Tài liệu Báo cáo khoa học: Design, structure and biological activity of b-turn peptides of CD2 protein for inhibition of T-cell adhesion ppt

Eur J Biochem 271, 2873–2886 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04198.x Design, structure and biological activity of b-turn peptides of CD2 protein for inhibition of T-cell adhesion Liu Jining1, Irwan Makagiansar3, Helena Yusuf-Makagiansar3, Vincent T K Chow2, Teruna J Siahaan3 and Seetharama D S Jois1 Department of Pharmacy and 2Department of Microbiology, National University of Singapore, Singapore; 3Department of Pharmaceutical Chemistry, The University of Kansas, Lawrence, KS, USA The interaction between cell-adhesion molecules CD2 and CD58 is critical for an immune response Modulation or inhibition of these interactions has been shown to be therapeutically useful Synthetic 12-mer linear and cyclic peptides, and cyclic hexapeptides based on rat CD2 protein, were designed to modulate CD2–CD58 interaction The synthetic peptides effectively blocked the interaction between CD2– CD58 proteins as demonstrated by antibody binding, E-rosetting and heterotypic adhesion assays NMR and molecular modeling studies indicated that the synthetic Accessory molecules, CD2–CD58 receptor-ligand pair [1–4] are important for adhesion and costimulation in the normal immune response The CD2 molecule is a transmembrane glycoprotein expressed on all subsets of T-cells, NK cells and lymphokine-activated killer cells, all known to be effectors of autoimmune disease and allograft rejection Its ligand, CD58 or leukocyte function associated antigen-3 (LFA-3), is also a transmembrane glycoprotein, distributed widely on T and B lymphocytes, erythrocytes, endothelium, platelets, and granulocytes It has been found that this heterophilic adhesion facilitates initial cell–cell contact before specific antigen recognition, and also enhances T-cell receptor (TcR) triggering by fostering interaction with peptide-class II major histocompatability complex (pMHC) The affinity of CD2–CD58 interaction is relatively low (Kd % lM), with very rapid koff and kon that supports dynamic binding with rapid counter-receptor exchange This creates an optimal intercellular membrane ˚ distance (% 135 A) on opposing cell surfaces suitable for TcR-pMHC or NK receptor–MHC interactions to foster immune recognition Hence, in the presence of human CD2–CD58 interaction, T-cells recognize the correct Correspondence to S D S Jois, Department of Pharmacy, 18 Science drive 4, National University of Singapore, Singapore 117543 Fax: + 65 779 1554, Tel.: + 65 874 2653, E-mail: phasdsj@nus.edu.sg Abbreviations: AET, 2-aminoethylisothiouronium hydrobromide; BCECF-AM, bis-carboxyethyl-carboxyfluorescein, acetoxymethyl; FITC, fluorecein isothiocynate; hCD2, human CD2; hCD58, human CD58; MEM-a, minimum essential medium-a; MTT, [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide]; PAL-resin, 5-(4-aminomethyl-3,5-dimethoxyphenoxy)valeryl-resin (Received January 2004, revised 22 April 2004, accepted 30 April 2004) cyclic peptides exhibit b-turn structure in solution and closely mimic the b-turn structure of the surface epitopes of the CD2 protein Docking studies of CD2 peptides and CD58 protein revealed the possible binding sites of the cyclic peptides on CD58 protein The designed cyclic peptides with b-turn structure have the ability to modulate the CD2–CD58 interaction Keywords: CD2, b-turn, cyclic peptide, E-rosetting, LFA-3 (CD58) pMHC with a 50- to 100-fold greater efficiency than its absence [4] In addition, endothelial cells (EC) in rheumatoid arthritis (RA) have been shown to express elevated levels of CD58, and RA lymphocytes in synovial fluid (SF) express increased levels of CD2 and CD58 relative to RA or normal peripheral blood lymphocytes [5,6] Thus, the inhibition of CD2–CD58 interaction can potentially be used for the treatment of autoimmune diseases It has been shown that blockade of the CD2–CD58 interaction [7,8] and/or modulation of the CD2 costimulatory pathway [9–12] can result in prolonged tolerance towards allografts The soluble CD58–Ig fusion protein Amevive (LFA3TIP) has been used to treat psoriasis [13] However, antibodies are huge protein molecules and therapeutic antibodies are nonhuman in origin, these often elicit significant side-effects attributed to their immunogenicity The humanized versions of antibodies BTI-322 [14] and MEDI-507 [15] have been tested for the treatment of acute organ rejection and graft-vs.-hostdisease Furthermore, MEDI-507 is also investigated for autoimmune and other inflammatory diseases Antibodies are susceptible to enzymatic degradation and hence pose a challenge for formulation and delivery To circumvent this problem, one approach is to design short peptides or small molecular mimics that will bind to critical areas in target proteins (CD58) and, like antibodies, interfere with their activity Currently, no peptide or small molecules targeting CD2 or CD58 have been yet reported Therefore, this study was undertaken to design small peptides based on CD2 protein epitopes to modulate CD2–CD58 interaction We designed linear and cyclic peptides (Table 1) from the b-turn regions of rat CD2 protein (Fig 1), and evaluated their ability to inhibit cell adhesion using antibody, E-rosetting and heterotypic-adhesion Ó FEBS 2004 2874 L Jining et al (Eur J Biochem 271) Table Peptides used in this study that are derived from rat CD2 protein The sequence number refers to the residues from the second position in the peptide to eleventh position Pen1 and Cys12 were introduced for cyclization purpose purchased from Biodesign International (Saco, ME, USA) and Immunotech, respectively The Jurkat, MOLT-3 T-leukemia and the human colon adenocarcinoma (Caco-2) cell lines were obtained from the American Type Culture Collection (Rockville, MD, USA) Jurkat and MOLT-3 cells were maintained in suspension in RPMI1640 medium supplemented with 10% (w/v) heatinactivated fetal bovine serum and 100 mgỈL)1 of penicillin/ streptomycin Caco-2 cells were maintained in minimum essential medium-a containing 10% (w/v) fetal bovine serum, 1% (v/v) nonessential amino acids, mM Na-pyruvate, 1% (v/v) L-glutamine and 100 mgỈL)1 of penicillin/streptomycin Caco-2 cells were used between passages 50 and 60 Sheep blood in Alsever’s solution was purchased from TCS Biosciences Ltd., Singapore Code Name Sequence number in the native protein lER cER lVY cVY cEL cYT Control peptide PenERGSTLVAEFC Cyclo(1,12) PenERGSTLVAEFC PenVYSTNGTRILC Cyclo (1,12) PenVYSTNGTRILC Cyclo (1, 6) ERGSTL Cyclo (1, 6) YSTNGT KGKTDAISVKAI 36–45 36–45 85–94 85–94 36–41 85–90 91–80a a Sequence from human CD2 The sequence was reversed, Tyr81 and Ty86 were replaced by Ala inhibition assays In order to understand structure–function relationship of peptides, we also carried out detailed NMR, molecular modeling and docking studies of peptide-protein complexes Our results indicate that the designed peptides are useful for inhibition of the T-cell adhesion mechanism Materials and methods Peptides The linear and cyclic peptides lER, lVY, cER, cVY, cEL and cYT (Table 1) were designed and purchased from Multiple Peptide Systems (San Diego, CA, USA) The pure product was analyzed by HPLC and fast atom bombardment mass spectrometry (FABMS) The HPLC chromatogram showed that the purities of peptides were more than 90%, and FABMS showed the correct molecular ion for the peptides The control peptide was synthesized using automatic solid-phase peptide synthesizer (Pioneer, Perspective Biosystem, Foster, CA, USA) using Fmoc chemistry with PAL resin The Fmoc-protected amino acids were obtained from Novabiochem All the solvents used in the Pioneer peptide synthesizer were obtained from Applied Biosystems Peptide was purified by preparative HPLC (Waters 600 HPLC system), on a reversed-phase C18 column (Inertsil, ˚ 10 · 250 mm, lm, 300 A) with a linear gradient of solvent A (0.1% (v/v) trifluoroacetic acid/H2O) and solvent B (0.1% (v/v) trifluoroacetic acid/acetonitrile) The peptides were detected by UV absorbance at k ¼ 215 and 280 nm The purity of each peptide was verified by an analytical HPLC (Shimadzu LC-10AT VP) using a reverse-phase C18 column (Lichrosorb RP18, 4.6 · 200 mm, 10 lm) with the same solvent system as in the preparative HPLC The molecular mass of the peptide was determined by using electro-spray ionization mass spectrometry (ESI-MS, Finnigon MAT) Antibodies Fluorescence-conjugated monoclonal antibody to human CD58 (FITC-anti-CD58) and CD2 (FITC-anti-CD2) were Cell lines CD2 detection and flow-cytometry assay To detect CD2 expression, 106 Jurkat cells were washed with 0.5% (w/v) BSA/NaCl/Pi, and incubated with FITCCD2 monoclonal antibody (mAb) for h at 37 °C After washing three times with 0.5% (w/v) BSA/10 mM Hepes/ NaCl/Pi, the cells were fixed using 1% (v/v) paraformaldehyde/NaCl/Pi and analyzed with a flow cytometer (FACScan apparatus, Becton Dickinson) equipped with the CELL QUEST software Ten thousand cells were counted for every sample during acquisition Inhibition of antibody binding MOLT-3 cells were grown and activated with 0.2 lM of phorbol 12-myristate-13-acetate (PMA) (Sigma) in 75 cm2 tissue culture flasks at 37 °C in a saturating humidified atmosphere of 95% air and 5% CO2 Cells were pelleted at %100 g for min, and re-suspended in serum-free medium to reach a cell count of 2.5 · 106 per mL Peptide stock solution was prepared in phosphate buffered saline (NaCl/Pi) and 0.25% (v/v) dimethylsulfoxide Cell suspensions and peptide solutions (80, 200 and 500 lM) were aliquoted into a 48-well cell culture cluster and incubated at 37 °C for h At the end of incubation, unbound peptide was removed by washing with 10 mM Hepes/NaCl/Pi FITC-anti-CD58 was added to the cell pellets, followed by incubation on ice for 45 After washing three times with 10 mM Hepes/ NaCl/Pi, the cells were fixed using 2% (v/v) paraformaldehyde/NaCl/Pi and analyzed with a flow cytometer (FACScan, Becton Dickinson) equipped with CELL QUEST software Binding of FITC-anti-CD58, following incubation with Fc blocker (Biodesign International) was used as a positive control Median values of fluorescence intensity were taken as the binding intensities As many as 10 000 cells were counted for every sample during acquisition, and each experiment was performed in triplicate The control histogram (cells without peptide treatment) was placed within 100–101 on the log scale of FL1-Height by adjusting the FL1 detector The data were represented as their relative inhibition or enhancement to the positive control Ó FEBS 2004 Design of peptides for T-cell adhesion inhibition (Eur J Biochem 271) 2875 Fig Sequence alignment of rat CD2 and human CD2 (hCD 2; CLUSTALW alignment) Residues of domain and are shown #, Interface contact residues in the hCD – hCD 58 structure; *, residues in the interface; b-turn regions are in bold letters; designed 12 amino acid residue peptide sequences are underlined E-rosetting Sheep red blood cells (SRBCs) were isolated by centrifuging sheep blood in Alsever’s solution at 200 g for SRBCs were washed three times with NaCl/Pi and incubated with four volumes of 2-aminoethylisothiouronium hydrobromide (AET) solution (Sigma) at 37 °C for 15 The cells were washed three times in NaCl/Pi, and resuspended in RPMI-1640 containing 20% fetal bovine serum to give a 10% suspension For use, the cell suspension was further diluted : 20 (0.5%) with medium Serial dilutions of peptides in NaCl/Pi were added to 0.2 mL of 0.5% (w/v) AET-treated SRBCs, and incubated at 37 °C for 30 After that, 0.2 mL of Jurkat cell suspension (2· 106 per mL) was added to the mixture, and incubated for another 15 The cells were pelleted by centrifugation (200 g, min, °C) and then incubated at °C for h The cell pellet was gently resuspended, and the E-rosettes counted with a haemocytometer [16] Cells with five or more SRBCs bound were counted as rosettes At least 200 cells were counted to determine the percentage of E-rosette cells The inhibitory activity was calculated by the following Eqn (1): inhibition %ị ẳ ẵnegative E- rosette %peptide À negative E-rosette %blank Þ= E-rosette %blank Š  100 1ị where, negative E-rosette %peptide ẳ (Jurkat cells without formation of E-rosettes/total Jurkat cells) · 100% Lymphocyte-epithelial adhesion assay Caco-2 cells were used between passages 50 and 60 and were plated onto 48-well plates at % · 104 cellsỈwell)1 When the cells reached confluency, the monolayers were washed once with MEM-a Jurkat cells were labeled the same day as the adhesion assay by loading with lM fluorescent dye biscarboxyethyl-carboxyfluorescein (BCECF-AM) at 37 °C for h Peptide dissolved in MEM-a was added at various concentrations to Caco-2 cell monolayers After incubation at 37 °C for 30 min, the labeled Jurkat cells (1 · 106 cellsỈwell)1) were added onto the monolayers After incubation at 37 °C for 45 min, nonadherent Jurkat cells were removed by washing three times with NaCl/Pi, and the monolayer-associated Jurkat cells were lysed with 2% (v/v) Triton X-100 in 0.2 M NaOH Soluble lysates were Ó FEBS 2004 2876 L Jining et al (Eur J Biochem 271) transferred to 96-well plates for reading with a microplate fluorescence analyzer Cell viability assay Peptides which exhibited effects on Jurkat/Caco-2 adherence were tested by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide (MTT) assay [17] to determine if their effects were due to frank toxicity A final concentration of 180 lM peptide was added to Caco-2 or Jurkat cells for or h, which is the time of exposure of Caco-2/Jurkat cells during the adherence assay The cell viabilities were validated by incubating with mgỈmL)1 MTT at 37 °C for h The MTT-labeled cells were lysed by dimethylsulfoxide and the absorbance was measured with a microplate reader at a wavelength of 570 nm NMR spectroscopy The samples for the NMR spectra of the peptide were prepared by dissolving mg of the peptides in 0.5 mL of 90% H2O/10% D2O For pH titration experiments, the pH of the solution was varied by the addition of DCl or NaOD (pH was not corrected for isotopic effects) The temperature dependence of the amide proton chemical shift was measured by collecting data from 283–303K in steps of 5K using a variable temperature probe The one- and two-dimensional NMR experiments were performed and processed on 300 MHz and 500 MHz Bruker DRX spectrometers equipped with a 5-mm broad-band inverse probe, at a proton frequency of 300.3414 MHz and 500.134 MHz, respectively, using XWINNMR version 1.0 software Spectra were acquired at 298K unless otherwise specified TOCSY [18], DQF-COSY [19] and rotating frame nuclear Overhauser spectroscopy (ROESY) [20] and NOESY [21] experiments were performed by presaturation of water during relaxation delay Data were collected by the TPPI method [22] with a sweep width of 5000 Hz ROE cross-peak volumes were measured using ROESY spectra with 300 ms spin-lock times and NOESY cross-peak volumes for hexapeptides were measured at 200 ms mixing time Coupling constants (3JHNa) were measured from the DQF-COSY spectrum Intensities were assigned as strong, medium and weak with upper and lower boundaries of distance for dNa (i, i), daN (i, i +1) and dNN (i, i +1), 1.9–3.0, 2.2–3.6 and ˚ 3.0–5.0 A, respectively [23] Side chain protons were not stereospecifically assigned; hence, ROE/NOE restraints for the side chain protons were calculated by considering pseudoatoms [23] performed for 20 ps to explore several possible conformations that the peptide can acquire The trajectory from high temperature dynamics was analyzed for similarities between the structures by comparing the root mean square deviations (rmsd) between each possible pair of structures, and was divided into several conformational families The average structure was taken from each family and chosen as starting structures for the calculation of corrected interproton distances from ROESY intensities using Matrix Analysis of Relaxation for Discerning the Geometry of an Aqueous Structure (MARDIGRAS) [27], which takes into account TOCSY contributions for the calculated intensities in the ROESY spectrum MARDIGRAS runs with correlation time (sc) of 0.25, 0.35, 0.45, 0.55 and 0.65 ns were performed with coupling constants calculated from starting model and observed 3JHNa The correlation time was expected to be in the range 0.25–0.65 ns as the observed intensities in 2D NOESY spectrum of this peptide were almost zero The interproton distances were calculated ˚ based on the distance of 1.78 A between the two GlyHa protons in peptide cVY and PheHb protons in peptide cER, respectively For cyclic hexapeptides NOESY crosspeaks were observed at 200 ms mixing time and GlyHa protons were used for calibration After high temperature dynamics with NOE constraints the folded peptide was cyclized by peptide bond to arrive at the starting structures for cyclic hexapeptides In the case of ROESY spectrum for 12 amino acid residue cyclic peptides (12-mers), the corrected interproton distances were used for subsequent calculation of the structure Each structure obtained during high temperature dynamics was then slowly cooled down to 400 K Each structure was then soaked with water molecules, followed by MD simulations at 300 K with all the ROE/NOE constraints These structures were further energy minimized with solvent molecules using the steepest descents and conjugate gradient methods until the rms ˚ derivative was 0.3 kcalỈmol)1ỈA)1 The resulting structures were analyzed again by MARDIGRAS, and the final structures chosen when two criteria were fulfilled: (a) the conformation of backbone had an interproton error of less ˚ than 0.2 A compared to upper and lower boundaries of distances from ROE/NOE data and (b) the conformation had / angles within 30° of the calculated /-values from JHNa [28] The final structures which satisfy most of the NMR distance constraints were clustered together based on the rms deviation of the backbone atoms and the structures which had similar NOE/ROE violations were clustered together as one family Each family/cluster had 10–12 structures An average structure was also chosen from this family as representative structure Determination of peptide structures Conformational space was searched for the peptides using the DISCOVER program version 2000 (Accelrys Inc., San Diego, CA, USA) to identify conformations consistent with the experimental ROE and coupling constant data [24,25] Briefly, the linear peptide was subjected to MD simulations in vacuo at 300K with ROE and disulfide bond constraints [26] The resulting structure was cyclized by forming disulfide bonds The cyclic structure obtained was slowly heated to 900 K in steps of 100 K dynamics for ps duration at each step At 900 K, MD simulations were Modeling of the peptide-CD58 complexes Complexes of CD58–CD2 peptide were generated by docking studies of CD2 peptide to CD58 protein crystal structure All docking studies were performed with the AUTODOCK program [29] (version 3.0) The coordinates of peptides were retrieved from the NMR determined structure (studies presented in this paper) and the coordinates of ligated hCD58 were retrieved from the Protein Data Bank (accession code 1qa9; the monomer of hCD58 was unmerged from the complex of hCD2–hCD58) [30] Ó FEBS 2004 Design of peptides for T-cell adhesion inhibition (Eur J Biochem 271) 2877 Fig Ribbon diagram of crystal structure of CD2–CD58 complex and crystal structure of rat CD2 (A) Ribbon diagram of crystal structure of CD2–CD58 (LFA-3) complex Starting positions of peptides for docking studies are shown in the figure The residues of hCD2 that are in b-turn region are shown as red sticks Tyr86 from CD2 is shown in green Residues from CD58 that are important in the interaction of CD2–CD58 are shown in the following colors: Lys32, Glu25 (purple); Asp33, Lys29, Glu37 (blue); Lys30 (magenta) Residues, Asp33 and Lys29 were shown to be important in binding to peptides from CD2 in docking studies (B) Crystal structure of ratCD2 Residues in the b-turn region are shown as sticks and labeled Hydrogen atoms were added to the protein using INSIGHTII (Accelrys Inc., San Diego, CA, USA) The appropriate partial atomic charges were assigned by consistent valence force field (Cvff) To eliminate the steric hindrance between peptide and protein, and to relax the hydrogen added to the protein, the peptide and protein were merged and minimized before docking Atomic solvation parameters and fragmental volumes were assigned to the protein atoms by the auxiliary program, ADDSOL Affinity grid files were generated using the other auxiliary program, AUTOGRID The dimension of the grid box was chosen to cover the whole ˚ protein with grid-point spacing of 0.375 A and centered at the positions describe below As there are two major cavities in the top and bottom of hCD58 besides the binding sites in the hCD2–hCD58 complex, the starting positions of peptides were generated at three sites on CD58 protein surface (Fig 2A) The parameters were set as the default values of the AUTODOCK Lamarckian genetic algorithm First, a randomized rigid docking (blind docking) was performed and the conformers with lowest energy or in significant clusters were chosen to perform further docking studies with flexible docking During flexible docking, the dihedrals of backbone of the ligand were kept rigid, whereas the dihedrals of side chain were allowed to rotate After docking, all structures generated were assigned to clusters based on a tolerance ˚ of A all-atom rmsd from the lowest energy structure The energies were listed in the increasing order of energy If the ˚ rmsd of a structure is less than A compared to the lowest energy structure in that starting position, that was grouped together with the lowest energy structure forming a cluster of structure The clusters were ranked by the lowest energy representative of each cluster Only low energy structures with more number of conformers in each cluster were used for final analysis Results Biological activity of the peptides 12-mer linear and cyclic peptides The inhibitory activities of the peptides designed from rat CD2 were assayed by three methods In the first method, the inhibition of anti-CD58 binding to CD58 expressed on the surface of MOLT-3 cells was evaluated Figure shows that the peptide lVY enhanced the binding of FITC-anti-CD58 in a concentration-dependent manner, while the peptides lER, cVY and cER inhibited antibody binding Compared with the two cyclic peptides, linear peptide lER displayed less inhibitory activity, inhibiting only 6% at 500 lM The peptide cER showed better inhibitory activity (18% at 200 lM) compared to peptide cVY (7%) A second method, E-rosetting, was carried out to test the biological activity of peptides [16] E-rosetting is the most widely used method to identify T-cells by CD2– CD58 interaction SRBCs express CD58 protein, while Jurkat leukemic T-cells express CD2 protein on their surface The ability of Jurkat cells to express CD2 was 2878 L Jining et al (Eur J Biochem 271) Fig Inhibition or enhancement of FITC-labeled CD58-antibody binding to MOLT-3 cells by synthetic peptides derived from CD2 examined by FACS MOLT-3 cells were activated by 0.2 lM PMA to induce CD58 expression FITC-anti-CD58 was added to the peptidetreated cells, followed by a further incubation Binding of FITC-antiCD58, following incubation with Fc blocker was used as a positive control Median values of fluorescence intensity were taken as the binding intensities As many as 104 cells were counted for every sample during acquisition The control histogram (cells without peptide treatment) was placed within 100–101 on the log scale of FL1-height The data were represented as their relative inhibition or enhancement to the positive control Each data point represents the mean of triplicate assay at different peptide concentration (lM) Fig Inhibition of E-rosette formation by synthetic peptides derived from CD2 protein Peptides were added to AET-treated Sheep Red Blood Cells (expressing CD58 protein) first and then an equal amount of Jurkat cells (expressing CD2 protein) were added later The cells were pelleted by centrifugation and incubated at °C The cell pellet was resuspended gently before counting the E-rosettes Cells with five or more SRBCs bound were counted as rosettes At least 200 cells were counted to determine the percentage of E-rosette cells Values are percentage inhibition of peptide-treated cells and expressed as the mean of three independent experiments measured by flow-cytometry assay Binding of Jurkat cells to SRBCs due to CD2 and CD58 interaction results in the formation of E-rosettes The ability of each of the designed CD2 peptides to inhibit CD2–CD58 interaction was evaluated by inhibition of E-rosette formation between Jurkat cells and SRBCs As depicted in Fig 4, the CD2 peptides showed 30–40% inhibitory activity at 200 lM When the concentration of the peptide was decreased, the inhibitory activity of the peptide was correspondingly decreased Even at 50 lM, peptide cVY displayed nearly 30% activity Among the four peptides (12-mers) studied, cVY showed the highest inhibitory Ó FEBS 2004 Fig Inhibition of lymphocyte-epithelial adhesion by synthetic peptides derived from CD2 protein CD58 and CD2 expressing on Caco-2 cells and Jurkat cells, respectively, were pre-examined Peptides were added to the confluent Caco-2 monolayer and then the BCECF labeled Jurkat cells were added to the mixture After the incubation for 45 at 37 °C, nonadherent Jurkat cells were removed by washing with NaCl/Pi and the monolayers associated Jurkat cells were lysed with Triton X-100 solution Soluble lysates are transferred to 96-well plates for reading in a microplate fluorescence analyzer Values are showed in the percentage inhibition of peptide-treated cells and expressed as the mean of three independent experiments activity of 40% at 100 lM concentration Both linear and cyclic ER peptides (lER and cER) showed similar inhibitory activities, whereas in the case of VY peptides, the cyclic cVY peptide showed increased activity compared to its linear counterpart lVY Correspondingly, a control peptide showed less than 5% inhibitory activity by the E-rosetting assay As a third method, inhibition of adhesion between Caco-2 cells and Jurkat cells was used to evaluate the biological activity of peptides designed Caco-2 cells express CD58 while Jurkat cells express CD2 protein The inhibitory activity observed between Caco-2 cells and Jurkat cells provides evidence that the peptides designed from CD2 can inhibit the adhesion between the heterotypic cells The inhibitory activities of designed CD2 peptides were measured by using fluorescently labeled Jurkat-cells by fluorescence spectrometer The activities of the peptides from CD2 in the heterotypic cell adhesion assay are shown in the Fig along with a control peptide Among the 12-mers, cER, lVY, cVY showed 30–50% inhibitory activity at 90 lM concentration The cyclic peptide cVY showed % 50% inhibition at 90 lM concentration However, as the peptide concentration was decreased to 10 lM, cVY showed less than 15% activity whereas lER and cER peptides retained 20% inhibitory activity Compared to linear peptides, cyclic peptides showed a slight increase in activity A control peptide showed less than 5% inhibitory activity at three different concentrations studied These peptides were also tested for their toxicity using the MTT assay [17] All the four peptides tested in the study resulted in 90–100% viability indicating that these peptides were not toxic to cells and the inhibition data observed were not due to changes in the cells arising from peptide toxicity Cyclic hexapeptides In order to understand the amino acid residues involved in the biological activity and to study the effect of reducing the chain length of peptides on Ó FEBS 2004 Design of peptides for T-cell adhesion inhibition (Eur J Biochem 271) 2879 biological activity, cyclic hexapeptides were designed These hexapeptides were truncated forms of the 12-mers described above, and were cyclized by peptide bonds The inhibitory activities of the cyclic hexapeptides are shown in Figs and In the E-rosetting assay (Fig 4), peptide cEL showed % 35% activity at a concentration of 200 lM, an increase in inhibitory activity compared to linear and cyclic 12-mer ER peptides However, the VY cyclic hexapeptide (cYT) lost activity upon truncation Similar trends were observed in the heterotypic adhesion assay of cyclic hexapeptides (Fig 5) The cEL peptide showed increased activity (50% at 90 lM), whereas cYT showed drastically diminished inhibitory activity compared to 12-mer VY peptides NMR structure determination The three-dimensional structures of the cyclic peptide were determined based on the NMR data of the cyclic peptides The one dimensional 1H NMR spectrum of the peptides cVY, cER, cEL and cYT showed good dispersion of the chemical shifts and the coupling patterns, indicative of a stable major conformer at the experimental temperature The structure of peptide cER NMR data of cER indicated the possibility of the b-turn structure in peptide cER The dNN (i, i +1) cross peaks between Gly4-Ser5 and the stronger daN (i, i +1) cross peaks between Arg3-Gly4 suggesting a possible b-turn at Glu2-Arg3-Gly4-Ser5 (Fig 6A) The two consecutive dNN (i, i +1) cross peaks between Leu7-Val8 and Val8-Ala9 suggest a tight b-turn at Leu7-Val8-Ala9-Glu10 The temperature-dependent coefficient of the chemical shift data indicated that the NH of Glu10 (Dd/DT ẳ )3.0 p.p.b.ặK)1) is intramolecular hydrogen bonded, suggesting a stable b-turn of Leu7Val8-Ala9-Glu10 [23] The temperature coefficient of chemical shift of Ser5 amide resonance showed a value >)3.0 p.p.b.ỈK)1 suggesting an open b-turn conformation at Glu2-Arg3-Gly4-Ser5 From ROE-restrained MD simulations and energy minimization, four families of conformers that satisfied the NMR data were obtained An average structure was taken from each family to Fig Summary of ROEs for peptides cER (A) and cVY (B) The thickness of bars indicate the intensity of ROE cross-peaks, and were assigned as strong, medium and weak ˚ represent the family Based on ROE violation > 0.2 A and allowed values of /, w in the Ramachandran map, only one family of structure that was consistent with NMR data was chosen to represent the conformation of peptide cER A family of low energy structures that were consistent with NMR data representing the conformation of cER is shown in Fig 7A The structure exhibits a well-defined b-turn spanning residues Glu2 to Ser5 The rmsd of the backbone atoms of the 12 structures in the chosen family was compared with the average structure in the same family It was found that the rmsd of all the backbone atoms in the ˚ peptide was 1.02 A, while that of residues at turn region ˚ Glu2-Arg3-Gly4-Ser5 was 0.32 A, indicating the stable nature of the b-turn conformation The /, w angles around Arg3-Gly4 and Val8-Ala9 of the structures showed the possibility of a type II b-turn at Glu2-Arg3-Gly4-Ser5 residues and a type III b-turn at Leu7-Val8-Ala9-Glu10, respectively [31] Therefore, the structure of peptide cER consists of two b-turns, located at the N- and C-termini A comparison of the b-turn structure of cER with the similar region in the crystal structure of rat and human CD2 was carried out In the rat CD2 crystal structure, the b-turn structure was exhibited by residues Arg37-Gly38-Ser39Thr40 The peptide cER displayed a b-turn structure with shift in one residue compared to the protein from which it is derived In ratCD2, the type of b-turn observed at Arg37Gly38-Ser39-Thr40 is a type II¢ b-turn whereas in cER peptide the b-turn is type II [31] This is due to the position of Gly amino acid in the b-turn which is flexible In human CD2, similar region (Fig 1) has a b-turn is around Thr38Ser39-Asp40-Lys41 and the turn observed was type I b-turn An additional b-turn was observed in the cER peptide structure at the Leu7-Val8-Ala9-Glu10 sequence compared with the corresponding part in rat CD2 (Fig 7A) The structure of peptide cVY Several lines of NMR evidence were consistent with the existence of a b-turn in the cVY peptide at Ser4-Thr5-Asn6-Gly7: (a) the Gly7 enantiotopic protons showed Dd-values > 0.4 p.p.m indicating the rigidity around this residue; (b) the dNN (i, i +1) cross peaks and medium range distance daN (i, i +1) cross peaks between Thr5-Asn6 and Gly7-Thr8 (Fig 6B); (c) the 3JHNa of Thr5 and Asn6 were close to those expected for a type I b-turn (i.e 3JHNa values of Hz and Hz are expected for the i +1 and i +2 turn residues, respectively); (d) the temperature dependence of the chemical shift data indicates that the NH of Gly7 (Dd/DT ẳ )2.9 p.p.b.ặK)1) was solvent shielded or intramolecular hydrogen-bonded Molecular modeling studies resulted in seven families of peptide cVY structures that best fit the ROE and dihedral angle data The family/cluster of structures that had ROE ˚ violation of £ 0.2 A was used to represent the final structure To check the convergence, the structures in each family were superimposed on the average structure in each family All structures presented a well-defined b-turn spanning residues Ser4-Gly7 [31] Lack of convergence was observed in the first residue and the last three residues in the peptide sequence The average rmsd of the backbone atoms of 12 structures compared to the average structure was ˚ 0.98 ± 0.35 A, while the average rmsd at the residue ˚ Ser4-Thr5-Asn6-Gly7 was 0.34 ± 0.06 A, indicating the Ó FEBS 2004 2880 L Jining et al (Eur J Biochem 271) Fig Superposition of 12 NMR-MD derived structures for the cyclic peptides with average structure for (A) cER and (B) cVY Only heavy atoms are shown for clarity The residues which are involved in b-turn conformation are labeled stable nature of the b-turn conformation A representative structure of peptide cVY families are shown in the Fig 7B The /, w angles around Thr5-Asn6 showed that the structure of the peptide deviated slightly from the ideal type I b-turn [31] A comparison of the b-turn structure of cVY with the similar region in the crystal structure of rat and human CD2 was carried out The type-I b-turn observed in cVY around Ser4-Thr5-Asn6-Gly7 was similar to that in rat CD2 crystal structure (Ser87-Thr88-Asn89Gly90) In human CD2, the similar region Asp87-Thr88Lys89-Gly90 exhibits a type I b-turn Superimposition of bturn regions from rat CD2 and human CD2 crystal structure with the b-turn region of cVY peptide indicated that the rmsd of the backbone atoms for four residues was ˚ less than A Thus, the overall backbone and side chain topologies of the turn region mimic those of rat CD2 Structure of cyclic hexapeptides – cEL peptide The chemical shifts of amide resonances of peptide cEL were well dispersed over a region of 1.2 p.p.m indicating the stable conformation of the peptide The Gly3 enantiotopic Ha protons were well separated in chemical shift, indicative of stable conformation around Gly3 The NH-NH region of the NOESY data showed connectivity between the amides of Glu1-Leu6, Glu1-Arg2, Gly3-Ser4, Ser4-Thr5 and Arg2Lue6 which is suggestive of a b-turn in the peptide and the proximity of amide protons due to compact nature of the structure However, the coupling constant of most of the amide protons was in the range of 6–8 Hz, suggestive of rapidly interconverting conformers that coexist in solution The temperature coefficients of chemical shift of amides Glu1 and Thr5 were near 3.8–4.0 p.p.b.ỈK)1 which may be due to intramolecular hydrogen bonding or solvent-shielded amide protons of Glu1 and Thr5 Molecular modeling studies indicated that the peptide exhibits two b-turns, i.e one at Arg2-Gly3-Ser4-Thr5 and the other at Thr5-Leu6Glu1-Arg2 A representative structure of cEL is shown in the Fig 8A The b-turn at Arg2-Gly3-Ser4-Thr5 was type II¢ b-turn as observed in the case of rat CD2 crystal structure Superimposition of backbone atoms of the residues in the b-turn region of rat and human CD2 (similar region) with cEL peptide b-turn region (Arg2-Gly3Ser4-Thr5) indicated that the rmsd of the backbone atoms ˚ ˚ was 0.67 A with rat CD2 and 1.2 A with human CD2 Thus, the peptide mimics the b-turn region of the protein from which it is derived from The peptide model also showed intramolecular hydrogen bonds between NH of Thr5 and CO of Arg2, as well as NH of Arg2 and CO of Thr5 Structure of cyclic hexapeptides – cYT peptide The NMR data of the cYT peptide were indicative of its flexible nature The chemical shift dispersion of amides was less than p.p.m., and the Gly5 Ha enantiotopic protons had a degenerate chemical shift usually indicative of flexible structure Amide region of the NOESY data suggested weak intensity NOE connectivities between Tyr1-Ser2, Ser2Thr3, Asn4-Gly5 and Gly5-Thr6 Most of the coupling constants were in the range of 6–8 Hz Ser2 NH showed a temperature coefficient of chemical shift value of 2.2 p.p.b.ỈK)1 which may be due to the hydrogen-bonded amide of Ser2 The ROE-based molecular modeling data on cYT resulted in a structure shown in Fig 8B The dihedral angles around Thr6 and Tyr1 exhibited the dihedral angles of a type I b-turn, and Gly5 exhibited c-turn dihedral angles The overall structure of the peptide was open/flexible as indicated in the superimposed 12 structures shown in Fig 8B Docking Recently, it has been shown that AUTODOCK can be used to dock peptide to proteins without prior knowledge of the binding site [32] Peptides derived from CD2 presumably modulate cell-adhesion by binding to CD58, hence inhibiting CD2–CD58 interaction Therefore, docking studies of peptides to CD58 protein were carried out in order to understand peptide–protein interactions by using autodock [29] In the docking of CD2 peptides to CD58 Ó FEBS 2004 Design of peptides for T-cell adhesion inhibition (Eur J Biochem 271) 2881 Fig Superposition of 12 NMR-MD derived structures for the cyclic hexapeptide with average structure for (A) cEL and (B) cYT Only heavy atoms are shown for clarity Table Peptide cER: CD58 docking results starting from the potential binding sites out of 100 runs Only the clusters with the lowest energy of docking are listed Starting position Final, low energy position of the peptide after docking CC¢ sheet Top cavity Top Cavity Top cavity Bottom cavity Top cavity protein, the grid was centered at three possible binding sites, ˚ with a 110 · 110 · 110 A cubic area to cover the whole CD58 protein Three positions were chosen on the protein surface of CD58 (Fig 2A), i.e (a) position 1, which is a CC¢ sheet and the interface of CD2–CD58 interaction; (b) position 2, the top cavity where the turn region from CD2 interacts with CD58; (c) position 3, the bottom cavity where a turn region of CD2 interacts with CD58 First, a randomized rigid docking (blind docking) was performed and the conformers with lowest energy or in significant clusters were chosen to perform further docking studies with flexible docking Peptide cER–CD58 complex The automated molecular docking calculations produced several possible binding sites and conformations for the peptide The conformation corresponding to the low energy of docking was chosen as the possible binding site The results from the docking studies of cER peptide-CD58 protein are shown in Table Although, different starting positions were chosen for the cER peptide on the CD58 protein surface, the final low energy docked conformers of the peptide were near the top cavity region on the protein Thus, the most probable binding site of cER peptide on CD58 is possibly near the top cavity Table lists the residues involved in intermolecular hydrogen bonding in the cER peptide and CD58 protein interface It is very clear that most of the residues that exhibit b-turn structure in the peptide (Glu2-Arg3-Gly4Ser5) were involved in hydrogen bonding with the protein receptor (CD58) The Ser5 residue in the turn region of peptide cER interacts with the key residue Asp33 of CD58 that is important in adhesion Thr6, the flanking residue of the b-turn region also forms a hydrogen bond with Asp33 Cluster Rank Lowest docked energy (kcalỈmole)1) Number of conformations in the cluster 2 )15.7 )15.5 )15.2 )13.2 )15.5 )14.9 1 Table Amino acid residues forming hydrogen bonds in the cER–CD58 – interface The residues in the turn region of peptide cER and in CD58 which are important for the CD2–CD58 interaction are shown in bold italic typeface Peptide cER ()15.5 kcalỈmol)1) CD58 Residue Atom Residue Atom Ser5 Phe11 Ser5 Arg3 Ser5 Thr6 Arg3 Phe11 Arg3 Hc O HN O O O NH1 O He Lys30 Lys30 Gln31 Gln31 Asp33 Asp33 Ser69 Ser70 Glu72 O Hf Od He HN Hd O Hc Oe which was shown to be important in CD2–CD58 interaction Peptide cVY–CD58 complex Docking studies of the cVY peptide and CD58 protein revealed that structures with low energy of docking were around the CC¢ sheet of CD58 protein (Table 4, Fig 2A) The CC¢ sheet is the interface of CD2–CD58 interaction Different starting positions yielded low docked energy conformations in the CC¢ sheet region, and hence the most possible binding site may be near the CC¢ sheet The amino acid residues that are involved in the cVY peptide–CD58 protein interaction are shown in Ó FEBS 2004 2882 L Jining et al (Eur J Biochem 271) Table Peptide cVY: CD58 docking results starting from the potential binding sites out of 100 runs Only the clusters with the lowest docked energy are listed CC¢ sheet CC¢ sheet Top cavity CC¢ sheet Bottom Cavity Cluster Rank CC¢ sheet Table Amino acid residues forming hydrogen bonds in the cVY– CD58 interface The residues in the turn region of peptide cVY and in CD58 which are important for the CD2–CD58 interactions are shown in bold italic typeface Peptide cVY ()10.7 kcal/mol) CD58 Residue Atom Residue Atom Tyr3 Thr5 Asn6 O Oc Hd Lys29 Lys29 Asp33 Hz Hz Od the Table It is very clear that the residues in the b-turn (Ser4-Thr5-Asn6-Gly7) region of the cVY peptide are involved in hydrogen bonding interaction with key residues Asp33 and Lys29 of CD58 Cyclic hexapeptide–CD58 docking The NMR-derived cyclic hexapeptide structures were used to perform docking studies of peptide–CD58 protein interaction Docking studies of cEL starting from different possible positions on CD58 resulted in low energy structures that were clustered around the top cavity of CD58 protein The lowest energy docked structure indicated that the Arg2 side chain of the peptide formed intermolecular hydrogen bonding with key residue Lys34 on CD58 Ser4 (backbone NH) and Thr5 (side chain) in the peptide were also involved in intermolecular hydrogen bonding with Gln31 and Glu72 of the CD58 protein, respectively Thus, the involvement of key residue Lys34 in CD58 protein with hydrogen bonding to peptide may result in inhibition of CD2–CD58 interaction The cYT peptide did not show binding site specificity The lowest energy clusters obtained after docking calculations were near the starting position of the peptide The low energy docked structures also indicated that Gly5 carbonyl carbon and Ser2, Asn4 side chains were involved in intermolecular hydrogen bonding with the protein However, none of the hydrogen bonds were with the key residues that are essential for CD2–CD58 interactions on the CD58 protein This supports the low biological activity of cYT observed in the E-rosetting and heterotypic adhesion assays Lowest docked energy (kcalỈmole)1) Number of conformations in the cluster 2 Starting position Final, low energy position of the peptide after docking )10.7 )10.0 )10.4 )10.0 )9.7 )9.4 1 Discussion Inhibition of CD2–CD58 interaction has important implications in controlling immune responses in autoimmune diseases In this study, we designed 12-mer linear and cyclic peptides (lVY, cVY, lER, and cER) as well as cyclic hexapeptides (cEL and cYT) that were derived from the rat CD2 sequence Initially, the design of small molecular inhibitors was based on the crystal structure of rat CD2 [33– 37] The CD58 (LFA-3) binding ability of CD2 is known to reside in domain-1 of CD2 protein CD2 peptide mapping and mutagenesis indicated that the binding surface of CD2 consists of b-sheet formed by strands GFCC¢C¢¢ The crystal structure of CD2 (Fig 2B) indicated that the rather flat b-sheet surface does not provide a complementary shape to bind to CD58, and hence does not have well-defined epitopes to design small molecular inhibitors The structure of CD2 is similar to CD4 and other IgSF molecules In the D1 domain of CD4, the b-turn near CC¢ appears to be important for binding to its receptor [38] b-Turn peptides based on CD4 have been shown to be effective in inhibiting CD4 interactions [38] Analysis of the crystal structure of CD2 revealed that on either side of the binding surface of CD2, there are b-turns which stabilize the b-strands Thus, we hypothesized that these b-turns may serve as good surface epitopes for the design of peptides to inhibit CD2– CD58 interactions Meanwhile, the crystal structure of human CD2–CD58 became available [30] Examination of the CD2–CD58 crystal structure indicated that the interface of the CD2–CD58 complex has poor shape complementarity in the center region of interaction (Fig 2A) Most of the interaction is via salt-bridges with charge neutralization and hydrogen bonds Furthermore, the b-strand surface of CD2 that interacts with CD58 is stabilized by b-turns on either side These b-turn regions seem to be important in holding the CD2–CD58 interface intact with b-sheet and salt-bridges Rat CD2 and human CD2 share sequence similarity (Fig 1) The residues in the b-turn of rat CD2 sequence are Arg37-Gly38-Ser39-Thr40 and Ser87-Thr88Asn89-Gly90, while those in human CD2 are at Thr38Ser39-Asp40-Lys41 and at Asp87-Thr88-Lys89-Gly90 Lys41 of the b-turn at Thr38-Ser39-Asp40-Lys41 is involved in the hydrogen bonding interaction with CD58 Similarly, in the b-turn at Asp87-Thr88-Lys89-Gly90, the Gly90 backbone carbonyl carbon is involved in hydrogen bonding interaction with CD58 The flanking residue Tyr86 of the b-turn at Asp87-Thr88-Lys89-Gly90 has been shown to be Ó FEBS 2004 Design of peptides for T-cell adhesion inhibition (Eur J Biochem 271) 2883 important in CD2–CD58 interaction [3] Therefore, we hypothesized that the stable b-turn conformation mimicking the native CD2 surface-binding region with CD58 may be important for inhibitory activity of the peptide Since human CD2 shares sequence homology with rat CD2, we designed the peptide based on the b-turn region of rat CD2 sequence These peptides have been shown to inhibit cell– cell adhesion Several lines of evidence suggest that peptides from rat CD2 may bind to human CD58 Human-CD2 and rat-CD2 have high 3D structure similarity Comparison of 3D structures of hCD2 from CD2–CD58 complex [30] and the structure of rat CD2 [36] using DS modeling (Accelrys Inc., Sandiego, CA, USA) suggested that the two structures overlap very well The pairwise r.m.s.d of 95 residues (covering the interaction region of hCD2–hCD58) of ˚ hCD2 and rat CD2 was 0.9 A The b-turn regions of hCD2 (Thr38-Ser39-Asp40-Lys41 and Asp87-Thr88Lys89-Gly90) overlap with b-turn regions of rat CD2 (Arg37-Gly38-Ser39-Thr40 and Ser87-Thr88-Asn89Gly90) In CD2–CD58 complex, the residues in the b-turn and flanking residues of this b-turn have been shown to be important in interface contact of hCD2–hCD58 The peptides designed from rat-CD2 sequence are in the b-turn region and they overlap very well with b-turn regions of human CD2 Point mutation data indicated that flanking residue of the b-turn region in human CD2 (Tyr86) is important in binding to CD58 Similarly Gly90 in hCD2 makes a contact with hCD58 Glu36 a flanking residue of the b-turn (Glu36-Arg37-Gly38-Ser39-Thr40) is also an important residue in the interface of hCD2–hCD58 Comparison of hCD2 and rat CD2 3D structures indicated that Tyr86 and Glu36 side chains overlap in hCD2 and rat CD2 and oriented in the same direction in both the structures To stabilize the b-turn structure in the designed peptides, cyclic versions of the peptides were synthesized We chose to cyclize the peptides by disulfide bonds with the introduction of amino acids penicillamine (Pen) and cysteine at the two ends of the peptide sequence Cyclization by disulfide bond is relatively easy and yields good yield after purification Penicillamine with two bulky methyl groups is known to stabilize disulfide bonds [39] Pen is used in the position because in previous work we have been successful in improving conformational stability of the cyclic peptides by using Pen at position [24,40] Initially, 12-mer peptides were designed (Table 1) After preliminary examination of biological activity of 12-mer linear and cyclic peptides, the peptides were truncated to six amino acid residues in order to elucidate the minimum number of critical amino acids necessary for biological activity These hexapeptides were cyclized by amide bonds to stabilize the structure In the hexapeptides, cyclization by disulfide bond was not designed since addition of penicillamine and cysteine to form a disulfide bond will increase the number of amino acids in the peptide The information obtained from truncating the peptides will also be useful in the design of future generation pharmacophores A control peptide (Table 1) was designed to compare the importance of primary and secondary structures in the designed peptides The control peptide sequence was chosen from the Ôhot-spotÕ [3] region of the hCD2–hCD58 interface on the CD2 protein The sequence was then reversed and the important amino acid Tyr86 as well as Tyr81 was replaced by Ala to generate the control peptide sequence shown in Table The abilities of our designed peptides (except peptide lVY) to inhibit binding of anti-CD58 to CD58 protein expressed on the surface of MOLT-3 cells suggested that they may interrupt the interaction between CD2–CD58 In addition, cyclic peptide cVY exhibited higher activity than the linear peptide derived from the same region, supporting the conformational dependence of the peptide inhibition Two different methods of inhibition of cell adhesion were used to evaluate the biological activities of peptides Both the assays indicated that peptides designed from CD2 were potent inhibitors of cell-adhesion While the 12-mer cyclic peptide cVY showed % 45–50% inhibition in the E-rosetting and heterotypic adhesion assays, linear and cyclic ER peptides showed similar inhibition activity In the case of the cVY peptide, cyclization seemed to improve the activity of the peptide Thus, cyclization stabilizes the b-turn conformation in the peptide which may be necessary for biological activity exhibited by the peptides Comparing the antibody binding assay and cell adhesion assay, there is a difference in the trend of activities of the peptides particularly in the case of linear peptide lVY This peptide showed enhancement in the antibody binding assay, whereas in celladhesion inhibition assay, the peptide inhibited cell-cell adhesion While all the three methods used to evaluate biological activities of the peptides in this study were expected to generate similar results, there were differences due to the nature of different cell types used in the experiments The binding sites on the surface of these cells may be unique MOLT-3 and Jurkat cells are derived from lymphocytes, while sheep blood cells are erythrocytes Furthermore, CD58 may have different epitopes to bind to antibody in the antibody binding assay Comparing the concentration of peptides used in inhibition studies by all the three methods, the antibody binding inhibition assay was the least sensitive and required higher concentrations of peptides to observe inhibition Also, the peptide lVY is a linear and may exhibit random structure Overall, the E-rosetting and heterotypic adhesion assays provide concrete evidence that the designed peptides can inhibit the mechanism of cell–cell adhesion Truncation of the 12-mer peptide cER to the cyclic hexapeptide cEL resulted in higher biological activity This provides the evidence that amino acids in the b-turn region and stable conformation of the peptide are important for biological activity Truncation of the cVY peptide to the hexapeptide cYT resulted in drastically reduced biological activity suggesting the lack of stable structure and amino acids important for biological activity of the peptide This suggests that b-turn conformation of the residues in the b-turn region may be important in the inhibitory mechanism, in addition to the primary structure of the peptide To confirm our hypothesis that the b-turn structures are important for the inhibitory activities of the peptides, the structures of the cyclic peptides were determined by NMR The results proved that the stable b-turn conformation exists in the cyclic peptides The b-turn regions in these peptides appeared to be more stable with flexibility at the Ó FEBS 2004 2884 L Jining et al (Eur J Biochem 271) terminal regions, thus reinforcing our hypothesis that the Ôactive coreÕ is located in this turn region or b-turn exposes the important residues to the receptor Moreover, the modeling of the peptide-protein complex suggested that the residues in the b-turn region played a key role in the interaction Molecular modeling studies predicted that the b-turn in the cyclic peptides closely mimics the conformational feature of b-turn in CD2 protein In addition, the binding sites of the peptides to CD58 protein predicted by autodock were near the CD2–CD58 interface, either in the top cavity of CD58 or near the CC¢ sheet which explained the inhibitory activity of the cyclic peptides There was no direct correlation between low energy docked structure of the peptide from docking studies and biological activity in terms of cell-adhesion inhibition by two methods explained Docking studies using autodock not take into account the presence of CD2 receptor in competing with peptides from CD2 to bind to CD58 It provides only probable binding sites with relatively low energy of interaction between peptide and protein analyzed in the study However, docking studies were in agreement with antibody binding inhibition assay Comparison of docking energies of the two cyclic peptides indicated that the cER peptide had low docked energy compared with the cVY peptide Biological activity of the peptides by antibody binding assay suggested that cER peptide inhibits antibody binding better than cVY peptide However, E-rosetting assay and heterotypic cell adhesion assay indicated that cVY exhibits cell-adhesion inhibition better than cER The cER peptide may potentially bind to the CD58 protein as indicated by the docking studies The lowest energy docked structure was at the top cavity On the other hand, cVY peptide has relatively high energy compared to cER, but the position of docking is near the CC¢ sheet which involves many salt bridges and hydrogen bonds in CD2–CD58 interaction Thus, in terms of inhibition of cell adhesion, cVY has more potential since it may directly interrupt the binding mechanism of CD2–CD58 Inhibition of cell adhesion involves interrupting the interaction of key residues at the protein– protein interface If we consider hydrogen bonding interaction between the peptide and CD58, the cER peptide forms hydrogen bonds with Asp33 in the lowest energy docking position Asp33 is one of the key residues in CD2–CD58 protein–protein interaction [3] The peptide cVY is involved in hydrogen bonding interaction with Asp33 and Lys29 on CD58 in the lowest docked energy structure Both Asp33 and Lys29 are very important residues in CD2– CD58 interaction Mutational studies have indicated that removal of these two interactions can result in loss of CD2–CD58 interaction This correlates with higher inhibitory activity of the cVY peptide compared to the cER peptide In the case of cyclic hexapeptide cEL, the lowest energy docked structure showed that cEL hydrogen bonds with Lys34, whereas cYT is not involved in hydrogen bonding with key residues that are important for CD2–CD58 interaction This correlates with the higher inhibitory activity of the cEL peptide and the very low inhibitory activity of the cYT peptide compared to other peptides The flanking residue of the b-turn, i.e Tyr3 of the cVY peptide appears to be important in exhibiting inhibitory activity Mutation of Tyr86 to alanine in the native protein reduces the affinity of the CD2–CD58 complex by more than 1000-fold Thus, Tyr86 on the CD2 surface has been defined as a Ôhot-spotÕ [3], and may represent a useful target for the design of small molecules This may explain the higher inhibitory activity of the cVY peptide (which contains Tyr3) compared to the cER peptide in the E-rosetting and heterotypic adhesion assays However, in the case of cyclic hexapeptide cYT, although it contains an important residue Tyr1, the peptide exhibits low inhibition activity in E-rosetting and heterotypic adhesion assays The peptide cYT is expected to mimic the region with Tyr86 in the protein NMR and molecular modeling studies suggested that cYT does not acquire a stable b-turn structure Cyclization may change the orientation of Tyr1 in the small cyclic peptide which may result in the loss of activity Docking studies of cYT peptide resulted in nonspecific binding to CD58 surface at three positions In conclusion, we have designed peptides from the b-turn region of the CD2 protein that are critical for inhibiting CD2–CD58 interaction NMR studies indicated that the cyclic peptides acquire b-turn structures in solution Cell viability assays clearly suggested that the peptides are not toxic to cells tested in the studies, and represent potential lead compounds for immunomodulation Thus, the designed peptides have been shown to mimic the epitopes on CD2 protein that interact with CD58 Acknowledgements We gratefully acknowledge the Super Computer and Visualization Unit, National University of Singapore for the use of computational facilities; Prof Arthur J Olson, the Scripps Research Institute, USA, for providing us with the AutoDock software This research was funded by an academic research grant (R-148-000-026-112) from the National University of Singapore, Singapore References Bierer, B.E & Burakoff, S.J (1988) T cells adhesion molecules FASEB J 2, 2584–2590 Wang, J.H & Reinherz, E (2000) Structure basis of cell–cell interactions in the immune system Curr Opin Structural Biol 10, 656–661 Kim, M., Sun, Z.Y.J., Byron, O., Campbell, G., Wagner, G., Wang, J.-H & Reinherz, E.L (2001) Molecular dissection of the CD2-CD58 counter–receptor interface identifies CD2 Tyr86 and CD58 Lys34 residues as the functional Ôhot spotÕ J Mol Biol 312, 711–720 Davis, S.J & van der Merwe, P.A (1996) The structure and ligand interactions of CD2: implications for T cell function Immunol Today 17, 177–187 Bachmann, M.F., Barner, M & Kopf, M (1999) CD2 sets quantitative thresholds in T cell activation J Exp Med 190, 1383–1391 Mojcik, C.F & Shevach, E (1997) Adhesion molecules: a rheumatologic perspective Arthritis Rheumatism 6, 991–1004 Kaplon, R.J., Hochman, P.S., Michler, R.E., Kwiatkowski, P.A., Edwards, N.M., Berger, C.L., Xu, H., Meier, W., Wallner, B.P., Chisholm, P & Marboe, C.C (1995) Short course single agent therapy with an LFA-3-IgG1 fusion protein prolongs primate cardias allograft survival Transplantation 61, 356–363 Sultan, P., Schechner, J.S., McNiff, J.M., Hochman, P.S., Hughes, C.C.W., Lorber, M.I., Askenase, P.W & Pober, J.S (1997) Ó FEBS 2004 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Design of peptides for T-cell adhesion inhibition (Eur J Biochem 271) 2885 Blockade of CD2–LFA)3 interactions protects human skin allografts in immunodeficient mouse/human chimeras Nature Biotechnol 15, 759–762 Qin, L.H., Chavin, K.D., Lin, J.X., Yagita, H & Bromberg, J.S (1994) Anti-CD2 receptor and anti-CD2 ligand (CD48) antibodies synergize to prolong allograft survival J Exp Med 179, 341–346 Hirahara, H., Tsuchida, M., Wanatabe, T., Haga, M., Matsumoto, Y., Abo, T & Eguchi, S (1995) Long-term survival of cardiac allografts in rats treated before and after surgery with monoclonal antibody to CD2 Transplantation 59, 85– 90 Sido, B., Otto, G., Zimmermann, R., Muller, P., Meuer, S & ă Dengler, T.J (1996) Prolonged allograft survival by the inhibition of costimulatory CD2 signals but not by modulation CD48 (CD2 ligand) in the rat Transpl Int 9, S323–S327 Sido, B., Otto, G., Zimmermann, R., Muller, P., Meuer, S & ă Dengler, T.J (1997) Modulation of the CD2 receptor and not disruption of the CD2/CD48 interaction is the principal action of CD2-mediated immunosuppression in the rat Cell Immunol 182, 57–67 Aruffo, A & Hollenbaugh, D (2001) Therapeutic intervention with inhibitors of co-stimulatory pathways in autoimmune disease Curr Opin Immunol 13, 683–686 Przepiorka, D., Phillips, G.L., Ratanatharathorn, V., Michele, C.F., Sehn, L.H., Antin, J.H., Lebherz, D., Awwad, M., Hope, J & McClain, J.B (1998) A phase II study of BTI-322, a monoclonal anti-CD2 antibody, for treatment of steroid-resistant acute graftversus-host disease Blood 92, 4066–4071 Branco, L., Barren, P., Mao, S.Y., Pfarr, D., Kaplon, R., Psotema, C., Langermann, S., Koenig, S & Johnson, S (1999) Selective deletion of antigen-specific, activated T cells by a humanized mAb to CD2 (MEDI-507) is mediated by NK cells Transplantation 68, 1588–1596 Albert-Wolf, M., Meuer, S.C & Wallich, R (1991) Dual function of recombinant human CD58: inhibition of T-cell adhesion and activation via the CD2 pathway Int Immunol 3, 1335– 1347 Mosmann, T (1983) Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays J Immunol Meth 5, 55–63 Bax, A & Davis, D.G (1985) MLEV-17-based two-dimensional homonuclear magnetization transfer spectroscopy J Magn Reson 65, 355–360 Rance, M., Sorensen, O.W., Bodenhausen, G., Wagner, G., Ernst, R.R & Wuthrich, K (1983) Improved spectral resolution in ă COSY 1H NMR spectra of protein via double quantum filtering Biochem Biophys Res Commun 117, 479–485 Bax, A & Davis, D.G (1985) Practical aspects of two-dimensional transverse NOE spectroscopy J Magn Reson 63, 207–213 Kumar, A., Wagner, G., Ernst, R.R & Wuthrich, K (1981) Buildup rates of the nuclear Overhauser effect measured by twodimensional proton magnetic resonance spectroscopy J Am Chem Soc 103, 3654–3658 Marion, D & Wuthrich, K (1983) Application of phase sensitive ă two-dimensional correlated spectroscopy (COSY) for measurements of 1H-1H spin-spin coupling constants in proteins Biochem Biophys Res Commun 113, 967–974 Wuthrich, K (1986) NMR of Proteins and Nucleic Acids John ˆ Wiley & Sons, New York Jois, S.D.S., Siahaan, T.J., Tibbetts, S.A & Chan, M.A (1999) A Ca2+ binding cyclic peptide derived from the a-subunit of LFA-1: inhibitor of ICAM-1/LFA-1 mediated T-cell adhesion J Peptide Res 53, 18–29 Sutcliffe, M.J., Structure determination from, NMR data II Computational approaches In NMR of Macromolecules: a 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 Practical Approach (Roberts, G.C.K ed.), pp 359–390 Oxford University Press, New York Srinivasan, N., Sowdhamini, R., Ramakrishnan, C & Balaram, P (1990) Stereochemical modeling of disulfide bridges Criteria for introduction into proteins by site-directed mutagenesis Int J Peptide Protein Res 36, 147–155 Liu, H., Banville, D.L., Basus, V.J & James, T.L (1995) Deriving accurate interproton distances for ROESY spectra with limited knowledge of scalar coupling constants via the carnival algorithm: an iterative complete-relaxation-matrix approach J Magn Reson 107, 51–59 Bystov, V.F (1976) Spin-spin coupling and the conformational states of peptide systems Progr NMR Spectrosc 10, 41–81 Morris, G.M., Goodsell, D.S., Halliday, R.S., Huey, R., Hart, W.E., Belew, R.K & Olson, A.J (1998) Automatic docking using a Lamarckian genetic algorithm and an empirical binding free energy function J Compu Chem 19, 1639–1662 Wang, J.H., Smoluar, A., Tan, K., Liu, J.H., Kim, M., Sun, Z.Y.J., Wagner, G & Reinherz, E.L (1999) Structure of a heterophilic adhesion complex between human CD2 and CD58 (LFA-3) counter-receptors Cell 97, 791–803 Rose, G.D., Gierasch, L.M & Smith, J.A (1985) Turns in peptides and proteins Adv In Protein Chem 37, 1–97 Hetenyi, C & Van Der Spoel, D (2002) Efficient docking of peptides to proteins without prior knowledge of the binding site Protein Sci 11, 1729–1737 Ikemizu, S., Sparks, L.M., van der Merwe, A., Stuart, D.I., Jones, E.Y & Davis, S.J (1999) Crystal structure of the CD2binding domain of CD58 (lymphocyte function-associated ˚ antigen 3) at 1.8 A resolution Proc Natl Acad Sci USA 96, 4289– 4294 Bodian, D.L., Jones, E.Y., Harlos, K., Stuart, D.I & Davis, S.J (1994) Crystal structure of the extracellular region of the human ˚ cell adhesion molecule CD2 at 2.5 A resolution Structure 2, 755– 766 Withka, J.M., Wyss, D.F., Wagner, G., Arulanandam, A.R.N., Reinherz, E.L & Recny, M.A (1993) Structure of the glycosylated adhesion domain of human T lymphocye glycoprotein CD2 Structure 1, 69–81 Jones, E.Y., Davis, S.J., Williams, A.F., Harlos, K & Stuart, D.I ˚ (1992) Crystal structure at 2.8 A resolution of soluble form of the cell adhesion molecule CD2 Nature 360, 232–239 Sun, Z.Y.J., Dotsch, V., Kim, M., Li, J., Reinherz, E.L & ˜ Wagner, G (1999) Functional glycan-free adhesion domain of human cell surface receptor CD58: design, production and NMR studies EMBO J 18, 2941–2949 Huang, Z., Li, S & Korngold, R (1997) Immunoglobulin superfamily proteins: Structure, mechanisms, and drug discovery Biopolymers 43, 367–382 Olson, G.L (1990) Design and synthesis of a protein beta-turn mimetic J Am Chem 112, 323–333 Tibbetts, S.A., Jois, S.D.S., Siahaan, T.J., Benedict, S.H & Chan, M.A (2000) Linear and cyclic LFA-1 and ICAM-1 peptides inhibit T cell adhesion and function Peptides 21, 1161– 1167 Supplementary material The following material is available from http://blackwell publishing.com/products/journals/suppmat/ejb/ejb4198/ ejb4198.htm Appendix Chemical shift, coupling constants and temperature dependence of amide proton resonance data at 298K for peptides, cER, cVY, cEL 2886 L Jining et al (Eur J Biochem 271) Appendix The backbone dihedral angles (in deg.) at residues Glu2-Arg3-Gly4-Ser5 and at residues Ser4-Thr5Asn6-Gly7 for conformations of peptide cVY Appendix Peptides cEL and cYT: CD58 docking results starting from the potential binding sites out of 100 runs Fig S1 Amide region of 500 MHz ROESY spectra of cER cyclic peptide in 90% H2O/10% D2O Fig S2 Amide region of 500 MHz ROESY spectrum of cVY cyclic peptide in 90% H2O/10% D2O Ó FEBS 2004 Fig S3 Amide region of 500 MHz ROESY spectrum of cEL cyclic peptide in 90% H2O/10% D2O Fig S4 Amide region of 500 MHz NOESY spectrum of cVT cyclic peptide in 90% H2O/10% D2O Fig S5 Measurement of CD2 expression on the surface of Jurkat cells by flow cytometry Fig S6 Cell viabilities of peptide-treated Caco-2 and Jurkat cells  ... peptides from rat CD2 may bind to human CD58 Human -CD2 and rat -CD2 have high 3D structure similarity Comparison of 3D structures of hCD2 from CD2? ??CD58 complex [30] and the structure of rat CD2 [36] using... biological activity and to study the effect of reducing the chain length of peptides on Ó FEBS 2004 Design of peptides for T-cell adhesion inhibition (Eur J Biochem 271) 2879 biological activity, ... N- and C-termini A comparison of the b-turn structure of cER with the similar region in the crystal structure of rat and human CD2 was carried out In the rat CD2 crystal structure, the b-turn structure