Portal de Periódicos da CAPES

Functional inhibition of CCR3-dependent responses by peptides derived from phage libraries

2001; Wiley; Volume: 31; Issue: 12 Linguagem: Inglês

10.1002/1521-4141(200112)31

1521-4141

Mehdi Houimel, Pius Loetscher, Marco Baggiolini, Luca Mazzucchelli,

Chemical Synthesis and Analysis

AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Abstract So far chemokine antagonists have been identified by modification of the NH2-terminus of known chemokines or by screening large number of compounds in functional assays. Here we used phage display peptide libraries to identify hexapeptides that antagonize the interaction between eotaxin and its receptor CCR3. The peptide sequence CPWYFWPC was recovered by panning phage libraries on CCR3-transfected murine pre-B cells after elution with eotaxin. The synthetic, structurally constrained peptide effectively competed 125I-eotaxin binding to CCR3 (IC50 = 20 μM). Furthermore, it had weak agonistic effects on Ca2+ mobilization in CCR3 transfectants that underwent heterologous desensitization by subsequent exposure to eotaxin. The peptide inhibited chemotaxis of CCR3 transfectants induced by a broad panel of CCR3 ligands. Specificity was tested with the CXCR1, CXCR2, CXCR3 and CCR5 receptors. In experiments aimed at characterization of residues necessary for eotaxin binding, we affinity purified the linear eotaxin-binding peptide VTPRQR, and showed that the peptide displaced the binding of radiolabeled eotaxin to CCR3 (IC50 = 300 μM) ina dose-dependent manner, inhibited eotaxin induced increases in intracellular Ca2+, and migration of CCR3-transfected cells. Specificity was affirmed using other CCR3 ligands. This is the first de novo identification of chemokine antagonists by direct screening on target proteins. Abbreviations: IP10: Interferon-inducible protein 10 I-TAC: Interferon-inducible T cell RANTES: Regulated on activation normal T cell expressed and secreted MCP: Monocyte chemoattractant protein Mig: Monokine induced by γ-interferon MIP: Macrophage inflammatory protein CCR: CC chemokine receptor CXCR: CXC chemokine receptor TU: Transducing unit 1 Introduction Eosinophils are pro-inflammatory granulocytes that play a major role in allergic diseases, such as asthma 1, allergic rhinitis 2, atopic dermatitis 3, eosinophilic gastroenteritis 4 and parasitic infections 5. Human eosinophils respond to a variety of CC chemokines including eotaxin, eotaxin-2,eotaxin-3, regulated on activation normal T cell expressed and secreted (RANTES), monocyte chemoattractant protein (MCP)-2, MCP-3, and MCP-4 6, through binding to the CC chemokine receptor-3 (CCR3), a seven transmembrane domain G-coupled receptor. CCR3 is expressed not only on eosinophils but also on Th2 T cells 7, 8, basophils 9, and mast cells 10. Thus, because of its action on eosinophils and on a whole array of cell types that are crucial for induction of an allergic response, eotaxin and the CCR3 receptor have become an interesting target for drug intervention. It is well known that mice genetically deficient for eotaxin 11 have diminished eosinophilia during the development of allergic inflammatory responses. In addition, anti-CCR3 mAb 7B11 was shown to completely block the binding and signaling of the known ligands for CCR3, and consequently to inhibit eosinophil recruitment 12. Modification of CCR3 agonists, such as RANTES, by N-terminal methionine elongation (Met-RANTES) 13, or addition of aminooxypentane (AOP-RANTES) 14, can generate antagonists able to inhibit inflammation and HIV infection, respectively. Recently, an antagonist of CCR3 was derived from the CC chemokine macrophage inflammatory protein (MIP)-4 15. This antagonist, termed Met-Ckβ7, specifically binds to CCR3, and blocks eosinophil chemotaxis to eotaxin at concentrations as low as 1 nM. Finally, potent small nonpeptide molecules antagonists of CCR3 have been identified by functional screening of large numbers of organic compounds 16, 17. Recently, it was also demonstrated that the CXCR3 ligands, IFN-inducible T cell (I-TAC), IFN-inducible protein 10 (IP10), and monokine induced by γ-IFN (Mig) are effective CCR3 antagonists 18. The identification of chemokine antagonists by classical approaches such as modification of the NH2-terminal region of chemokines, screening of large number of chemical compounds, or characterization of a natural cross-competition between receptor ligands is often laborious, time intensive, and not necessarily effective. The identification de novo of small molecules binding to the chemokine receptors or their ligands by direct screening of the target proteins is therefore a promising and appropriate alternative strategy for the development of new drugs. In this context, the utilization of combinatorial peptide libraries expressed on filamentous bacteriophage 19 is a powerful tool for studies on ligand-receptor interaction 20, and for the discovery of pharmacologically active lead compounds, such as peptide agonists or antagonists 21. In this study, we used two phage peptide libraries displaying either a linear or a structural constrained 6-mer peptide to identify functional antagonists of the eotaxin-CCR3 interaction. First, we identified peptides that bind to and block CCR3 by panning phage libraries on CCR3-transfected cells. Second, in an attempt to identify residues critical in eotaxin binding, we characterized soluble eotaxin ligands able to specifically compete with eotaxin for binding to its receptor. Our data indicate that short peptides unrelated to known chemokine or chemokine receptor sequences retain the ability to engage specific interaction with these molecules and may be used to develop new therapeutic agents. 2 Results 2.1 Affinity isolation of CCR3 binding phage Phage clones binding to CCR3 were selected by incubation of the structural constrained hexapeptide library with murine pre-B cells 300-19 expressing the human CCR3 receptor (approximately 20,000 receptors/cell) at their surface. We opted for phage displaying a cyclic peptide because the binding domain of eotaxin is presumed to be structurally constrained and, in general, because libraries displaying disulfide-constrained motif may be more appropriate to identify high-affinity ligands 22. To avoid receptor internalization upon phage binding, the incubations were performed at 4 °C. Bound phage were first specifically eluted by competition with excess of eotaxin (3 μM). As a negative control nonspecific cell binding phage were then recovered with a triethylamine solution (pH 11). After three rounds of such selection, an enrichment of approximately 102 was achieved for specifically eluted phage, suggesting that biopanning was effective. Conversely, the number of phage particles recovered by nonspecific elution from the CCR3 transfectants progressively decreased over the three rounds of selection (data not shown). After the third round of affinity purification, ELISA screening was used to further select phage clones binding with high affinity to the target protein. The random inserts of the 10 clones giving the strongest ELISA signals were then sequenced. The deduced amino acid sequences identified 5 phage peptides, suggesting PWY(F/L)WP as possible consensus motif (Table 1). Analysis of the selected motif using Gen-Bank/EMBL did not reveal any significant homology with known proteins, nor in particular with known chemokines Table 1. Peptide motif selected from the structurally constrained phage peptide library binding to CCR3 Target/Library Sequence Number of recovered clones CCR3 transfected CPWYLAPC 2 cells/Constrained CPWYFLWC 2 (pIII-C-X6-C) CPWYFWPC 4 CWANGWPC 1 CHGPLVWC 1 2.2 CCR3 binding and functional assays with affinity-purified phage The CCR3-binding phage clones were tested for their ability to inhibit 125I-eotaxin binding to CCR3. As a positive control we used excessive amounts of unlabeled eotaxin, or of the anti-CCR3 mAb 7B11. The selected phage clones were all equally able to compete for 125I-eotaxin binding to CCR3, whereas a control phage had no effect (Fig. 1). To determine the ability of the CCR3 binding phage to antagonize the Ca2+ mobilization induced by eotaxin, the selected phage clones (1.0 × 1010 TU) were added to Fura-2-loaded CCR3 transfected cells 60 s before stimulation with 5 nM eotaxin. As shown in Fig. 2, none of the tested clones had an agonistic effect, as outlined by the failure to induce [Ca2+]i changes in CCR3 transfectants. Conversely, compared to the [Ca2+]i rise observed after exposure of CCR3 transfectants to eotaxin alone, a decrement in the magnitude of [Ca2+]i mobilization was observed by prior exposure of the transfected cells to CCR3-binding phage clones. The strongest antagonistic effect was obtained with the phage clone displaying CPWYFWPC, i.e., approximately 80 % reduction of [Ca2+]i mobilization. A control phage had no effect. 2.3 CCR3 binding and functional assays with synthetic peptides To confirm that the interactions between the selected phage clones and CCR3 were specific for the displayed random motif, binding assays were performed with synthetic peptides. The CPWYFWPC peptide was synthesized because of the strong inhibition of [Ca2+]i mobilization obtained in assays with the corresponding phage clone. An irrelevant cyclic hexapeptide was used as a control. The CPWYFWPC peptide was able to markedly antagonize the 125I-eotaxin binding to CCR3-transfected cells. Half maximal inhibition (IC50) was seen at 20 μM (Fig. 3). In contrast, even high peptide concentration (500 μM) did not affect binding competition assays with other radiolabeled CCR3 ligands known for their antagonistic effect, such as 125I-IP10 and 125I-I-TAC (Fig. 4). Experiments were then performed to investigate the ability of the CPWYFWPC peptide to inhibit intracellular Ca2+ flux in response to eotaxin. In dose-response assays, the addition of peptide alone induced a weak increase in [Ca2+]i. However, as shown in Fig. 5, the prior exposure to the peptide engendered a decrement of approximately 50 % in the magnitude of eotaxin-induced [Ca2+]i rise in CCR3 transfectants. The migration of CCR3 transfectants in response to eotaxin was also inhibited by preincubating the cells with the CPWYFWPC peptide. The inhibitory effect was already detectable at low peptide concentrations (1 μM). At higher concentrations (100 μM), however, 20 % of the CCR3 transfectants were still migrating (Fig. 6 A). Finally, the specificity of the peptide CPWYFWPC was tested in migration assays using murine pre-B cells transfected with CCR5, CXCR1 CXCR2 or CXCR3 and appropriate ligands (Fig. 6 B). A cross-reactivity was not observed. As several chemokines bind to CCR3, we also tested the ability of the CPWYFWPC peptide to inhibit CCR3 transfectants migration induced by ligands other then eotaxin. As shown in Fig. 7, migration induced by eotaxin-2, RANTES, MCP-3, and MCP-4 was markedly inhibited by prior incubation of the CCR3 transfectants with 100 μM of the peptide. Interestingly, eotaxin-2-induced migration was more efficiently inhibited by this peptide concentration than migration induced by eotaxin. Figure 1Open in figure viewerPowerPoint Ability of phage clones (1.0 × 1010 TU) selected by screening a structurally constrained hexapeptide library on CCR3-transfected cells and recovered by specific elution to compete for 125I-eotaxin binding to CCR3. Cold eotaxin and the anti-CCR3 mAb 7B11 served as positive controls. A control phage had no effect. Results are expressed as percentage (mean ± SD) of 125I-eotaxin binding. Experiments were performed in duplicate and repeated three times. Figure 2Open in figure viewerPowerPoint Antagonist effect of CCR3 binding phage clones on eotaxin-induced [Ca2+]i mobilization. Fura-2 loaded CCR3 transfectants were incubated for 60 s with the indicated phage-displayed constrained hexapeptides (1.0 × 1010 TU) selected against CCR3 before stimulation with 5 nM eotaxin. A control phage was used as a negative control. Values are representative of two comparable experiments. Figure 3Open in figure viewerPowerPoint CCR3 binding assays with synthetic peptide: The CCR3 binding structurally constrained peptide CPWYFWPC (black circle) antagonizes 125I-eotaxin binding to CCR3 transfected cells in a dose-dependent manner (IC50 = 20 μm). The control peptide had no effect (black square). The data points represent the mean ± SD from three experiments. Figure 4Open in figure viewerPowerPoint CCR3 and CXCR3 binding assays of 125I-IP10 (A) or 125I-I-TAC (B) in the presence of 500 nM of IP10, I-TAC, eotaxin, or 500 μM of the synthetic peptide CPWYFWPC. The CPWYFWPC peptide does not affect 125I-IP10 and 125I-I-TAC binding to both receptors. The values were normalized by setting the specific binding of labeled IP10 or I-TAC to 100 %. As positive control 125I-eotaxin was used (data not shown). The data represent the mean values ± SD from three experiments. Figure 5Open in figure viewerPowerPoint [Ca2+]i mobilization assays with the synthetic CCR3 binding cyclic peptide CPWYFWPC. CCR3 transfectants were preincubated (60 s) with the cyclic peptide (125 μM) and subsequently exposed to eotaxin (5 nM). The time points at which the synthetic peptide, and eotaxin were added are indicated by the downward arrows. The peptide alone elicits a calcium signal in CCR3 transfectants that undergoes heterologous desensitization after exposure to eotaxin. Experiments with other peptide concentrations (not shown) were consistent with a dose-dependent response. Figure 6Open in figure viewerPowerPoint Chemotaxis inhibition by CPWYFWPC peptide. (A) eotaxin (10 nM) induced migration of CCR3 transfectants was analyzed in the presence of serial dilutions of the CPWYFWPC peptide or of control peptide (100 μM). Data represent mean values ± SD of three experiments performed with same cells. (B) Transwell migration assays were performed using CCR3-, CCR5-, CXCR1-, CXCR2-, and CXCR3-transfected murine preB cells. Transfectants were incubated with CPWYFWPC (100 μM) for 15 min prior stimulation with eotaxin (10 nM), RANTES (100 nM), IL8 (10 nM), IP10 (10 nM) or I-TAC (10 nM). A cross-reactivity of the CPWYFWPC peptide with CCR5, CXCR1, CXCR2 and CXCR3 was not observed. Data represent mean values ± SD of three experiments. Figure 7Open in figure viewerPowerPoint Effect of CPWYFWPC peptide in chemotaxis induced by eotaxin-2, RANTES, MCP-3, and MCP-4. CCR3 transfectants were incubated for 15 min with the cyclic peptide CPWFYFWPC (100 μM) prior to stimulation with eotaxin (10 nM), eotaxin-2 (10 nM), RANTES (100 nM), MCP-3 (100 nM), and MCP-4 (100 nM). Migration is quantitated as percent of controls that were performed in the presence of chemokines without competing peptide. Data represent the mean values ± SD from two experiments. 2.4 Affinity isolation and functional characterization of eotaxin binding phage Crucial binding domains of chemokine receptors are supposed to be located on the presumably linear NH2-terminal region. Thus, phage clones binding to human eotaxin were initially identified by incubating the linear hexapeptide phage library with eotaxin coated on immunotubes. Bound phage were eluted under basic conditions (pH 12). The enrichment of recovered phage after the third round of biopanning was 102. ELISA screening was then used to further select phage binding with high affinity to eotaxin. The peptide sequences of the ten phage clones giving the strongest ELISA signals are shown in Table 2. One identical peptide motif, MTRSIA, was recovered seven times. Although a consensus sequence was not apparent, the sequences showed some weak homologies. The most conserved residue, displayed in the second position after the NH2-terminus, was Thr. The amino acids Leu, Arg and Asp were all found twice as first, fourth and fifth residues, respectively. Conserved substitutions (L/V/M or T/S) suggested an even greater homology for the first two residues displayed at the NH2-terminus. Table 2. Peptide motif selected from the linear phage peptide library binding to eotaxin Target/Library Sequence Number of recovered clones Human Eotaxin/ LTDYQM 1 Unconstrained VTPRQR 1 (pIII-X6) MTRSIA 7 LSSRKI 1 Despite the lack of a clear consensus sequence, all selected eotaxin binding phage clones, but not an irrelevant control phage, were effective in inhibiting the binding of radiolabeled eotaxin to CCR3 transfectants (Fig. 8). Further, the [Ca2+]i rise induced by eotaxin in CCR3-expressing cells was abolished by incubation of eotaxin with eotaxin binding phage clones (1.0 × 1010 TU) for 1 h prior cell stimulation whereas no inhibition of [Ca2+]i mobilization was observed with a control phage (data not shown). Figure 8Open in figure viewerPowerPoint Phage clones (1.0 × 1010 TU) selected by screening a linear hexapeptide library on eotaxin inhibit the binding of 125I-eotaxin to CCR3 transfected cells. A control phage had no effect. 2.5 Functional effects of eotaxin-binding peptides Based on the weak sequence homologies among the motif displayed on eotaxin binding phage, on the functional assays outlined above, and on the predicted peptide solubility, we decided to test the linear peptide VTPRQR for binding and functional assays. As a control, an irrelevant linear hexapeptide was used. The peptide VTPRQR was able to inhibit 125I-eotaxin binding to CCR3 in a dose-dependent manner (IC50 = 300 μM), whereas the control peptide had no effect (Fig. 9). Experiments were then performed to further define the effect of synthetic peptide in elicitation of [Ca2+]i changes. In contrast to the findings obtained in [Ca2+]i mobilization assays with phage clones, eotaxin, in the presence of VTPRQR peptide, was still able to engage a reaction with CCR3 transfectants and induce a weak but detectable rise in [Ca2+]i. Nevertheless, the synthetic peptide, partially inhibited the transient rise in [Ca2+]i induced by eotaxin in a dose-dependent manner (Fig. 10 A). Here the IC50 was estimated at 250 μM, and eotaxin efficacy was unaffected by 500 μM of control peptide. Stimulation of the transfectants with the VTPRQR peptide alone (500 μM) and a control peptide did not induce [Ca2+]i mobilization (data not shown). To further define the specificity of the peptide VTPRQR for eotaxin, [Ca2+]i mobilization experiments were performed with other CCR3 ligands, namely eotaxin-2, RANTES and MCP-3. As shown in Fig. 10 B, the VTPRQR peptide (300 μM) did not affect the [Ca2+]i rise induced by these CCR3 agonists. Finally, the ability of the peptide VTPRQR to inhibit migration was tested in a transwell system using CCR3 transfectants. Inhibition of migration was also observed to be dose dependent but required higher concentrations of peptide (500 μM) to achieve 50 % inhibition (data not shown). Figure 9Open in figure viewerPowerPoint Binding of 125I-eotaxin of CCR3 transfectants is inhibited by increasing amount of the eotaxin binding peptide VTPRQR (open circle). The IC50 was estimated to be about 300 μM. The control peptide had no effect (black square). Figure 10Open in figure viewerPowerPoint [Ca2+]i mobilization assays with the synthetic eotaxin binding VTPRQR peptide. (A) Dose-response curve for eotaxin-induced calcium mobilization expressed as percent inhibition of the maximal magnitude obtained by exposure of the Fura-2-loaded cells to eotaxin alone. (B) The peptide (300 μM) partially prevents signaling through CCR3 induced by eotaxin whereas it does not affect [Ca2+]i rise induced by eotaxin-2 (5 nM), RANTES (50 nM) and MCP-3 (50 nM). 3 Discussion At the present time, the most commonly used antagonists of CCR3 are two newly identified small non-peptide molecules 16, 17, amino-terminally modified natural ligands, such as Met-RANTES 13 and AOP-RANTES 14, or a modified form of the β-chemokine MIP4, called Met-chemokine-β-7 15. Recently, it has also been demonstrated that the ligands of the CXCR3, I-TAC, Mig and IP10 are natural antagonists for CCR3 18. In this study we identified novel hexamer peptides, unrelated to known chemokine and chemokine receptor sequences, binding either to eotaxin or to its receptor CCR3, and able to inhibit eotaxin activity on CCR3-transfected murine pre-B cells. Although, more potent antagonists have been identified recently using other techniques 16, 17, our studies represent the first de novo identification of chemokine antagonists, and they indicate that similar procedures may be extended to studies on other chemokine receptors. In fact, the CCR3 binding motif was identified by biopanning experiments with phage display peptide libraries on whole cells expressing the CCR3 receptor in its natural conformation. This investigative approach is of particular interest sincechemokine receptors, as all members of the superfamily of the G-protein-coupled receptors, are 7-transmembrane-spanning proteins that may not be purified in vitro without losing their quaternary structure and binding properties. Panning on entire transfected cells may therefore represent a valid alternative to identify new ligands for receptors with a complex structure. After three rounds of affinity purification we were able to isolate several phage clones displaying the structurally constrained motif CPWY(F/L) WPC. All phage clones were able to abrogate binding of radiolabeled eotaxin to CCR3 transfectants and to inhibit the eotaxin induced intracellular calcium release in the CCR3-transfected murine pre-B cells. Considering that these effects were obtained with 1.0 × 1010 phage particles corresponding to picomolar concentrations of displayed peptide, one would expect to obtain a similar effect with the synthetic peptide at least in the nanomolar range. The synthetic circular peptide was able to displace up to 80 % of radiolabeled eotaxin from CCR3 in a concentration-dependent manner, however, much less efficiently (IC50 = 20 μM) than the respective phage clones. Further, in contrast to the results obtained with the phage clones, the CPWYFWPC peptide induced weak but detectable release of intracellular calcium in CCR3 transfectants. The rather weak efficiency of the free peptide in displacement of radiolabeled eotaxin from the CCR3 receptor may reflect the importance of multivalent binding interaction between the phage, which display five copies of the hexapeptide on the pIII protein, and CCR3. It is also possible that the conformation of the peptide fused on the pIII protein significantly differs from that of the soluble synthetic peptide. Alternatively, as recently shown with small non-peptide compound blocking eosinophil functions 16, the peptide identified in the present study may only partially block the eotaxin binding site on CCR3 and also engage an interaction with other domains of this receptor. The lack of an effect of the CPWYFWPC peptide to compete the binding of CCR3 antagonists, such as IP10 and I-TAC, also suggests the existenceof binding domains on this receptor that differ at least partially with respect to agonistic and antagonistic ligands. The ability of the CPWYFWPC peptide to antagonize chemotaxis was investigated in vitro against several chemokine receptors. The cyclic peptide was able to inhibit eotaxin-induced migration of the CCR3 transfectants with slightly higher potencies than its inhibition of eotaxin binding. This functional response may be due to partial blocking of the CCR3 receptor or desensitizationof the transfectants through the weak agonistic effect of the peptide, or to both factors. Importantly, however, the peptide alone was not able to induce chemotaxis even at high concentrations. Remarkably, the CPWYFWPC peptide was effective in inhibiting migration induced by other CCR3 ligands, such as RANTES, MCP-3, MCP-4, and in particular eotaxin-2. These findings are in agreement with thenotion that all agonistic ligands share a common binding pocket on the CCR3 receptor 23, 24. Finally, the specificity of the peptide was tested in migration assays with CCR5, CXCR1, CXCR2 and CXCR3 transfected murine pre-B cells. Migration of these transfectants was unaffected even with high peptide concentrations. In general, the identification of residues engaged in chemokine binding may be crucial for the characterization of receptor binding sites and consequently for the development of new drugs. In the present study we selected soluble peptides unrelated to known chemokine receptor sequences, but able to specifically bind eotaxin and abrogate its activity through the CCR3 receptor. An intervention at the level of chemokine by antagonists may not be appropriate due to the high redundancy of this system 25. However, several aspects indicate that this approach is worth trying. High redundancy exists also at the receptor level. For instance MIP-1α acting through the CCR1 receptor, is also very potent in recruiting eosinophils to the site of an allergic reaction 26. Conversely, some chemokines, such as BCA-1 and SDF-1α specifically bind to only one receptor. Therefore, the blockade of the chemokine may be effective, if we consider thatreceptor inactivation may be due to transient mechanisms, such as receptor internalization and cell desensitization. Lastly and more importantly, it is also conceivable that a stable peptide-chemokine interaction may prevent chemokine binding to extracellular matrix proteins, enabling therefore the generation of a stable chemotactic gradient in vivo 27. In our study, we demonstrated that the synthetic linear hexapeptide VTPRQR was able to inhibit, in a dose-dependent manner, the 125I-eotaxin binding to CCR3, the eotaxin-induced cellular calcium release, and the eotaxin-induced migration of CCR3 transfectants. Further, the peptide had no effect on other CCR3 natural ligands, such as eotaxin-2, RANTES and MCP-4, suggesting therefore high specificity, in spite of a homology among these molecules ranging from 34 % to 64 %. Interestingly, peptides consisting of at least 16 residues, derived from the N-terminal region of the CCR3 receptor 28, bind to eotaxin with an affinity similar to those obtained with the hexamer we identified. Taken together, these data indicate that several binding sites may exist on eotaxin and suggest that the characterization of these binding properties may extend the availability of reagents aimed at a the disruption of the chemokine-receptor interaction. In this report we describe inhibition of chemokine functions with low molecular weight compounds (hexapeptide) identified de novo, and not derived from modifications of previously known chemokine ligands, nor by functional screening of large numbers of chemical compounds. Several issues, however, must be addressed before drawing conclusions on the potential therapeutic applications of such peptides. For instance, a functional effect was obtained only at relative high concentrations, most likely because short constrained peptides might fail to mimic a complex ligand structure. Additional open questions rely to the specificity, stability and clearance from blood of these molecules in vivo. Nevertheless, newly identified peptides could enable the characterization of critical binding residues and could be envisaged as scaffolds for the design and chemical synthesis of novel molecules. Thus, analysis of binding peptides identified with phage libraries may help to identify candidate molecules among the large numbers of compounds generated by combinatorial chemistry, and in general, allow more efficient functional screening procedures. 4 Materials and methods 4.1 Chemokine synthesis and CCR3 blocking antibody All chemokines were chemically synthesized using tBoc solid-phase chemistry 29, purified by HPLC, and analyzed by electronspray mass spectrometry. The anti-human CCR3 mAb (7B11) was a gift of Dr. C. Dahinden, University of Bern, Bern, Switzerland. 4.2 Phage display libraries and peptide synthesis Phage peptide libraries displaying either a random linear or a structurally constrained hexapeptide have been previously described 30. Peptides were chemically synthesized using solid phase Fmoc chemistry with free N-termini and subjected to analytical HPLC and MALDI-TOF mass spectrometry. Constrained random peptide flanked by two cysteine residues have a cyclic structure. Thus, the corresponding synthetic peptides were oxidized by stirring a 1 mM solution of peptide in PBS for 36 h at room temperature. The oxidation was monitored at 412 nm by measuring the free-SH content with Ellman's reagent (5-5′ dithiobenzoic acid, Sigma Chemical Co., St. Louis, MO). 4.3 Chemokine receptor transfectants Chemokine receptor-expressing cells were obtained by transfecting murine pre-B cells 300-19 with cDNA for CCR3, CCR5, CXCR1, CXCR2 and CXCR3 as previously described 18, 31. Transfectants were grown in a humidified 5 % CO2 atmosphere at 37 °C and maintained in RPMI 1640 (Gibco BRL, Life Technologies, Basel, Switzerland) supplemented with 10 % FCS, nonessential aminoacids, sodium pyruvate (1 mM) penicillin (100 U/ml), streptomycin (100 μg/ml), L-glutamine (2 mM) and 1.5 μg/ml puromycin (Sigma Chemical Co.) or for CXCR3transfectants, G418 sulfate antibiotic (Gibco BRL) at 800 μg/ml. 4.4 Selection of CCR3-binding phage Affinity selection of CCR3-binding motif was performed by biopanning CCR3-transfected cell lines with a phage library displaying a constrained hexapeptide. For the first round of panning, 2 × 107 cells were harvested from culture medium, and after washing, resuspended in 500 μl of PBS s