HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Full Text (PDF) Supplemental Data All Versions of this Article: 18/14/1704 04-1513fjev1    most recent Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by SCHWARZ, M. Articles by PETER, K. Search for Related Content PubMed PubMed Citation Articles by SCHWARZ, M. Articles by PETER, K.

Single-chain antibodies for the conformation-specific blockade of activated platelet integrin _IIbß₃ designed by subtractive selection from naive human phage libraries

MEIKE SCHWARZ^*,,1, PETER RÖTTGEN, YOSHIAZUKU TAKADA, FABRICE LE GALL, STEFAN KNACKMUSS, NICOLE BASSLER^*, CLAUDIA BÜTTNER, MELVYN LITTLE, CHRISTOPH BODE^* and KARLHEINZ PETER

^* Department of Cardiology, University of Freiburg, Freiburg, Germany;
The Scripps Research Institute, La Jolla, California, USA;
Affimed Therapeutics AG, Heidelberg, Germany; and
University of North Carolina at Chapel Hill, North Carolina, USA

¹Correspondence: Department of Cardiology, University of Freiburg, Breisacherstr. 33, Freiburg 79106, Germany. E-mail: schwarz{at}med1.ukl.uni-freiburg.de

SPECIFIC AIMS

The aim of this study was to design a new type of _IIbß₃ inhibitor specific for the activated conformation of this platelet integrin receptor. For this purpose we established a phage display strategy allowing the selection of single chain antibodies (scFv) against function-dependent epitopes on complex receptors from human naive phage libraries.

PRINCIPAL FINDINGS

1. Selection of conformation-dependent phages from a naive natural and a naive synthetic library using a novel subtractive panning protocol
Two distinct naive, highly complex scFv-phage libraries were constructed: 1) a natural library with randomly recombined human V_H and V_L-sequences (complexity=1.8x10⁹); and 2) a synthetic library with randomized complementary-determining regions 3 (CDR3) of the heavy chain of two well established scFv frameworks (complexity=7.1x10⁸). For selection of activation specific anti-_IIbß₃ clones from these large libraries, we established a novel cell-based panning strategy (Fig. 1 ). The libraries were first preabsorbed on cells expressing the nonactivated receptor and then selection of phages binding specifically on the activated _IIbß₃ was performed on cells expressing the activated receptor.

View larger version (51K): [in this window] [in a new window]   Figure 1. Scheme of the differential panning strategy. This strategy was developed to reduce nonspecific (depicted in gray) and nonactivation specific binding phages (in blue) and to specifically select phages binding to epitopes on IIbß3 expressed solely on the activated receptor (in red). The first round (upper row) was performed on human platelets. Initially a depletion step was performed, where all phages that bound to nonactivated IIbß3 or the platelet background were separated by centrifugation and discarded. The supernatant was used for the next selection step in which the nonbinding phages in the supernatant were discarded and the binding phages were rescued and eluted by lowering the pH or by competition with the IIbß3 blocker abciximab. The phages were then amplified in Escherichia coli and used for the next round. In the ensuing two rounds, the cell background was changed by using recombinant IIbß3-expressing CHO cells to avoid enrichment of phages binding to activation-specific platelet epitopes other than on the IIbß3 molecule.

An important precondition for a successful depletion and selection was to obtain the receptor reliably in an activated or nonactivated conformation. With this in mind, we used two sources of _IIbß₃: 1) washed human platelets which were activated with ADP when needed; and 2) recombinant _IIbß₃-expressing CHO cells, where the activated conformation could be achieved by introduction of a VGFFK-deletion in the cytoplasmic tail of _IIb.

By using these two cell types with distinct cell surface backgrounds in subsequent rounds, nonspecific binding was reduced. We performed washing steps at low pH for the reduction of unspecific binding.

Selected antibodies were further characterized in flow cytometry on nonactivated and ADP-activated human platelets (Fig. 2 ). About 20% of natural and 50% of synthetic clones revealed binding exclusive to the activated receptor. Sequencing of newly obtained activation-specific clones revealed that nearly all contained an RXD-sequence in their V_H CDR3 in analogy to the _IIbß₃-binding RGD sequence in fibrinogen. With one exception, natural clone MA4, which also does not contain an RXD-pattern, all scFvs competed with the binding of GRGDSP-peptides. Thus, these scFvs are likely to bind at the fibrinogen-binding site.

View larger version (20K): [in this window] [in a new window]   Figure 2. Binding in flow cytometry. Sample flow cytometry histograms of scFv binding on resting (gray histograms) and ADP-stimulated (white histograms) platelets.

2. Affinity maturation of synthetic clones by V_L CDR3 optimization
In contrast to clones of the natural library, clones obtained from the synthetic library demonstrate rather weak binding, which might be explained by their limited sequence variability (only V_H CDR3 was randomized). Nevertheless, use of a synthetic library seems to be desirable, since it offers various additional advantages (high yield expression of stable scFv with low immunogenicity, and the possibility of affinity improvement by CDR-shuffling). We carried out a secondary panning with a light-chain sublibrary of selected primary clones and achieved optimization of the V_L CDR3. This time, we modified the panning procedure with the goal of selecting phages with high affinity. In contrast to the first specificity-focused panning protocol, we omitted a depletion step. We employed a smaller quantity of cells (10⁵ instead of 4x10⁶) to enhance competition between phages. In the fifth round we performed a competitive elution with gradually increasing concentrations of the _IIbß₃ inhibitor eptifibatide. All resulting clones demonstrated activation-specific binding and many demonstrated binding in flow cytometry comparable to the natural clones.

3. Selected scFvs demonstrate an inhibitory potency equal to presently used _IIbß₃ blockers
We evaluated whether the obtained scFvs were capable of inhibiting fibrinogen binding in a comparable manner to existing _IIbß₃ blockers. Representative antibody-clones, MA2 and MA4 (natural clones), SA2 (synthetic clone), and SCE5 (synthetic clone with optimized LCDR3) were transferred to the expression vector pHOG-21, produced, and purified at a large scale. Binding of fibrinogen to ADP-stimulated platelets at various scFv concentrations was measured in flow cytometry. As depicted in Fig. 3 , MA2 completely inhibited fibrinogen binding with half-maximal binding at 1.1 µg/mL (36.6 nM). The clinically used _IIbß₃ blocker abciximab demonstrated a comparable inhibition (half maximal at 0.75 µg/mL (15 nM)). One of the best synthetic clones (SCE5) also had a comparable inhibitory potency with an IC₅₀ of 1.2 µg/mL (37 nM). Comparison with the primary clone SA2 clearly demonstrates the improvement of affinity by optimization of the V_LCDR3. As a final functional evaluation, effective inhibition of platelet aggregation induced by ADP, collagen, and thrombin was demonstrated using light transmission aggregometry.

View larger version (14K): [in this window] [in a new window]   Figure 3. Inhibitory potency of the generated scFvs. Fibrinogen binding to activated platelets at increasing concentrations of abciximab (A), scFv MA2 (B), scFv MA4 (C) and the scFvs SA2 and SCE5 (D), as detected in flow cytometry.

4. Selected clones demonstrate a binding site close to the one of natural _IIbß₃ ligand fibrinogen
To obtain more detailed information about the binding sites of these antibodies, we performed binding studies with mutant _IIbß₃, transiently expressed on the cell surface of CHO cells. First, we tested mouse-human domain switch mutants. Individual regions of human _IIbß₃ were "murinized" by the exchange of all nonhomologous amino acids within the respective region. In the ß₃ subunit, mutants W129S/N133T (B3) and D179N/T182N/T183A (B6) demonstrated considerably reduced antibody binding, indicating that these regions are important for antibody binding. No single mutation induced complete inhibition of binding of clones MA2 and SCE5. Thus, we also tested the combination of mutations B3 and B6, which demonstrates a complete inhibition of binding for all clones. In the _IIb subunit we detected discrete differences between distinct clones, especially for the A16 (V156A/N158S/D159) mutant, which does bind clone MA2 unaltered, but does not bind clones MA4, SA2 and SCE5.

For a detailed analysis of ligand-binding sites on the _IIb subunit, we used several alanine exchange point mutations based on a computed model of this region (ß-propeller model). Several discontinuous individual residues in this region are critically involved in fibrinogen binding of _IIb. We performed binding studies with the newly generated scFvs on these point mutants. Results of the binding studies with point mutants of tested activation-specific scFvs matched with binding data obtained for fibrinogen. Mapping studies indicate almost identical binding sites of newly generated scFvs and fibrinogen.

CONCLUSIONS AND SIGNIFICANCE

In this paper we describe for the first time _IIbß₃-blocking single-chain antibodies that bind to the receptor selectively in its activated state and are completely of human origin.

Until now phage display was mainly performed on immobilized target proteins, which made selection of conformation dependent antibodies on complex transmembrane receptors difficult. Libraries from immunized donors or with default sequence patterns were used in a previous study to achieve reduction of unspecific binding phages. This approach results in a limitation of resulting sequence variability. The protocol presented here combines a large initial library complexity with a rigorous selection procedure that allows the selection of highly conformation specific antibodies.

Human single-chain fragments obtained in this study are a promising basis for a new type of _IIbß₃-blocking drug with several important advantages, compared with the currently used _IIbß₃ blocker: 1) activation-specific blockers are not expected to inhibit platelet adhesion to immobilized fibrinogen which is exposed on the injured vessel wall, and thus formation of a platelet monolayer at the site of vascular injury may still be possible. 2) Activation-specific scFvs do not bind and thus do not induce LIBS expression on resting platelets. Since thrombocytopenia is mainly attributed to an immune response directed against LIBS epitopes, _IIbß₃ inhibitor-induced thrombocytopenia may therefore be avoided. 3) scFv antibodies, which have already been used successfully clinically, demonstrate several advantages in comparison to mouse derived antibodies or their fragments (e.g., Fab, smaller size, complete human origin, negligible immunogenicity, facility of genetic engineering, and further optimization).

Antibody-fragments obtained here represent useful tools for the investigation of integrin-ligand interactions, markers of integrin and platelet activation, and promise to be the basis for development of a new type of activation-specific _IIbß₃ inhibitor. This is the first report of successful cloning of conformation-specific human single-chain antibodies directed against epitopes on a cell surface from nonimmune phage scFv libraries. This method promises to be generally applicable for cell surface molecules, especially integrins.

FOOTNOTES

To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.04-1513fje;

This article has been cited by other articles:

				Z. Li, M. A. Nardi, J. Wu, R. Pan, W. Zhang, and S. Karpatkin Platelet Fragmentation Requires a Specific Structural Conformation of Human Monoclonal Antibody against {beta}3 Integrin J. Biol. Chem., February 8, 2008; 283(6): 3224 - 3230. [Abstract] [Full Text] [PDF]

				P. Stoll, N. Bassler, C. E. Hagemeyer, S. U. Eisenhardt, Y. C. Chen, R. Schmidt, M. Schwarz, I. Ahrens, Y. Katagiri, B. Pannen, et al. Targeting Ligand-Induced Binding Sites on GPIIb/IIIa via Single-Chain Antibody Allows Effective Anticoagulation Without Bleeding Time Prolongation Arterioscler. Thromb. Vasc. Biol., May 1, 2007; 27(5): 1206 - 1212. [Abstract] [Full Text] [PDF]

				S. U. Eisenhardt, M. Schwarz, N. Schallner, J. Soosairajah, N. Bassler, D. Huang, C. Bode, and K. Peter Generation of activation-specific human anti-{alpha}M{beta}2 single-chain antibodies as potential diagnostic tools and therapeutic agents Blood, April 15, 2007; 109(8): 3521 - 3528. [Abstract] [Full Text] [PDF]

				N. Bassler, C. Loeffler, P. Mangin, Y. Yuan, M. Schwarz, C. E. Hagemeyer, S. U. Eisenhardt, I. Ahrens, C. Bode, S. P. Jackson, et al. A Mechanistic Model for Paradoxical Platelet Activation by Ligand-Mimetic {alpha}IIb{beta}3 (GPIIb/IIIa) Antagonists Arterioscler. Thromb. Vasc. Biol., March 1, 2007; 27(3): E9 - E15. [Abstract] [Full Text] [PDF]

				M. Schwarz, G. Meade, P. Stoll, J. Ylanne, N. Bassler, Y. C. Chen, C. E. Hagemeyer, I. Ahrens, N. Moran, D. Kenny, et al. Conformation-Specific Blockade of the Integrin GPIIb/IIIa: A Novel Antiplatelet Strategy That Selectively Targets Activated Platelets Circ. Res., July 7, 2006; 99(1): 25 - 33. [Abstract] [Full Text] [PDF]

This Article Full Text (PDF) Supplemental Data All Versions of this Article: 18/14/1704 04-1513fjev1    most recent Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by SCHWARZ, M. Articles by PETER, K. Search for Related Content PubMed PubMed Citation Articles by SCHWARZ, M. Articles by PETER, K.