PEDS Advance Access originally published online on January 19, 2006
Protein Engineering Design and Selection 2006 19(3):121-128; doi:10.1093/protein/gzj011
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Characterization of an Fc
RI-binding peptide selected by phage display
1Department of Molecular Biosciences, University of Oslo, 2Dynal Biotech ASA, Oslo and 3Norwegian Institute of Public Health, Oslo, Norway 4Present address: Akershus University Hospital, Lørenskog, Norway
5 To whom correspondence should be addressed. E-mail address: inger.sandlie{at}imbv.uio.no
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
The high-affinity IgG receptor, Fc
receptor I (FcRI), is expressed exclusively on myeloid cells, and there is a great interest in the targeting of vaccine antigens to FcRI using anti-human FcRI antibodies or fragments derived from such molecules. In order to reduce the size and complexity of the targeting reagent, we have searched for FcRI binding peptides in peptide libraries displayed on phage. The human monocytic cell line U937 was used as target. Phages that displayed the consensus peptide CLRSGXGC were selected and revealed increased binding to IFN- stimulated versus non-stimulated U937 cells as well as to FcRI transfected versus non-transfected IIA1.6 cells. Furthermore, they bound the extracellular domains of soluble FcRI, but neither FcRIIA, FcRIIB nor FcRIIIB. Binding was inhibited by a synthetic version of the peptide, whereas neither human IgG nor the FcRI-specific monoclonal antibodies (mAb) mAb22 and 32.2 interfered. Flow-cytometry analysis and internalization studies showed that a synthetic biotin-conjugated peptide ADGACLRSGRGCGAAK-bio was able to target U937 cells and FcRI transfected IIA1.6 cells, and further to promote internalization and vesicular degradation of streptavidin coupled to 1 µm magnetic beads. These peptides may have potential as FcRI targeting reagents.
Keywords: cell internalization/Fc receptors/peptide selection/phage display/U937 cells
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Fc
receptors (FcRs) are expressed on a majority of haematopoietic cells and play a critical role in both cellular and humoral immunity, as they serve to link antibody-mediated immune responses with cellular effector functions. Three distinct classes of FcRs are recognized by IgG, namely FcRI (CD64), FcRII (CD32) and FcRIII (CD16). In humans the latter two classes are further divided into FcRIIA, FcRIIB, and FcRIIC, and FcRIIIA and FcRIIIB, respectively. FcRI exhibits high affinity for monomeric IgG (Ka = 108–109 M–1), whereas FcRII and FcRIII exhibit relatively weak affinities (Ka < 107 M–1) and therefore bind IgG only as immune complexes (ICs) in an aggregated form (Daeron, 1997
).
Engagement of FcRs can lead to either activating or inhibitory signalling depending on the specific FcR being engaged, with FcRIIB being the sole inhibitory receptor. Activation results in release of inflammatory mediators, phagocytosis of microorganisms, antibody-dependent cellular cytotoxicity, endocytosis of ICs and maturation of dendritic cells (DC) (Daeron, 1997; Gessner et al., 1998; Ravetch and Bolland, 2001). Furthermore, antigen presentation on MHC I and II has been demonstrated (Regnault et al., 1999; Machy et al., 2000), which may lead to priming of both CD4 and CD8 T-cell responses. While FcRII and FcRIII are expressed on a variety of cell types, the expression of FcRI is restricted to antigen presenting cells (Pan et al., 1990; Fanger et al., 1997). Selective targeting of activating FcRs has been suggested for DC maturation and T-cell activation while circumventing the inhibitory activity of FcRIIB (Kalergis and Ravetch, 2002).
Owing to its high affinity for IgG, FcRI is saturated in vivo, and several mAbs have been developed which bind FcRI outside the IgG-binding site (Guyre et al., 1989), and trigger FcRI functions in vivo (Heijnen et al., 1996; Keler et al., 2000). Bispecific and humanized variants of such an antibody (mAb22) are currently developed as anticancer therapeutics (James et al., 2001; Repp et al., 2003). However, production of therapeutic full-length mAbs and bispecific Abs is costly and simpler targeting agents are being sought.
Binders to a large variety of cells and cell surface receptors have been selected from libraries of peptides with random amino acid sequences (Mori, 2004). The human monocytic cell line U937 constitutively expresses FcRI and FcRII but not FcRIII (van de Winkel and Anderson, 1991). FcRI is reported to be expressed at 
13 000–18 000 molecules per cell and may be upregulated 5- to 10-fold by IFN- stimulation (Guyre et al., 1983). In contrast, FcRII is not affected by IFN- stimulation. Thus, peptides with increased binding to IFN- stimulated versus non-stimulated U937 cells could represent a putative source of FcRI targeting molecules.
In this paper, we describe the use of IFN- stimulated U937 cells for the isolation of phage displayed FcRI binding peptides that can discriminate between FcRI and FcRIIA, FcIIB as well as FcRIIIB. Furthermore, the peptides bind to a site on FcRI different from the IgG binding site and the binding sites of two monoclonal anti-FcRI antibodies, namely mAb22 and 32.2. A synthetic biotin-conjugated variant of one selected peptide was shown to target streptavidin to U937 cells and promote receptor mediated uptake and intracellular degradation of streptavidin complexed on 1 µm magnetic beads.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Cells, phages, phage display libraries, antibodies and other reagents
The cell lines U937 (CRL-1503) and K562 (CCL-243) were purchased from American Type Culture Collection (ATCC). An FcR and FcR -chain negative murine cell line, IIA1.6, as well as IIA1.6 cells co-transfected with human FcRIA and -chain were kindly provided by J.G.J. van de Winkel, University Medical Centre, The Netherlands. All cell lines were cultured in RPMI 1640 (Bio Whittaker, Belgium) supplemented with 10% heat inactivated FCS (Integro b.v, The Netherlands), 2 mm L-glutamine, 25 U/ml penicillin and 25 µg/ml streptomycin (all from Bio Whittaker) under standard conditions. The transfected IIA1.6 cells were cultured in the presence of 5 µM methotrexate (Sigma, St Louis, MO) to ensure high levels of FcRIA expression. Peripheral blood mononuclear cells (PBMC) were purified from whole blood by lymphoprep density gradient centrifugation (Axis-Shields, Norway). Human-soluble FcRIIIB, produced in Escherichia coli inclusion bodies followed by reconstitution in vitro (Zhang et al., 2000), was kindly provided by P.D.Sun, NIAID, NIH Rockville, MD, USA.
Two fUSE5 (Scott and Smith, 1990) gpIII display libraries of six (Cys6) or nine (Cys9) random amino acids constrained by flanking cysteine residues have previously been described (Lauvrak et al., 2004). E.coli K91K cells as well as the fUSE5 vector were gifts from G.P.Smith, University of Missouri USA. A fUSE5 phage displaying the peptide CLLGGLGC (LLGG-phage) was created by insertion of double-stranded oligonucleotides encoding the peptide-sequence as a gene III fusion using the Sfi sites of the fUSE5 vector essentially as described (Smith and Scott, 1993; Lauvrak et al., 2004). Human IgG3 was produced at the National Institute of Public Health (Norway) and human normal IgG was purchased from Aventis (NJ, USA) and aggregated as previously described (Kawata et al., 1996). The monoclonal anti-human FcRI antibodies mAb22 and 32.2 were obtained from Abcam (Cambridge, UK). Goat anti-human IgG conjugated to horse radish peroxidase (HRP) was purchased from Sigma (Germany) and FITC-labelled goat anti-mouse IgG was obtained from Molecular Probes (OR, USA). FITC-labelled streptavidin was obtained from Dako (Denmark) and streptavidin-coupled magnetic beads (Dynabeads, myoneTM, diameter: 1 µm) was obtained from Dynal Biotech (Norway). 125I was obtained from Amersham (England). IFN- was purchased from Genzyme (Cambridge, MA). Concanamycin A was obtained from Sigma and E64d was obtained from Calbiochem (Germany).
Synthetic peptides
Cyclic synthetic peptides were produced at a 5–10 mg scale with purity of >75% (Eurogentec, Seraing, Belgium). Peptides were either produced in the form NH2-adgacxncga-COOH or as biotin-conjugated peptides in the form NH2-ADGACXnCGAAK-Bio, where ADGA and GA(A) are flanking amino acids as found in the fUSE5 phage, K-Bio represents a biotin-conjugated lysine residue and CXnC represents the selected peptides CLRSGLGC (C6-1), CLRSGRGC (C6-2) as well as the control peptides CDIFGRDC (C2) and CWTSGARWRLC (C3). Lyophilized peptides were stored at –20°C and dissolved before use in distilled H2O to a concentration of 10 mg/ml.
Affinity selection
U937 cells were treated with 200 U/ml IFN- for 48 h as described (Nambu et al., 1989; Sarmay et al., 1992). Upregulation of FcRI was verified by flow cytometry. For each selection, portions of 107 IFN- stimulated U937 cells were washed with 20 ml phosphate-buffered saline (PBS; pH 7.4) supplemented with 1% BSA (Sigma-Aldrich, Steinheim, Germany), 1 mM CaCl2 and 10 mm MgCl2 (PW buffer). Cells were recovered by centrifugation, resuspended and blocked in 900 µl PW buffer at 4°C for 30 min. Portions of each phage display library corresponding to 1010 transducing units (TUs) were diluted and blocked for 30 min in 100 µl PW buffer. Pre-blocked phages were allowed to react with pre-blocked IFN- stimulated U937 cells for 1 h at 4°C with agitation. Low temperature was chosen to avoid induction of phagocytosis. Unbound phages were removed by eight washes with 2 ml PW buffer. Bound phages were eluted with 200 µl 0.1 M HCl–glycine (pH 2.2) for 10 min on ice, and the eluates were neutralized with 17 µl 1.5 M Tris (pH 9.1), amplified by infection of E.coli K91K and precipitated with polyethylene glycol as described (Smith and Scott, 1993). A second round of affinity selection was performed with 1010 TUs of amplified phages from the first selection. Phage supernatants from individual clones collected after the second round of selection were PCR-amplified with the primers 5'-CTATTCTCACTCGGCCGACG-3' and 5'-TTCAACAGTTTCGGCCCCAG-3' and characterized by sequencing (GATC, Germany).
Phage binding to monocytic cells, IIA1.6 cells and PBMC
Portions of 3 x 106 IFN- stimulated and non-stimulated U937 cells, FcRI negative K562 cells, FcRI transfected and untransfected IIA1.6 cells as well as PBMC were washed as described above and blocked in 400 µl PW buffer at 4°C for 30 min. Portions of 109 TUs from individual phage clones isolated by two rounds of selection (C6-1, C6-2, C6-3 and C9-1) or a phage clone with the insert CLLGGLGC (LLGG phage) and the C1 (CGPGGYVGYTC) or C2 control phages were diluted in PW buffer to a total volume of 100 µl and incubated with U937, K562, IIA1 cells and/or PBMC for 1 h at 4°C with agitation. Cells were washed eight times in 2 ml PW buffer and bound phages were eluted and neutralized as described above. The number of bound phages was determined as TUs in acid eluates.
Inhibition of phage binding to U937 cells
Portions of 3 x 106 IFN- stimulated and PW buffer pre-blocked U937 cells were incubated at 4°C with either 5-fold dilutions of the C6-1 synthetic peptide starting at a concentration of 400 µM, or the C2 synthetic peptide at a concentration of 400 µM, human IgG3 at 1 and 12 mg/ml or each of the two anti-FcRI antibodies mAb22 and 32.2 at 1 mg/ml. Following 1.5 h incubation, 109 TUs of PW buffer pre-blocked phages expressing the C6-1 peptide were added and allowed to react with the cells for 1 h at 4°C with agitation. Cells were washed eight times in 2 ml PW buffer and bound phages were eluted as described above. The number of bound phages was determined as TUs of acid eluates.
Soluble FcRI, FcRIIA and FcRIIB
The extracellular domains of human FcRI, FcRIIA and FcRIIB were cloned and expressed as soluble fusions to GST in 293-E cells as described (Berntzen et al., 2005). The expressed proteins were purified from cell supernatants by GSTrap FF columns (Amersham Pharmacia Biotech, UK). Bound proteins were eluted with 50 mM Tris–HCl, 10 mM reduced glutathione (Sigma), pH 8.0. Elution buffer was exchanged with PBS (pH 7.4) using Centricon-50 concentrators (Millipore, Bedford, MA). Proteins were analysed on 10% SDS-polyacrylamide gels (NOVEX, San Diego, CA) and their concentration determined using the GST 96-well detection module with recombinant GST as standard (Amersham Pharmacia Biotech). Briefly, two anti-GST antibodies were used, one as coat and one HRP-conjugated as detecting agent.
Receptor activity was evaluated by human IgG3 binding as follows: 100 µl FcRI, FcRIIA and FcRIIB fusion proteins in PBS were incubated in microtitre wells (Nunc, maxisorp, Denmark) overnight (ON) at 4°C and blocked with PBS supplemented with 1% BSA (PBS/BSA) for 1 h. Heat aggregated human IgG (3 µg/ml) was added to the wells and incubated for 1 h at room temperature (RT). Wells were washed four times with PBS containing 0.05% Tween-20 (PBS/T). Goat anti-human IgG conjugated to HRP (1:1000-dilution) was added and incubation for 1 h at RT was continued. Following four new washes with PBS/T, the plates were developed with ABTS (Sigma) in citrate buffer. Absorbance at 405 nm was read after 60 min.
Phage binding to FcRI, FcRIIA, FcRIIB and FcRIIIB.
For the phage binding assay, microtitre wells were coated with FcRI, FcRIIA and FcRIIB GST-fusion proteins at 1 µg/ml as well as FcRIIIB at 10 µg/ml and blocked as above. Portions of 50 µl of selected C6-1 and control C2 phage supernatants (5 x 109 TUs) diluted 1:1 with PW buffer were added to the wells. The phages were allowed to react with the receptors for 1 h at RT. Wells were washed eight times in PW buffer and bound phages were eluted with 100 µl 0.1 M HCl–glycine (pH 2.2) for 10 min at RT. The eluates were neutralized with 9 µl 1.5 M Tris (pH 9.1). The number of eluted phages was determined as TUs.
Flow cytometry
Biotin-conjugated C6-2 and C3 peptides were incubated with FITC-conjugated streptavidin (1 mg/ml in PBS) at a 4:1 molar ratio for 30 min at RT to form peptide–strep–FITC complexes. Portions of 3 x 106 PW buffer blocked IFN- stimulated or non-stimulated U937 cells as well as untransfected and FcRI transfected IIA1.6 cells in volumes of 450 µl PW buffer were incubated with 50 µl of the peptide–strep–FITC complexes for 1 h at 4°C and then washed three times with PW buffer. Fluorescence intensity of the samples was measured on a FACS Calibur (Becton Dickinson, NJ, USA) using cellquest software. As a positive control U937 cell surface FcRI expression was detected by incubation of the same amount of cells with mAb22 (4 µg/ml), washed three times with PW buffer and incubated similarly with a saturating concentration of FITC-labelled goat anti-mouse IgG (1:100 dilution).
Receptor-mediated internalization
A portion of 7 x 108 streptavidin-coupled magnetic beads in 200 µl PBS was incubated with 125I (40 MBq) in an iodogen-coated tube (IODO-GEN, Bio-Rad). After 60 min the labelled beads were washed with PBS to remove free 125I using a magnet (Dynal). The 125I–streptavidin-coupled beads were incubated with 4 µl biotin-conjugated C6-2 peptide (10 mg/ml) at RT for 60 min to form bead–125I–streptavidin–peptide complexes. The beads were washed to remove unbound peptides by using the magnet, and re-suspended in PBS containing 0.1% BSA. Portions of 2.5 x 106 IFN- stimulated as well as non-stimulated U937 cells were washed in serum-free RPMI supplemented with 1% BSA and pre-incubated for 30 min at 37°C in 6-well plates (Nunc, Denmark) before the addition of bead–125I–streptavidin–peptide complexes, at approximately three to four beads per cell. Half of the wells containing IFN- induced cells were further supplemented with 0.1 µM concanamycin A or 50 µM E64d. Samples of supernatant were collected after incubation at 37°C for various time points. In each case, acid insoluble material was precipitated with 10% TCA for 15 min on ice followed by centrifugation, and acid-soluble radioactivity that represents cell degraded 125I-labelled streptavidin was counted in a gamma counter (CoBRA II, Perkin Elmer, USA). Cells incubated with bead–125I–streptavidin–C3 peptide were used as a control for non-specific uptake. To determine the amount of acid-soluble radioactivity released as a result of receptor mediated uptake, a background of acid-soluble count in wells without cells was subtracted.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Selection of U937-reactive peptides
To isolate FcRI targeting peptides, two libraries of peptides displayed on phage, Cys6 and Cys9, were separately screened for phages that bind IFN- stimulated U937 cells. Two successive rounds of selection were performed and portions of pooled amplified phages from each round of selection were analysed for binding to stimulated U937 cells. Clones from the second round of selection were chosen at random and sequenced and analysed for binding to IFN- stimulated and non-stimulated U937 cells as well as to the FcRI-negative monocytic cell line K562. Binding was compared with that of equal numbers of a randomly picked clone from the Cys6 library. The binding specificity of the selected phages was further studied by investigating their binding to untransfected and FcRI-transfected IIA1.6 cells as well as to PBMC.
Flow cytometry analysis of the U937 cells confirmed that the surface expression of FcRI was upregulated more than 5-fold by IFN- stimulation (data not shown), making FcRI a prominent structure on the U937 cell surface. For both libraries the percentage of rescued phage increased with a factor of 104 after the second round of selection (Figure 1), indicating that phages with affinity for U937 cells were accumulating. Sixteen of eighteen selected Cys6 clones shared the consensus sequence CLRSGXXC, where X is a variable amino acid. The majority of phages with the CLRSGXXC motif also shared a Gly residue as the penultimate amino acid. Among ten selected Cys9 clones six shared the sequence CSWIPGVGLVC, but no consensus sequence was identified. The sequences of peptides presented by random phage clones after the second round of selection are shown in Table I.
|
|
In one representative experiment using an input of 109 TUs of the C6-1 and C9-1 phage clones the mean phage recoveries from IFN-
stimulated U937 cells were 5 x 106 and 3.2 x 106 TUs, respectively. Recovery from non-stimulated cells gave a mean value of 2.4 x 106 for the C6-1 phage and 2.6 x 106 for the C9-1 phage. In contrast, only 4 x 104 TUs of the C1 control phage were recovered from both IFN- stimulated and non-stimulated cells. Less significant difference in binding of the three phage clones to the FcRI negative cell line K562 was seen (Figure 2A). In a similar experiment the C6-2 and C6-3 phages, that both differ from C6-1 in one amino acid, showed equal binding levels as C6-1 phages to both stimulated and non-stimulated U937 cells, as well as to the K562 cells (data not shown). As the difference in binding to the IFN- stimulated cells versus the non-stimulated cells was more pronounced for the selected C6-phages compared with the C9-1 phage, the C6-phages were chosen for further investigation. Furthermore, the C6-2 phage showed increased binding to FcRI transfected versus untransfected IIA1.6 cells (Figure 2B), which also suggested that FcRI is the molecular target of the selected C6 phages. For additional validation of the results, phage binding to freshly isolated PBMC was investigated. Compared with the C2 control phage, the C6-2 phage showed five to ten times increased binding to PBMC (which contained 5% FcRI positive cells) (data not shown).
|
To test the influence of substituting the Arg and Ser residues of the selected C6-1 motif to Leu and Gly, a phage displaying the peptide CLLGGLGC (LLGG-phage), that contains the motif LLGG found in the IgG lower hinge, was produced and examined for binding to IFN-
stimulated U937 cells. In contrast to the U937 selected C6-1 phage, the LLGG-phage bound only at the level of the C2 control phage (Figure 2C), indicating that the Arg and Ser residues at positions three and four of the insert are essential for binding.
Phage binding to recombinant FcRI, FcRIIA, FcRIIBand FcRIIIB
To investigate if the selected C6-1 phage bound to any of the three FcRs; FcRI, FcRIIA or FcRIIB, expressed by U937 cells, recovery of phages that bound to recombinant receptors coated in microtiter wells was evaluated. In each case aggregated human IgG3 was shown to bind the receptors at various concentrations, indicating that the receptors were correctly folded so as to be functional (Figure 3A). At receptor concentration of 2 µg/ml, the recovery of C6-1 phage from FcRI was 100-fold increased compared with the C2 control phage. In contrast, C6-1 phage binding to a coat of 2 µg/ml of FcRIIA and FcRIIB was only at the level of the C2 control phage (Figure 3B). No binding of C6-1 to FcRIIA and FcRIIB could be seen at receptor concentrations up to 6 µg/ml (data not shown). Furthermore, cross-reactivity with the third class of Fc receptors was investigated. The bacterially produced FcRIIIB showed binding to aggregated human IgG3 (not shown), but as for the FcRIIs no binding to the C6-1 phage was detected (Figure 3C). In conclusion, we found that the selected C6-1 phage, in contrast to human IgG, interacted with FcRI but not FcRIIA, FcRIIB nor FcRIIIB.
|
Inhibition of phage binding to IFN-
stimulated U937 cells by human IgG, anti-FcRI mAbs and synthetic peptides
To further study the specificity of the interaction, inhibition of C6-1 phage binding to IFN- stimulated cells by human IgG3, two anti-FcRI Abs (mAb22 and 32.2) and synthetic peptides was tested. The cells were pre-incubated with an excess of the inhibitors before the phages were added. The number of recovered phages was determined as E.coli K91K-TUs. Neither human IgG nor the two anti-FcRI mAbs at concentrations up to 1 mg/ml inhibited C6-1 phage binding to the cells (Figure 4A), indicating that the C6-1 phages bind to an epitope on FcRI different from the binding site for human IgG and the two characterized anti-FcR mAbs. Similarily, IgG at serum concentrations (12 mg/ml) did not inhibit phage binding to cells (data not shown). In contrast, a synthetic version of the C6-1 peptide that contains the binding sequence of the phage flanked by seven residues from the phage gpIII, showed significant dose-dependent inhibition of C6-1 phage binding to IFN- stimulated U937 cells. At a concentration of 4 µM synthetic C6-1 peptide >50% reduction in phage recovery was seen. In contrast, no inhibition was seen using 400 µM of the C2 control peptide (Figure 4B). A synthetic biotin-conjugated version of the C6-2 peptide revealed the same inhibitory ability as the C6-1 peptide (data not shown).
|
Targeting of streptavidin to Fc
RI-expressing cells
To investigate whether a synthetic version of the selected peptides was able to target a protein to FcRI expressing cells, biotin-conjugated C6-2 peptide complexed to FITC-labelled streptavidin was incubated with IFN- stimulated and non-stimulated U937 cells as well as untransfected and FcRI transfected IIA1.6 cells, followed by flow cytometry analysis. The presence of FcRI on the U937 cells was verified by mAb22 and 32.2 binding followed by FITC-labelled goat anti-mouse IgG. When complexed with the synthetic C6-2 peptide, FITC-labelled streptavidin showed significant binding to non-stimulated U937 cells. Furthermore, as for mAb22 and 32.2, the C6-2 streptavidin complex revealed increased binding to the IFN- induced U937 cells (Figure 5A). A similar experiment with the C3 control peptide showed no binding. Similarly, the C6-2 streptavidin complex showed increased binding to the FcRI transfected IIA1.6 cells compared with the untransfected cells (Figure 5B). In conclusion, a synthetic version of the selected C6-2 peptide was able to target streptavidin to FcRI-expressing U937 cells and FcRIA/-chain transfected IIA1.6 cells.
|
Receptor-mediated internalization
In an approach to determine whether the selected peptide was able to trigger FcRI functions, the biotin-conjugated C6-2 peptide bound to iodinated streptavidin on 1 µm magnetic beads was investigated for the ability to induce receptor-mediated internalization. IFN- stimulated U937 cells were incubated with beads–125I–streptavidin charged with the C6-2 peptide at 37°C for various time periods and intracellular degradation of streptavidin was estimated as described in Materials and methods. Samples of IFN- stimulated cells incubated with beads–125I–streptavidin with C6-2 peptide in the presence of concanamycin A or E64d was included, as concanamycin A and E64d raises the vacuolar pH and inhibits lysosomal cysteine proteinases, respectively, and consequently prevents intracellular degradation of endocytosed material. In each case cell supernatants were treated with trichloroacetic acid, and the radioactivity in the acid-soluble fraction, which represents cell degraded 125I-labelled streptavidin, was counted.
IFN- stimulated cells incubated with beads–125I–streptavidin charged with C6-2 peptide showed a significant increase of streptavidin degradation compared with non-stimulated cells and to stimulated cells incubated with beads–125I–streptavidin charged with the C3 control peptide (Figure 6). The presence of concanamycin A or E64d inhibited streptavidin degradation, indicating that the degradation takes place in intracellular vesicles. In conclusion, this result demonstrates that the selected C6-2 peptide was able to target protein complexes for receptor-mediated vacuolar degradation.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Phage display selection on whole cells has many advantages compared with selection on a soluble protein. It ensures that the selected peptide binds to its target in the presence of many other biological molecules. Furthermore, it allows for selection against membrane proteins that are difficult to express and purify. The latter is the case for human Fc
RI, as the recombinant production of this receptor has proven both by us (Berntzen et al., 2005) and others (Sondermann and Oosthuizen, 2002) to result in very low levels of secreted protein. Therefore, the human monocytic cell line U937 was used as target in an attempt to isolate peptides selectively interacting with FcRI. This cell line constitutively expresses FcRI, FcRIIA and FcRIIB. The level of FcRI was upregulated by IFN- stimulation, making this receptor the most abundant of the FcRs.
Two libraries were included in the selection procedure, Cys6 and a Cys9, displaying six and nine random amino acids, respectively, both constrained by two cysteine residues and expressed as fusion to the phage gpIII coat protein. Both libraries have previously been the source of a number of different peptides with short consensus motifs when using monoclonal antibodies as target (Lauvrak et al., 2004).
The phage peptides selected from the Cys6 library shared a clear consensus motif that showed increased binding to IFN- stimulated U937 cells versus non-stimulated cells. When one of these phages was tested in ELISA for binding to recombinant FcRI, FcRIIA, FcRIIB and FcRIIIB, binding to FcRI only, was observed. Selection from the Cys9 library only gave rise to one single clone that revealed less increase in binding to IFN- stimulated cells versus the non-stimulated cells compared with the selected Cys6 peptides, making this phage a less attractive candidate for further investigations.
It has been reported that synthetic peptides based on the IgG Fc-region are able to bind to human monocytes (Ratcliffe and Stanworth, 1982), human FcRI on human monocytes (U937 cells) and macrophages (THP-1 cells) (Sheridan et al., 1999), to recombinant human FcRIII (Radaev and Sun, 2001) and to a lymphoma cell line (ST486) transfected with human FcRIIB (Medgyesi et al., 2004). All these peptides share the same FcR binding site as intrinsic IgG and must compete with circulating IgGs and ICs for their binding to all FcRs in vivo. In contrast, the binding of the selected Cys6 phages described in this paper was not inhibited by human IgG3, suggesting that the phage presented peptides bind to an epitope on FcRI distinct from the binding site for IgG. Likewise, the two characterized monoclonal anti-FcRIs mAbs, mAb22 and 32.2 (Guyre et al., 1989) did not block the interaction of these phages with U937 cells, indicating that the peptide binds to a fourth epitope on FcRI. Several in vivo studies in mice demonstrate that activating FcR-dependent mechanisms are crucial for the antitumor effect of cytotoxic antibodies (Clynes et al., 2000; Uchida et al., 2004). As FcRIIB engagement inhibits effector cell activation (Dijstelbloem et al., 2001), minimal binding to inhibitory FcRIIB is a clear advantage (Clynes et al., 2000). Thus, in a therapeutic perspective it is highly encouraging that the selected Cys6 phages bind to FcRI but not FcRIIB.
Search in the Swissprot database was performed in order to identify proteins with homology to the consensus sequence. A 100% match was found in the MMP-9 protein (617LRSGRG622), which is a member of the matrix metalloproteinases, a family of enzymes that catalyze degradation of extracellular matrix and basement membrane. Surprisingly, Pasqualini et al. isolated an identical binding motif when searching for peptide binders to MMP-9 (Koivunen et al., 1999) as well as to echovirus 22 (EV22) (Pulli et al., 1997). Based on these results it was suggested that the LRSGRG motif in MMP-9 is involved in dimerization of MMP-9 and, furthermore, that EV22 interacts with MMP-9 via the same motif. Whether this implies that MMP-9 as well as EV22 may interact with FcRI, and the functional relevance of such an interaction, remains to be investigated.
Both C6-1 and C6-2 were produced as cyclic peptides, and MS analysis of the peptides revealed monomeric peptides of expected size. However, gel filtration indicated that the peptide might associate into a non-covalent dimer. Synthetic version of the C6-1 and the C6-2 peptides showed a concentration-dependent inhibition of the C6-1 phage interaction with the U937-cells with >50% reduced recovery at a peptide concentration of 4 µM, suggesting that the peptides are bioactive at concentrations in the lower µM range.
Flow cytometry analysis showed that the synthetic biotin-conjugated C6-2 peptide was able to target FITC-conjugated streptavidin to U937 cells and that the amount of bound streptavidin was increased by IFN- stimulation of the cells. Furthermore, the C6-2 peptide showed increased binding to FcRI transfected IIA1.6 cells compared with untransfected cells. In conclusion, the peptide clearly retained its ability to target FcRI expressing cells in a molecular context different from the phage coat protein III.
FcRI aggregation by ICs is critical to generate signals triggering cell activation. To try to induce aggregation of FcRI in U937 cells by means of the C6-2 peptide, a complex between biotin-conjugated C6-2 peptides and 125I-labelled streptavidin-coupled magnetic beads was prepared. To determine whether the peptide complex was able to promote internalization and transport of the beads to degrading compartments, streptavidin degradation was measured in the presence and absence of concanamycin A and E64d. Degraded streptavidin was represented by acid-soluble radioactivity in the cell cultures. This assay has previously been used to study other receptor ligands and cell types (Mousavi et al., 2005). The results showed that the beads were internalized in a peptide-dependent manner and that 125I-labelled streptavidin was degraded in an intracellular acid compartment, as the degradation was inhibited by concanamycin A or E64d.
In conclusion, we isolated peptides targeting FcRI by use of phage display selection on the human monocytic cell line U937. One of the selected peptides was shown to induce FcRI-mediated internalization of a linked antigen followed by intracellular degradation. The peptide may be applicable to the targeting of vaccine agents or therapeutic drugs to FcRI bearing cells, as well as to the study of receptor function.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
We thank Professor Jan G.J. van de Winkel for providing the Fc
RIA/-chain transfected and untransfected IIA1.6 cells and Professor Peter D. Sun for providing soluble human FcRIIIB. This work was supported by the Research Council of Norway (144971/330 and 155237/300). |
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Berntzen,G., Lunde,E., Flobakk,M., Andersen,J.T., Lauvrak,V. and Sandlie,I. (2005) J. Immunol. Methods, 298, 93–104.[CrossRef][Web of Science][Medline]
Clynes,R.A., Towers,T.L., Presta,L.G. and Ravetch,J.V. (2000) Nat. Med., 6, 443–446.[CrossRef][Web of Science][Medline]
Daeron,M. (1997) Annu. Rev. Immunol., 15, 203–234.[CrossRef][Web of Science][Medline]
Dijstelbloem,H.M., van de Winkel,J.G. and Kallenberg,C.G. (2001) Trends Immunol., 22, 510–516.[CrossRef][Web of Science][Medline]
Fanger,N.A., Voigtlaender,D., Liu,C., Swink,S., Wardwell,K., Fisher,J., Graziano,R.F., Pfefferkorn,L.C. and Guyre,P.M. (1997) J. Immunol., 158, 3090–3098.[Abstract]
Gessner,J.E., Heiken,H., Tamm,A. and Schmidt,R.E. (1998) Ann. Hematol., 76, 231–248.[CrossRef][Web of Science][Medline]
Guyre,P.M., Morganelli,P.M. and Miller,R. (1983) J. Clin. Invest., 72, 393–397.[Web of Science][Medline]
Guyre,P.M., Graziano,R.F., Vance,B.A., Morganelli,P.M. and Fanger,M.W. (1989) J. Immunol., 143, 1650–1655.[Abstract]
Heijnen,I.A., van Vugt,M.J., Fanger,N.A., Graziano,R.F., de Wit,T.P., Hofhuis,F.M., Guyre,P.M., Capel,P.J., Verbeek,J.S. and van de Winkel,J.G. (1996) J. Clin. Invest., 97, 331–338.[Web of Science][Medline]
James,N.D., Atherton,P.J., Jones,J., Howie,A.J., Tchekmedyian,S. and Curnow,R.T. (2001) Br. J. Cancer, 85, 152–156.[CrossRef][Web of Science][Medline]
Kalergis,A.M. and Ravetch,J.V. (2002) J. Exp. Med., 195, 1653–1659.
Kawata,N., Takata,M., Suwaki,T., Tanimoto,Y., Soda,R., Takahashi,K. and Kimura,I. (1996) Int. Arch. Allergy Immunol., 109, 27–34.[Web of Science][Medline]
Keler,T., Guyre,P.M., Vitale,L.A., Sundarapandiyan,K., van De Winkel,J.G., Deo,Y.M. and Graziano,R.F. (2000) J. Immunol., 165, 6738–6742.
Koivunen,E. et al. (1999) Nat. Biotechnol., 17, 768–774.[CrossRef][Web of Science][Medline]
Lauvrak,V., Berntzen,G., Heggelund,U., Herstad,T.K., Sandin,R.H., Dalseg,R., Rosenqvist,E., Sandlie,I. and Michaelsen,T.E. (2004) Scand J. Immunol., 59, 373–384.[CrossRef][Web of Science][Medline]
Machy,P., Serre,K. and Leserman,L. (2000) Eur. J. Immunol., 30, 848–857.[CrossRef][Web of Science][Medline]
Medgyesi,D., Uray,K., Sallai,K., Hudecz,F., Koncz,G., Abramson,J., Pecht,I., Sarmay,G. and Gergely,J. (2004) Eur. J. Immunol., 34, 1127–1135.[CrossRef][Web of Science][Medline]
Mori,T. (2004) Curr. Pharm. Des., 10, 2335–2343.[CrossRef][Web of Science][Medline]
Mousavi,S.A., Sato,M., Sporstol,M., Smedsrod,B., Berg,T., Kojima,N. and Senoo,H. (2005) Biochem. J., 387, 39–46.[CrossRef][Web of Science][Medline]
Nambu,M., Morita,M., Watanabe,H., Uenoyama,Y., Kim,K.M., Tanaka,M., Iwai,Y., Kimata,H., Mayumi,M. and Mikawa,H. (1989) J. Immunol., 143, 4158–4165.[Abstract]
Pan,L.Y., Mendel,D.B., Zurlo,J. and Guyre,P.M. (1990) J. Immunol., 145, 267–275.[Abstract]
Pulli,T., Koivunen,E. and Hyypia,T. (1997) J. Biol. Chem., 272, 21176–21180.
Radaev,S. and Sun,P.D. (2001) J. Biol. Chem., 276, 16478–16483.
Ratcliffe,A. and Stanworth,D.R. (1982) Immunol. Lett., 4, 215–221.[CrossRef][Web of Science][Medline]
Ravetch,J.V. and Bolland,S. (2001) Annu. Rev. Immunol., 19, 275–290.[CrossRef][Web of Science][Medline]
Regnault,A. et al. (1999) J. Exp. Med., 189, 371–380.
Repp,R. et al. (2003) Br. J. Cancer, 89, 2234–2243.[CrossRef][Web of Science][Medline]
Sarmay,G., Lund,J., Rozsnyay,Z., Gergely,J. and Jefferis,R. (1992) Mol. Immunol., 29, 633–639.[CrossRef][Web of Science][Medline]
Scott,J.K. and Smith,G.P. (1990) Science, 249, 386–390.
Sheridan,J.M., Hayes,G.M. and Austen,B.M. (1999) J. Pept. Sci., 5, 555–562.[CrossRef][Web of Science][Medline]
Smith,G.P. and Scott,J.K. (1993) Methods Enzymol., 217, 228–257.[Web of Science][Medline]
Sondermann,P. and Oosthuizen,V. (2002) Biochem. Soc. Trans., 30, 481–486.[CrossRef][Web of Science][Medline]
Uchida,J., Hamaguchi,Y., Oliver,J.A., Ravetch,J.V., Poe,J.C., Haas,K.M. and Tedder,T.F. (2004) J. Exp. Med., 199, 1659–1669.
van de Winkel,J.G. and Anderson,C.L. (1991) J. Leukoc. Biol., 49, 511–524.[Web of Science][Medline]
Zhang,Y., Boesen,C.C., Radaev,S., Brooks,A.G., Fridman,W.H., Sautes-Fridman C. and Sun,P.D. (2000) Immunity, 13, 387–95.[CrossRef][Web of Science][Medline]
Received July 28, 2005; revised December 12, 2005; accepted December 15, 2005.
Edited by Sally Ward
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Bonetto, L. Spadola, A. G. Buchanan, L. Jermutus, and J. Lund Identification of cyclic peptides able to mimic the functional epitope of IgG1-Fc for human Fc{gamma}RI FASEB J, February 1, 2009; 23(2): 575 - 585. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Berntzen, J. T. Andersen, K. Ustgard, T. E. Michaelsen, S. A. Mousavi, J. D. Qian, P. E. Kristiansen, V. Lauvrak, and I. Sandlie Identification of a High Affinity Fc{gamma}RIIA-binding Peptide That Distinguishes Fc{gamma}RIIA from Fc{gamma}RIIB and Exploits Fc{gamma}RIIA-mediated Phagocytosis and Degradation J. Biol. Chem., January 9, 2009; 284(2): 1126 - 1135. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Flobakk, I. B. Rasmussen, E. Lunde, T. Frigstad, G. Berntzen, T. E. Michaelsen, B. Bogen, and I. Sandlie Processing of an Antigenic Sequence from IgG Constant Domains for Presentation by MHC Class II J. Immunol., November 15, 2008; 181(10): 7062 - 7072. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||