a review, see Ref. 1Zasloff M. Nature. 2002; 415: 389-395Crossref PubMed Scopus (6852) Google Scholar). Apart from this ubiquitous mechanism, some antimicrobial peptides, such as ramoplanin, enduramycin, and janiemycin and the glycopeptides vancomycin and teicoplanin, use specific targets that play a central role in the cell wall synthesis, namely lipid II, to achieve their bioactivity with much higher efficiency (2Brötz H. Josten M. Wiedemann I. Schneider U. Götz F. Bierbaum G. Sahl H.-G. Mol. Microbiol. 1998; 30: 317-327Crossref PubMed Scopus (348) Google Scholar, 3Cudic P. Kranz J.K. Behenna D.C. Kruger R.G. Tadesse H. Wand A.J. Veklich Y.I. Weisel J.W. McCafferty D.G. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 7384-7389Crossref PubMed Scopus (69) Google Scholar, 4Nagarajan R. J. Antibiot. (Tokyo). 1993; 46: 1181-1195Crossref PubMed Scopus (173) Google Scholar). Lipid II also serves as a target for the pore-forming peptides nisin and epidermin, which belong to the type A lantibiotic (lanthionine-containingantibiotic) family (2Brötz H. Josten M. Wiedemann I. Schneider U. Götz F. Bierbaum G. Sahl H.-G. Mol. Microbiol. 1998; 30: 317-327Crossref PubMed Scopus (348) Google Scholar, 5Breukink E. Wiedemann I. van Kraaij C. Kuipers O.P. Sahl H.-G. de Kruijff B. Science. 1999; 286: 2361-2364Crossref PubMed Scopus (636) Google Scholar). Some type B lantibiotics also possess functions of targeted cell wall synthesis inhibition; one of the best studied peptide is mersacidin (see Fig. 1A) (6Brötz H. Bierbaum G. Leopold K. Reynolds P.E. Sahl H.-G. Antimicrob. Agents Chemother. 1998; 42: 154-160Crossref PubMed Google Scholar, 7Brötz H. Bierbaum G. Reynolds P.E. Sahl H.-G. Eur. J. Biochem. 1997; 246: 193-199Crossref PubMed Scopus (159) Google Scholar). Although these lipid II-targeting antimicrobial peptides share a common binding molecule, the recognition epitopes among these peptides are somewhat different: co-incubation of vancomycin or other inhibitors of transglycosylases or transpeptidases with mersacidin does not impede their lipid II binding capacity (6Brötz H. Bierbaum G. Leopold K. Reynolds P.E. Sahl H.-G. Antimicrob. Agents Chemother. 1998; 42: 154-160Crossref PubMed Google Scholar). As antibiotic resistance is becoming more and more severe, the diversity of such a targeting action is of great interest. The sophisticated chemical composition of lipid II provides the complexity that can be targeted in various ways. It consists of a peptidoglycan head group that serves as the building block for the cross-linked cell wall matrix and of a pyrophosphate-undecaprenyl lipid tail that functions as the carrier for the transport of the peptidoglycan moiety from the cytoplasm to the extracellular domain (see Fig. 1B). Although the targeted antimicrobial activity of mersacidin is evident, little detail is known, however, about its mechanism of recognition and inhibition. Mersacidin is a 20-residue peptide with nine post-translationally modified amino acids and a single negatively charged residue, Glu-17 (Fig. 1A) (for reviews, see Refs. 8Brötz H. Sahl H.-G. J. Antimicrob. Chemother. 2000; 46: 1-6Crossref PubMed Scopus (99) Google Scholar and 9Jack R.W. Jung G. Curr. Opin. Chem. Biol. 2000; 4: 310-317Crossref PubMed Scopus (127) Google Scholar). It contains four ring structures: two separate ones in the N-terminal part and two intertwined ones in the C-terminal part. The three-dimensional structure of mersacidin has been solved both by solution NMR spectroscopy (10Prasch T. Naumann T. Markert R.L. Sattler M. Schubert W. Schaal S. Bauch M. Kogler H. Griesinger C. Eur. J. Biochem. 1997; 244: 501-512Crossref PubMed Scopus (46) Google Scholar) and x-ray crystallography (11Schneider T.R. Karcher J. Pohl E. Lubini P. Sheldrick G.M. Acta Crystallogr. Sect. D Biol. Crystallogr. 2000; 56: 705-713Crossref PubMed Scopus (35) Google Scholar). Unlike type A lantibiotics, which are mostly extended and flexible, the structure of mersacidin is globular and compact. In both the crystalline and solution states the local ring structures are tightly confined by the lanthionine linkages. The overall conformations obtained from the two different methods are similar except for a minor difference in the orientation of the glycine-rich ring (residues 5–11). These structures were, however, both solved in pure methanol, due to the poor solubility of mersacidin in aqueous solution, and in the absence of lipid II, which is required for its bioactivity. To understand the mechanism of action of mersacidin, knowledge of its structure upon binding to lipid II under physiological conditions is crucial. We report here high resolution NMR spectroscopy studies of the interaction between mersacidin and its binding target lipid II in dodecylphosphocholine (DPC)1micelles, which were used as a membrane mimic.15N-1H HSQC titration experiments provide residue-specific insight into the interaction with lipid II. The solution NMR structures of mersacidin in various sample environments, namely free in methanol/water solution and DPC-bound in the absence and presence of lipid II, are presented. The effect of binding to DPC micelles and lipid II on the mersacidin dynamics are characterized by means of 15N relaxation together with gradient-edited diffusion experiments. Despite the large number of published solution NMR structural and relaxation dynamics studies of membrane proteins and peptides (12Opella S.J. Nat. Struct. Biol. 1997; 4: 845-848PubMed Google Scholar, 13Opella S.J. Ma C. Marassi F.M. Methods Enzymol. 2001; 339: 285-313Crossref PubMed Scopus (76) Google Scholar, 14Bader R. Bettio A. Beck-Sickinger A.G. Zerbe O. J. Mol. Biol. 2001; 305: 307-329Crossref PubMed Scopus (111) Google Scholar, 15Neidigh J.W. Fesinmeyer R.M. Prickett K.S. Andersen N.H. Biochemistry. 2001; 40: 13188-13200Crossref PubMed Scopus (188) Google Scholar, 16Williams K.A. Farrow N.A. Deber C.M. Kay L.E. Biochemistry. 1996; 35: 5145-5157Crossref PubMed Scopus (69) Google Scholar), only a few examples of high resolution NMR studies of protein/peptide-ligand interactions in the presence of membrane-like environments are available to date (17Kutateladze T. Overduin M. Science. 2001; 291: 1793-1796Crossref PubMed Scopus (146) Google Scholar, 18Hsu S.-T. Breukink E. de Kruijff B. Kaptein R. Bonvin A.M.J.J. van Nuland N.A.J. Biochemistry. 2002; 41: 7670-7676Crossref PubMed Scopus (62) Google Scholar). We will show that the differences in sample environments result in substantial conformational changes that modulate the charge accessibility. These changes in charge distributions most likely play a crucial role in the mechanism of mersacidin bioactivity. For the overproduction and purification of [15N]mersacidin, the producer strainBacillus sp. HIL Y-85,54728 was inoculated into 10 ml of tryptone soy broth and incubated overnight. The preculture served as inoculum for 10 ml of synthetic production medium (19Bierbaum G. Brötz H. Koller K.P. Sahl H.-G. FEMS Microbiol. Lett. 1995; 127: 121-126Crossref PubMed Google Scholar), which contained 50 mm15NH4Cl (>98%) (Cambridge Isotope Laboratories, Inc., Cambridge, UK) and 10 mm K2SO4. It was incubated at 30 °C to an optical density of 0.1 at 600 nm. 0.1 ml of this culture served as inoculum for 10 100-ml cultures in the same medium that were incubated for 72 h at 30 °C with vigorous agitation in 1-liter Erlenmeyer flasks. The culture supernatant was sterilized by filtration and loaded onto a 50-ml Serdolit PAD-I column (Serva Electrophoresis GmbH, Heidelberg, Germany), which had been washed with methanol and equilibrated with distilled water. The column was washed with 10 bed volumes of distilled water and 10 bed volumes of 50% methanol in 50 mm potassium phosphate, pH 7, and the peptide was eluted with 500 ml of acetonitrile, 0.1% trifluoric acid. The antibacterial activity of the fractions was checked in a bioassay usingMicrococcus luteus ATCC 4698 as an indicator strain. Active fractions were pooled and concentrated by rotary evaporation, and precipitated proteins were removed by centrifugation. Aliquots of the concentrate were applied to a POROS 20 R2 HPLC column (Applied Biosystems, Weiterstadt, Germany) using the following gradient (eluent A: 0.1% trifluoric acid in water, eluent B: 0.1% trifluoric acid in acetonitrile): 0 min 5% B, 12 min 30% B, 20 min 40% B, 22 min 100% B at a flow of 5 ml/min. Active fractions were pooled, concentrated by evaporation, and rechromatographed on the POROS column. After lyophilization of the active fractions, the peptide was applied to a reversed phase HPLC column (RP18) using the following gradient: 0 min 5% B, 30 min 50% B, 44 min 67.5% B, 47 min 100% B. The mass of the purified peptide, which eluted at 55% B, was checked by mass spectrometry. Lipid II was prepared as described previously (7Brötz H. Bierbaum G. Reynolds P.E. Sahl H.-G. Eur. J. Biochem. 1997; 246: 193-199Crossref PubMed Scopus (159) Google Scholar). Due to the solubility problem of mersacidin, freeze-dried mersacidin was first dissolved in perdeuterated d3-methanol (Cambridge Isotope Laboratories, Inc.) as a 10 mg/ml stock solution. It was then diluted with sodium phosphate buffer and DPC solution and water. Lipid II was taken from a stock solution (in CHCl3:MeOH = 1:1) and vacuum-dried before mixing. Mersacidin-containing DPC solution was then added to dissolve lipid II, and the resulting sample solution was transferred into an NMR tube for measurements. A short sonication was applied after each mixing step to ensure uniform mixing and proper micelle formation. The typical sample concentration was 2 mm mersacidin in 10 mmsodium phosphate buffer at pH 6.0 with a total volume of 500 μl. The sample hence contained 37% methanol and 63% H2O. For structure determination purposes, 4% (∼100 mm) perdeuterated d38-DPC (Cambridge Isotope Laboratories, Inc.) with or without 2 mm lipid II was added to obtain lipid II embedded in DPC micelles and a control sample of mersacidin in DPC micelles alone. For simplicity, MeOH/H2O, DPCbound, and lipid II are defined here as the three sample conditions of free mersacidin in the methanol/water mixture and in the DPC micelle solution with and without lipid II, respectively. All NMR experiments were carried out on Varian UnityPlus 500 and Bruker DRX600 and DRX750 spectrometers at 293 K. Spectra including two-dimensional NOESY with mixing times of 50, 100, and 200 ms, a total correlation spectroscopy with mixing time of 70 ms, and a three-dimensional NOESY-HSQC with mixing time of 100 ms were collected to obtain complete backbone and side chain proton resonance assignments (21Cavanagh J. Fairbrother W.J. Palmer III, A.G. Skelton N.J. Protein NMR Spectroscopy. 1st Ed. Academic Press, San Diego, CA1996Google Scholar). A two-dimensional double quantum filtered correlation spectroscopy was recorded to extract the backbone3JHN-Hα coupling constants in combination with a short mixing time (50 ms) NOESY as described previously (22Ludvigsen S. Andersen K.V. Poulsen F.M. J. Mol. Biol. 1991; 217: 731-736Crossref PubMed Scopus (166) Google Scholar). Stereospecific assignment of side chain methylene groups was achieved based on the intensity correlations of the spin systems (23Nilges M. Clore G.M. Gronenborn A.M. Biopolymers. 1990; 29: 813-822Crossref PubMed Scopus (106) Google Scholar, 24Xu R.X. Olejniczak E.T. Fesik S.W. FEBS Lett. 1992; 305: 137-143Crossref PubMed Scopus (29) Google Scholar, 25Basus V.J. Methods Enzymol. 1989; 177: 132-149Crossref PubMed Scopus (71) Google Scholar). All spectra were processed using the NMRPipe software package (26Delaglio F. Grzesiek S. Vuister G.W. Zhu G. Pfeifer J. Bax A. J. Biomol. NMR. 1995; 6: 277-293Crossref PubMed Scopus (11638) Google Scholar) and analyzed with NMRView (27Johnson B.A. Blevins R.A. J. Biomol. NMR. 1994; 4: 603-614Crossref PubMed Scopus (2686) Google Scholar) except for those of diffusion measurements (see below). The assigned chemical shift data sets of mersacidin in three different conditions, MeOH/H2O, DPCbound, and lipid II, were deposited into the BioMagResBank under accession codes 5581, 5582, and 5580, respectively. For the starting point corresponding to the free mersacidin in the 37% MeOH, 63% H2O mixture, 1 μmol of the 15N-labeled mersacidin stock solution ind3-methanol was diluted with 50 μl of 100 mm sodium phosphate buffer, and water was added to reach a total volume of 450 μl. In the first titration step, aliquots of a 40% DPC stock solution with 37% methanol content were successively added resulting in DPC concentrations of 0.2, 1, 2, 3, 4, 5, and 6%. This is respectively equivalent to a range of ratios of mersacidinversus DPC micelles of 20:1 up to 1:1.5, assuming that each DPC micelle consists of 50–55 monomers (28Wymore T. Gao X.F. Wong T.C. J. Mol. Struct. 1999; 486: 195-210Crossref Scopus (76) Google Scholar, 29Marrink S.J. Tieleman D.P. Mark A.E. J. Phys. Chem. B. 2000; 104: 12165-12173Crossref Scopus (265) Google Scholar). At the end of the first titration step (DPCbound) the total volume was 525 μl resulting in a 16% sample dilution but with identical methanol content. For the second titration step, portions of 0.05, 0.25, 0.5, 0.75, 1, 1.25, and 1.5 μmol of vacuum-dried lipid II were first prepared in separate containers and then successively dissolved in the sample taken from the NMR tube to reach the same concentration ratios as those of the DPC titration steps. The final sample (lipid II) therefore contained 1 μmol of [15N]mersacidin, ∼1.5 μmol of DPC micelles and 1.5 μmol of lipid II. In both titration steps, the chemical shift changes of mersacidin were all saturated by an excess of DPC micelles and lipid II at a 1:1.25 molar ratio and higher. 15N longitudinal relaxation timesT1, transverse relaxation timesT2, and 1H-15N steady-state NOEs of mersacidin were obtained from series of two-dimensional experiments with coherence selection achieved by pulse field gradients (PFGs) (30Farrow N.A. Muhandiram R. Singer A.U. Pascal S.M. Kay C.M. Gish G. Shoelson S.E. Pawson T. Forman-Kay J.D. Kay L.E. Biochemistry. 1994; 33: 5984-6003Crossref PubMed Scopus (2018) Google Scholar). For T1measurements, seven spectra were recorded with relaxation delaysT set to 10, 20, 100, 300, 500, 750, and 1000 ms. For the rotating frame relaxation times T1ρ, spectra were first obtained using a 15N spin lock continuous wave radio frequency (rf) with a field strength ν1 of 1 kHz for seven relaxation delays T of 20, 50, 70, 100, 120, 200, and 350 ms. The relaxation times T1 and T1ρ were then derived from a single exponential decay fitting of the peak intensities usingxcrfit (www.pence.ca/ftp/xcrvfit). TheT2 of each residue was subsequently derived from the observed relaxation time T1ρ by correcting for the offset Δν of the rf field to the resonance by use of the relation 1/T1ρ = (1/T1) cos2θ + (1/T2) sin2θ, where θ = tan−1(ν1/Δν).1H-15N heteronuclear NOEs were determined from the ratio of peak intensities (Ion/Ioff) with and without the saturation of the amide protons for 3 s. All15N relaxation experiments were carried out on a Varian UnityPlus 500 spectrometer at 293 K. Since the structure of mersacidin exhibits high flexibility, quantitative analysis of the relaxation data in terms of spectral density functions is not applicable (31Dayie K.T. Wagner G. Lefevre J.F. Annu. Rev. Phys. Chem. 1996; 47: 243-282Crossref PubMed Scopus (138) Google Scholar). Therefore, the relaxation data in the three different sample conditions were only compared in a qualitative way. A further simplification allows us to estimate the correlation time from the averaged T2 value (32Anglister J. Grzesiek S. Ren H. Klee C.B. Bax A. J. Biomol. NMR. 1993; 3: 121-126Crossref PubMed Scopus (90) Google Scholar) according to Equation 1,τc∼1/(5T2)Equation 1 The overall correlation time can also be expressed as a function of the molecular size and the bulk solvent viscosity η. For a spherical molecule of radius a rotating in a liquid of viscosity η, the rotational correlation time τc is given by the Stokes relation (33Wand A.J. Ehrhardt M.R. Flynn P.F. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 15299-15302Crossref PubMed Scopus (113) Google Scholar) according to Equation 2,τc=4πa3η/3kT=Vη/kT=1/6DrotEquation 2 where V is the volume of the molecule, kis the Boltzman constant, T is the absolute temperature, and Drot is the rotational diffusion constant for a spherical molecule. The molecular radius a is actually the effective hydrodynamic radius Rh of the molecule with the hydration shell. The overall correlation time estimated from NMR relaxation measurements can thus be compared with the effective molecular size and sample viscosity that are measured from PFG-NMR diffusion experiments. The hydrodynamic radius Rh can be calculated from the translational diffusion coefficient Dtransof the particles through another Stokes relation according to Equation3, Dtrans=kT/6πηRhEquation 3 PFG-NMR diffusion measurement with the PG-SLED (pulse gradient-stimulated echo longitudinal encode-decode) sequence enables us to obtain Dtrans, which is proportional to the decay rate d of the NMR signal attenuation as a function of gradient strength g (34Wilkins D.K. Grimshaw S.B. Receveur V. Dobson C.M. Jones J.A. Smith L.J. Biochemistry. 1999; 38: 16424-16431Crossref PubMed Scopus (830) Google Scholar), according to Equation 4,I(g)=Ioexp[−dg2]Equation 4 where Io is the NMR peak intensity in the absence of gradient pulses and g is the field strength of the bipolar gradient pulse pair. Changes in the solvent viscosity η in different sample environments can be monitored using the methanol signal as an internal standard, assuming that methanol does not interact with other solutes and thus that its hydrodynamic radiusRh is invariable in analogy to the use of dioxane for protein folding studies as described previously (34Wilkins D.K. Grimshaw S.B. Receveur V. Dobson C.M. Jones J.A. Smith L.J. Biochemistry. 1999; 38: 16424-16431Crossref PubMed Scopus (830) Google Scholar). The ratio of the rate constants of methanol in different conditions gives the relative change in bulk solvent viscosity η. Knowing this, the relative hydrodynamic radii Rh of mersacidin with respect to methanol in different environments can thus be extracted. In practice, each diffusion data set consists of a series of 40 one-dimensional 1H spectra with 2.5% increments of the gradient strength from 2.5 to 100% collected at 750 MHz with a three-axis gradient probe (x axis for bipolar gradient pulse pair and y and z axis for residual signal crushing). Data processing was performed with Felix from Biosym Technologies (San Diego, CA), and Origin7.0 from OriginLab (North Hampton, MA) was used for non-linear fitting to obtain the translational diffusion coefficients Dtrans. All structure calculations were performed with the program CNS (35Brünger A.T. Adams P.D. Clore G.M. DeLano W.L. Gros P. Grosse-Kunstleve R.W. Jiang J.S. Kuszewski J. Nilges M. Pannu N.S. Read R.J. Rice L.M. Simonson T. Warren G.L. Acta Crystallogr. Sect. D Biol. Crystallogr. 1998; 54: 905-921Crossref PubMed Scopus (16979) Google Scholar) using the ARIA setup and protocols (36Nilges M. Donoghue S.I. Prog. Nucl. Magn. Reson. Spectrosc. 1998; 32: 107-139Abstract Full Text PDF Scopus (224) Google Scholar). Semiautomated NOE assignment was used to assist the spectral assignment (37Linge J.P. O'Donoghue S.I. Nilges M. Methods Enzymol. 2001; 339: 71-90Crossref PubMed Scopus (333) Google Scholar). This was done from a partially assigned NMRView peak list. The initially unassigned cross-peaks were defined as ambiguous distance restraints with a lower weighing factor. The calibration of the cross-peak intensities against distances was done automatically at the beginning of each iteration. The additional unambiguously assigned cross-peaks were interactively re-examined with NMRView, and the checked peak list was then used as the input for the next calculation in an iterative way until all cross-peaks were assigned. Each semiautomated assignment step with ARIA consisted of eight iterations with successive reduction of the violation tolerance and a final refinement in explicit solvent using default ARIA parameters unless otherwise stated. The Parallhdg5.3 force field with the PROLSQ parameters was used (38Linge J. Williams M.A. Spronk C.A.E.M. Bonvin A.M.J.J. Nilges M. Proteins Struct. Funct. Genet. 2003; 50: 496-506Crossref PubMed Scopus (544) Google Scholar). The topologies of dehydroalanine (Dha), aminobutyric acid (d-Abu), 3-methyllanthionine, and the cyclized C terminus of the S-aminovinyl-methyl-cysteine (Tea) were constructed based on alanine, threonine, and cysteine and comparison of available data bases. Four thioether bridges were introduced. Nine backbone φ and four side chain χ1torsion angle restraints obtained from the stereospecific assignments of the methylene groups of the thioester-linked Ala-12 and Ala-18 and of Leu-14 and Glu-17 were used in the structure calculations. A torsion angle dynamics simulated annealing protocol was performed, initially at 10,000 K (8000 steps), followed by a first cooling stage to 50 K (50 K/step); Cartesian space refinement was used for the second cooling stage (from 2000 to 1000 K in 16,000 steps) and the subsequent third cooling stage (from 1000 to 50 K in 4000 steps) followed by 200 steps of energy minimization. The slow cooling process at the second stage ensures a better convergence of the calculated structures. The 50 structures with the lowest restraint energy were further subjected to explicit solvent refinement (OPLS water and Me2SO models) as described previously (39Bonvin A.M.J.J. Houben K. Guenneugues M. Kaptein R. Boelens R. J. Biomol. NMR. 2001; 21: 221-233Crossref PubMed Scopus (25) Google Scholar), and the best 20 were kept for clustering and structural analysis. A cluster is defined as a group of at least four structures with pairwise backbone (residue 3, 4, and 12–20) positional root mean square deviations (r.m.s.d.) lower than 0.3 Å. Structures were visualized and analyzed with MOLMOL (40Koradi R. Billeter M. Wüthrich K. J. Mol. Graph. 1996; 14: 51-55Crossref PubMed Scopus (6498) Google Scholar). DynDom was used to identify conformational changes and to define domains and effective hinge regions of the structures obtained under the three different sample conditions (41Feenstra K.A. Hess B. Berendsen H.J.C. J. Comput. Chem. 1999; 20: 786-798Crossref Scopus (670) Google Scholar). The coordinates of the three structure ensembles, MeOH/H2O, DPCbound, and lipid II, were deposited in the Research Collaboratory for Structural Bioinformatics Protein Data Bank under accession codes 1MQX, 1MQY, and 1MQZ, respectively. The assignment of the resonances of mersacidin in various sample environments was achieved by a standard multidimensional NMR protocol using the15N-labeled sample (21Cavanagh J. Fairbrother W.J. Palmer III, A.G. Skelton N.J. Protein NMR Spectroscopy. 1st Ed. Academic Press, San Diego, CA1996Google Scholar). A two-step15N-1H HSQC titration of perdeuterated DPC micelles followed by unlabeled lipid II was then used to investigate the interaction of mersacidin and lipid II. Chemical shifts are excellent probes for biomolecular interaction studies because of their sensitivity to the changes in the surrounding chemical environment. They are commonly used for the mapping of ligand binding sites in proteins (42Zuiderweg E.R. Biochemistry. 2002; 41: 1-7Crossref PubMed Scopus (477) Google Scholar), and yet, as will be discussed later, the chemical shift perturbations upon ligand binding can also be governed by structural rearrangements. The first step (DPC micelles) was used as a control to monitor how mersacidin was influenced by the membrane-mimicking environment. Both sequential titration experiments revealed significant chemical shift perturbations in mersacidin (Fig.2). The addition of DPC micelles strongly affects the backbone amide protons of Gly-7, Abu-13, Abu-15, and Glu-17 in the 1H dimension (mainly upfield shifts) and most of the C-terminal part in the 15N dimension (Fig.3). Along the 15N dimension, a downfield shift occurs at Dha-16, which is flanked by progressively increasing, almost symmetric, upfield shifts. Subsequent addition of lipid II gives rise to large downfield shifts for the amide protons of Gly-7, Abu-13, and Glu-17 in the 1H dimension; a similar chemical shift perturbation profile is observed in the 15N dimension. The general direction of change is, however, inverted compared with the first DPC micelles titration. Both titration steps result in clearly localized effects. Unlike the strong binding affinity between nisin and lipid II, which results in slow exchange cross-peak patterns in similar titration experiments (18Hsu S.-T. Breukink E. de Kruijff B. Kaptein R. Bonvin A.M.J.J. van Nuland N.A.J. Biochemistry. 2002; 41: 7670-7676Crossref PubMed Scopus (62) Google Scholar), mersacidin follows a fast exchange profile indicative of weaker binding under these conditions. The gradual changes in position can easily be followed during the course of both titration processes (Fig. 2). Unexpectedly the addition of DPC micelles seems to affect mersacidin more than its specific target, lipid II, does. The titration data suggest an interesting interconversion process. For Gly-7, Gly-8, Ala-12, Abu-13, and Glu-17, the 1H chemical shifts first move upfield and then conversely shift downfield. A similar behavior is monitored along the 15N dimension for most residues, e.g.Glu-17. Since chemical shifts are closely related to the surrounding chemical environment primarily defined by the three-dimensional structure, the direction and displacement of the cross-peaks suggest that the structure of mersacidin undergoes a substantial change upon addition of DPC micelles and is somehow restored close to its initial conformation, if not identical, upon the subsequent addition of lipid II. Yet this could also be an indication that mersacidin falls off the DPC micelles after addition of lipid II and is restored to a state simila
Continuous growth in international waterborne trade requires that global container handling operations become faster and more efficient. In response to this requirement, an increasing number of terminal operators have chosen to move from conventional to automated container handling systems. Automated systems have proven to be more efficient, reliable, and faster than their conventional counterpart. In moving to automated container handling systems, engineers realize that automated operations and equipment have unique infrastructure requirements that must be addressed during the design phase of terminal development. Communication between the terminal operator and infrastructure designer from the early stage of terminal design is critical for optimal terminal performance and to avoid costly changes in the future. This paper addresses critical interface elements between infrastructure design and automated operations.
The POU-domain transcription factor Pit-1 and Ets-1, a member of the ETS family of transcription factors, can associate in solution and synergistically activate the prolactin promoter by binding to a composite response element in the prolactin promoter. We mapped the minimal region of Ets-1 required for the interaction with the Pit-1 POU-homeodomain. Here, we describe a detailed NMR study of the interaction between the POU-homeodomain of Pit-1 and the minimal interacting region of Ets-1. By using heteronuclear single quantum coherence titration experiments, we were able to map exact residues on the POU-homeodomain that are involved in the interaction with this minimal Ets-1 interaction domain. By using our NMR data, we generated point mutants in the POU-homeodomain and tested their effect on the interaction with Ets-1. Our results show that phosphorylation of Pit-1 can regulate the interaction with Ets-1.
Proteins frequently contain unstructured regions apart from a functionally important and well-conserved structured domain. Functional and structural aspects for these regions are frequently less clear. The general human positive cofactor 4 (PC4), has such a domain organization and can interact with various DNA substrates, transcriptional activators, and basal transcription factors. While essential for the cofactor function, structural and functional knowledge about these interactions is limited. Using biochemical, nuclear magnetic resonance (NMR), and docking experiments, we show that the carboxy-terminal structured core domain (PC4ctd) is required and sufficient for binding to single-stranded DNA (ssDNA), double-stranded DNA (dsDNA), and the herpes simplex virion protein 16 (VP16) activation domain (VP16ad). We determined the interaction surfaces within PC4 and showed that VP16 and DNA binding are mutually exclusive. Although the amino-terminal domain of PC4 (PC4ntd) alone is devoid of any bioactivity, it increases the interaction with VP16ad. While it decreases the ssDNA-binding and DNA-unwinding activity, it does not influence dsDNA binding. Structural characterization of this domain showed that it is highly flexible and mostly unstructured both in the free form and in the complex. NMR titration experiments using various protein and DNA substrates of the individual domains and the full-length PC4 revealed local conformational or environmental changes in both the structured and unstructured subdomains, which are interpreted to be caused by inter- and intramolecular interactions. We propose that the unstructured PC4ntd regulates the PC4 cofactor function by specific interactions with the activator and through modulation and/or shielding of the interaction surface in the structured core of PC4ctd.
The long-lived light-induced intermediate (pB) of the E46Q mutant (glutamic acid is replaced by glutamine at position 46) of photoactive yellow protein (PYP) has been investigated by NMR spectroscopy. The ground state of this mutant is very similar to that of wild-type PYP (WT), whereas the pB state, formed upon illumination, appears to be much more structured in E46Q than in WT. The differences are most striking in the N-terminal domain of the protein. In WT, the side-chain carboxylic group of E46 is known to donate its proton to the chromophore upon illumination. The absence of the carboxylic group near the chromophore in the E46Q mutant prohibits the formation of a negative charge at this position upon formation of pB. This prevents the partial unfolding of the mutant, as evidenced from NMR chemical shift comparison and proton/deuterium (H/D) exchange studies.
