Advertisement

A cyclic disulfide peptide reproduces in solution the main structural features of a native antigenic site of foot-and-mouth disease virus

Description
A cyclic disulfide peptide reproduces in solution the main structural features of a native antigenic site of foot-and-mouth disease virus
Categories
Published
of 11
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  ELSEVIER International Journal of Biological Macromolecules 20 (1997) 209 219 ~T~N,~AL ~U~AL OF Biological ~cmmolecules ~T~UC ~ Jl~ I~T~q AND ~r ~A~0NS A cyclic disulfide peptide reproduces structural features of a native antigenic disease virus in solution the main site of foot and mouth Thomas Haack a Julio A. Camarero l a Xavier Roig a Mauricio G. Mateu b Esteban Domingo b David Andreu a Ernest Giralt a Department of Organic Chemistry, University of Barcelona, Marti i Franquks 1-11, E-08028 Barcelona, Spain b Severo Ochoa Center for Molecular Biology CSIC-UAM), E-28049 Madrid, Spain Received 12 November 1996; received in revised form 7 February 1997; accepted 21 February 1997 bstract A cyclic disulfide peptide corresponding to the G-H loop sequence 134-155 [replacement Tyr136 and Arg153 with Cys] of the capsid protein VP1 of foot-and-mouth disease virus (FMDV) isolate C-S8cl was examined by proton 2D-NMR spectroscopy in water and in 25% HFIP/water. In water, NMR data supported the presence of a non-canonical turn in the central, conserved cell adhesion RGD motif and suggested the presence of a nascent helix in the C-terminal part, stabilized and slightly extended upon addition of 25% HFIP, a secondary structure stabilizing cosolvent. The formation of the C-terminal helix was evidenced by combined analysis of NOE connectivities, Hc~ chemical shifts, 3JNH H~ oupling constants and amide temperature coefficients. Surprisingly, these global structural features of the cyclic peptide in solution show similarities to previous X-ray structure analysis of (a) a shortened linear peptide complexed with a antivirus antibody and (b) the G-H loop represented on the chemical reduced viral surface of a different serotype. Thus, even in entirely different biological environments the cyclic peptide reflect similar structural features, reinforcing the concept that this viral loop behaves as an independent structural and functional unit. © 1997 Elsevier Science B.V. Keywords: Capsid protein VPI; NMR; Structure in solution I Introduction * Corresponding author. Tel.: + 34 3 4021262; fax: + 34 3 3397878; e-mail: giralt@mafalda.qui.ub.es 1 Present address: Rockefeller University, 1230 York Av- enue, New York, NY 10021, USA. Foot-and-mouth disease virus FMDV) is the cause of one of the economically most important disease of farm animals [1-3]. A major antigenic 0141-8130/97/ 17.00 © 1997 Elsevier Science B.V. All rights reserved. PH S0141-8 1 30(97)01 1 63-X  210 T. Haack et al. International Journal of Biological Macromolecules 20 1997) 209 219 site (site A) of FMDV is located within a highly exposed and mobile region, the G-H loop of capsid protein VP1, on the surface of the viral particle [4,5]. This loop spans VP1 residues 130- 160 and attracts a major portion of the immune response against FMDV [6] (and references cited therein). A significant trait of all FMDV serotypes is the presence in the G-H loop of a highly conserved Arg-Gly-Asp (RGD) tripeptide serv- ing a dual purpose: on the one hand, the RGD motif is directly involved in binding to antibodies [7]; on the other hand, it enables FMDV to recognize and bind the host cell and thus propa- gate the viral infection [8]. The structural study of this loop is therefore of great interest for a molec- ular understanding of such processes as receptor and antibody recognition and neutralization of infectivity. The high mobility of the loop has prevented its study by X-ray diffraction of the native virion [4,5]. However, the crystal structures of the G-H loop on a chemically reduced FMDV of serotype O1-BFS [9], and on a complex be- tween a linear peptide representing this loop of isolate C-S8cl and the Fab fragment of a neutral- izing monoclonal antibody elicited against the complete virus [7] have been reported. In solution studies, several peptides corresponding to the G- H loop of different FMDV serotypes were found to be highly flexible in water and up to 50 helical upon addition of structure-stabilizing sol- vents such as TFE or HFIP [10,11]. Conformationally restricted peptides reproduc- ing continuous epitopes that adequately mimic the native structure have been used as experimental synthetic immunogens with some success [12,13]. In the present case, linear peptides faithfully mim- icked the antigenicity [14] and cell recognition properties [15,16] of the G-H loop of serotype C virus. To further improve this mimicry, conforma- tionally restricted cyclic peptides of the G-H loop of FMDV isolate C-S8cl were designed on the basis of the first three-dimensional structure deter- mined for FMDV (that of native O~BFS of serotype O; [4]). In this structure the disordered region of the G-H loop was delimited by VP1 residues 136 and 157 (equivalent to residues 136 and 153 of C-S8cl). A cyclic disulfide peptide reproducing VP1(134-155), with Cys replace- ments at Tyr136 and Arg153 (cGH[C136,C153], shown below), was as efficiently recognized by antivirus antibodies as its linear counterpart [17] and showed improved immunogenicity, suggesting a resemblance between the immunoactive confor- mation and the cyclic construct [Camarero et al., unpublished results]. This type of restricted mobil- ity analogs of the G-H loop should therefore be helpful in the structural evaluation of the active conformation of this antigenic site. 34 55 TTYT S RGDL HLTTTH RHL G-H loop sequence C-S8c I isolate) I 22 TTCT S RGDL HLTTTH CHL cyclic disulfide analog I I cGH[C136,C153] In the present paper we report the structural characterization of cGH[C136,C153], the cyclic disulfide peptide corresponding to residues 134 155 of VP1 of FMDV C-S8cl, in water and in 25 HFIP/water mixture by 2D-NMR spec- troscopy. Knowledge of the conformational con- straints induced by cyclization should help in the interpretation of the immunological properties of this mimic of the G-H loop. 2 Materials and methods The cyclic disulfide corresponding to the se- quence of the G-H loop of FMDV C-S8cl be- tween VP1 positions 134 155, with Tyr136 and Arg153 replaced by Cys, was synthesized by solid phase methods, purified and characterized as pre- viously described [17]. Homogeneity and identity of the peptide were confirmed prior to NMR by analytical RP-HPLC and ES-MS. NMR experi- ments were performed with 600 /tl of 2 mM samples containing 1.5 mM phosphate buffer and 0.01 mM NaN 3 dissolved either in 85°/,, H,O containing 15 D2 or in 65 H2 containing 25 d2-hexafluoroisopropanol (HFIP) and 10 D20. pH was adjusted to 5.2 using small aliquots of 0.1 N sodium hydroxide or phosphoric acid. Deuterium exchange spectra were recorded with lyophilized samples dissolved in either pure D20 or 75 D20/25 HFIP. The decay of the NH intensities was followed using 1D and short 2D TOCSY spectra. Dioxane was used as internal standard.  T. Haack et al./ International Journal of Biological Macromolecules 20 1997) 209-219 211 NMR spectra were recorded on a Varian VXR-500 instrument operating at a proton fre- quency of 500.013 MHz. Two-dimensional homonuclear clean-TOCSY [18-20], DQF- COSY [21] and NOESY [22,23] were used for spin system and sequential assignment. Amide temperature coefficients were determined be- tween 5 and 35°C in 5°C intervals by recording a series of 1D spectra for the water sample and a series of fast-TOCSY for the HFIP sample. 3JNH_H~ coupling constants were extracted di- rectly from high resolution ID-spectra in the case of the H20 sample or estimated from the antiphase splitting in DQF-COSY [24,25], or rel- ative aN cross-peak intensities in the TOCSY spectra using 25 HFIP as solvent. NOE cross- peaks were integrated at a mixing time of 400 ms A 90 ° pulse was generated applying a 12.4 ps pulse (transmitter power of 56 dB) with the transmitter on-resonance on the H20 signal. H20 suppression was achieved either via the WATERGATE sequence scheme [26] or with a low-power CW irradiation during relaxation de- lay for TOCSY, DQF-COSY, and additionally during the mixing period for NOESY experi- ments. TOCSY spectra were recorded with a spin-locking time of 80 ms and NOESY spectra at three different mixing times (100, 200 and 400 ms), to evaluate spin diffusion effects. Mix- ing times were varied randomly by 10 to avoid zero-quantum effects. The acquisition time for two dimensional spectra was set in general to 0.2 s, followed by a recovery delay of 1.1 s. A spectral width of 5000 Hz was used in both di- mensions containing 2048-1024 data points. Usually 24-32 scans for 256 increments were recorded to achieve sufficient resolution. Data processing was done applying a baseline correc- tion in both F1 and F2 dimension and multipli- cation of the FID by either a shifted gaussian or sine bell function. To achieve higher resolu- tion either zero filling was applied, resulting in a symmetrical matrix of 4096 data points in each dimension, or a linear prediction of the FID resulting in a matrix of 4096 x 2048 points. 3 Results Sequential assignment were done applying the standard two-step procedure developed by Wiithrich et al. [27] using homonuclear scalar (clean-TOCSY, DQF-COSY) and through-space connectivities (NOESY). The peak intensity of the NOESY spectra between 100-400 ms in both, H20 and 25 HFIP, were linear and thus ex- cluded spin diffussion effects leading to three spin NOEs. Aggregation for the cyclic peptide in both, water and 25 HFIP solutions, was ruled out by the invariance of the NMR spectra recorded after a 10 fold dilution from 2 to 0.2 M concentration and by the presence of fast exchanging amide protons (all NHs were exchanged after 5 min) determined by proton-deuterium exchange exper- iments. Nevertheless, substantial line broadening of the amide protons in 25 HFIP decreased cross-peak intensities rendering the conforma- tional analysis difficult. Sequential assignment and conformational analysis based upon NOE connectivities were complicated by the small chemical shift dispersion and sequence degeneracy that results in an exten- sive peak overlap in both, water and 25 HFIP. Nevertheless, assignment was achieved with ex- ception of Thrl based upon a combination of ~N(i,i+l), NN(i,i+l) and some flN(i,i+l) NOEs, shown in Fig. 3. In water, a continuous stretch of NN(i,i + 1) NOEs were observed from Ala7 to Thrl5 (Fig. 2A), extended in 25 HFIP from Ser6 to Alal9 (Fig. 2B). Tables 1 and 2 show the complete assignment of all protons from Thr2 to the C-terminus as extracted from the TOCSY spectra at 5°C (water) and 10°C (25 HFIP), respectively (Fig. 1). Interestingly, in 25 HFIP unlike water, some TOCSY amide-side chain cross-peaks (Leull, Leul4) decreased in intensity with decreasing temperature but were strongly present in the NOESY, pointing to a decrease of the peptide chain mobility. In water, as outlined in the NOE summary in Fig. 3A, a continuous stretch of c~N(i,i + 1) cross- peaks was observed nearly throughout the se- quence, accompanied by NN(i,i + 1) NOEs from Ser6 to Thrl5, especially strong from Aspl0 to Thrl5. Additionally, a single NN(i,i + 3) NOE  212 T. Haack et al. /International Journal of Biological Macromolecules 20 1997) 209 219 F (ppm~ 7.2- 7.4- 8,0. 8.2 ~ 8.4-~ 8.6- 9 ~ ~4, 4, R8 V ig~ 16 ~T1,~-~12~ ----~ {13 -~ ~ DI0~ L14~ ~ ~ ~ ' T17-~- LII~ ~ '¢ S6__ ~ ~' 7 T4~ R8 A~- A5 ~ C3 qm~ T2~ II I ' ' ' ' I .... I ' ' ' ' [ ' ' ' ' l ' ' ' ' I .... I I 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 F1 (ppm) Fig. 1. Part of the TOCSY spectrum (amide-alkyl zone) of cGH[CI36, CI53], 2 mM in 90 H20/10 D20 at 5°C and pH 5.23 (electrode reading). Connectivities are indicated by straight lines and residue numbers refer to the native, viral sequence. was observed between Leul4 and Thrl7 whereas long range NOEs were clearly absent under these conditions. Thus, NOE analysis in water suggests conformational averaging with some preferences to adopt a nascent helix, defined as an ensemble of turn-like structures over several adjacent residues [28], on the C-terminal side of the RGD motif. Upon addition of the structure stabilising solvent HFIP (25 ), beforehand shown by CD studies to enhance the helical content to 38 in comparison with 4 in water, an extended series of stronger sequential NN(i,i + 1) NOEs from Ala5 to Alal9 was apparent and several NN(i,i + 2) or eN(i,i + 2) of low intensity were detected from Aspl0 to Alal9 (Fig. 3B). Furthermore, several long range NOEs were found between His and Leu side chain protons and an eN(i,i+2) between Arg8 and Aspl0. Thus, 25 HFIP seems to stabilise the nascent helix and the fl-turn. A reliable structure calculation based upon NOE constraint were re- jected due to the insufficient amount of NOEs per residue (3-4). Additionally, calculation of inher- ent flexible, small peptide molecules would simply provide an average of the whole conformational ensemble, rather than defined structural trends. Although new approaches tackling this averaging problem have been reported [29-32] the scarce number of NOEs found prevented their applica- tion in the present case. Further secondary structure sensitive NMR parameters such as the amide temperature coeffi- cients (AdNH/AT) to detect solvent-shielded or exposed amide protons [33-35], 3JNH.H ~ coupling constants which have been used to derive back- bone • dihedral angles [36] and the conformation sensitive He chemical shifts (conformational shifts) [37 41] were used additionally to the NOE analysis to derive structural information. In con- trast to the NOE connectivities, which show a dependence on both distance (proportional to r 6) and correlation time, ~,, the conformational shifts and 3JNH.H ~ coupling constants are linear popula- tion-weighted averages over the different exchang- ing conformers and thus in general more appropriate to analyze flexible peptides such as the present one.  T. Haack et al./ International Journal of Biological Macromolecules 20 1997) 209-219 213 A) B) F F _ . (ppm (~qpm) 14-16 15 , 8 2 6 16 8.6 ~1 ~> 7.9 8.2 8.0 8.1 8.4 8.3 ~~~ 8.5 14 ~ ~© 8.6 cz~ ¢. ~ ~rTrv~ r , r~ zvTr'r , , I ' ~ r~, , , , I ' ' ~ , ' 7 r=~ 8.6 8.4 8.2 .... 8.0 8.5 8.3 8.1 7.9 7.7 F1 (ppm) F1 (ppm) Fig. 2. Comparison of the expanded NOESY contour plot (amide-amide zone) of cGH[C136, C153] recorded with a mixing time of 400 ms in (A) H20/10 D20 and (B) H20/25 HFIP and 10 D20. Experimental conditions are as described in Fig. 1. Sequential connectivities NN(i,i+ 1) are indicated by continous lines. Medium or long range NOEs are highligthed by the corresponding residue numbers. Whereas almost all coupling constants in H20 were close to the random coil value of~ 7 Hz (Fig. 3A), some coupling constants in 25 HFIP located in the region from Leull to Hisl8 were significantly smaller (Fig. 3B), suggesting dihedral angles in the a-helical conformational space. This was confirmed by the analysis of the Ha chemical shifts, shown in Fig. 4. in its typical representa- tion as chemical shift deviation from random coil values [40]. In water, three different regions are observed: the N- and C-termini show a typical pattern of end fraying, negative values or upfield shifts typical for a-helices in the N-terminal re- gion from Ala5 to Alal2 with a local minimum at Gly9, and positive values or downfield shifts typi- cal for fl-strands in the C-terminal part from Hisl3 to Alal9, with local minima at Leul4 and Alal9. The temperature dependence of the con- formational shifts (data not shown) defines two clearly distinct zones: N-terminal region from Thrl to Gly9, and a C-terminal section from Aspl0 to Thrl7. A strong and non-linear depen- dence of the A~5 Ha was observed in the N-termi- nal zone, suggesting higher flexibility. The region containing residues Aspl0-Thrl7 does not change AHa with temperature and thus indicates a stable conformation. Addition of HFIP provoke a significant upfield shift of the Ha and amide (less important but notable) protons in the C-ter- minal part from Leull to Alal9 and a small downfield shift in the N-terminal part containing Cys3-Ala7. Recently it was shown, that the influ- ence of trifluoroethanol (TFE) can provoke an increase in secondary structure content in peptides with residual structures in water, but does not change the Ha chemical shifts of 'random-coil' peptides [34]. HFIP used in the present study to increase the secondary structural content was pre- sumed to influence the conformational shift in a similar way to TFE due to their similar physico- chemical properties. Consequently, the results of the conformational shifts in 25 HFIP are consis- tent with the presence of a helical structure at the C-terminal side of the RGD motif from Leul 1 to Alal9 and a possible formation of a fl-strand in the N-terminal part. No general trend was found for the central RGD triad, possibly reflecting the inherent flexibility of turns.
Search
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks