Large-scale production, bacterial localization assessment and immobilized metal affinity chromatography purification of a human single-chain Fv antibody against alphaIIb-beta3 integrin.

Our objective was to investigate the Escherichia coli localization (such as supernatant, cytoplasm and inclusion bodies) of an anti-alphaIIb-beta3 (alphaIIbbeta3) scFv fragment referred to as scFv[EBB3] produced in batch fermentation. Immobilized metal affinity chromatography (IMAC) purification was performed on supernatant using expanded bed absorbed technology (EBA) and on sonicated cells in native conditions over an immobilized copper-ion affinity column. Inclusion bodies were solubilized before IMAC purification and the refolding procedure was performed on the column. The majority of scFv[EBB3] were present as inclusion bodies (55%), whereas 36% were found in the cytoplasm and only 9% secreted in the supernatant. The scFv activity was assessed by enzyme-linked immunosorbent assay (ELISA), flow cytometry and immunohistochemistry analyses performed on a thrombus induced in vivo on an atherosclerotic rabbit model.

Improvement in production and purification bioprocesses of bacterially expressed anti-alphaIIbbeta3 human single-chain FV antibodies.

Production of anti-alphaIIbbeta3 (anti-alphaIIbbeta3)-binding single-chain FV (scFv) fragments obtained from combinatorial libraries of IgG human antibodies is of broad interest for imaging and treatment of acute coronary syndromes. The objective of our work was to design an optimized production of one selected anti-alphaIIbbeta3-binding scFv fragment for subsequent in vivo animal studies. Fed-batch fermentation was initiated with 2TY media supplemented with 0.1 M glucose. This growing batch culture was used as a starting point for further fed-batch induction, in which a media without glucose containing 1 mM IPTG and 0.4 M saccharose was continuously added. Subsequent purification was performed on the whole cell extract in native conditions over an immobilized copper-ion affinity column. The improved conditions allowed the recovery of 5 mg of highly purified scFv fragments as determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). The bioactivity of the scFv fragments was further monitored by ELISA, cytometric and immunohistochemical methods.

Analysis of the V genes coding for a monospecific human antibody to myosin and functional expression of single chain Fv fragments.

A monospecific human IgM monoclonal antibody (mAb), reactive with myosin from human heart, has been obtained by EBV transformation. This mAb may have a diagnostic potential in the imaging of myocardial necrosis. However, owing to the fact that the molecular mass of an IgM is 900 kDa, a poor diffusion and a slow penetration inside necrotic myocytes could reduce its capacity for scintigraphic detection. In order to alleviate these problems, we constructed the scFv by cloning the VH and VL domains into the pHOG21 vector. Analysis of the V genes proved an unmutated configuration showing that the immortalized B cell issued from the primary IgM repertoire. The expression product in Escherichia coli was a 35 kDa scFv fragment with the antigen-binding specificity of the parental mAb.

PAICA: a method for characterizing platelet-associated antibodies--its application to the study of idiopathic thrombocytopenic purpura and to the detection of platelet-bound c7E3.

In idiopathic thrombocytopenic purpura (ITP), autoantibodies reacting with antigens on the platelet membrane bring about accelerated platelet destruction. We now report PAICA ("Platelet-Associated IgG Characterization Assay"), a method for detecting autoantibodies bound to specific membrane glycoproteins in total platelet lysates. This monoclonal antibody (MAb) capture assay takes into account the fact that antibodies on circulating platelets may be translocated to internal pools as well as being on the surface. A total of twenty ITP patients were examined by PAICA, and the results compared with those obtained by measuring (i) serum antibodies bound to paraformaldehyde-fixed control platelets by ELISA, (ii) IgG bound to the surface of the patient's own platelets by flow cytometry (PSIgG), (iii) total platelet-associated IgG (PAIgG) by ELISA and (iv) serum antibodies reacting with control platelets by MAIPA ("Monoclonal Antibody-specific Immobilization of Platelet Antigens"). Of twelve patients with elevated PAIgG, nine had increased PSIgG yet eleven reacted positively in PAICA. Of these, eight possessed antibodies directed against GP IIb-IIIa, two against GP Ib-IX and one patient possessed antibodies directed against GP IIb-IIIa and GP Ia-IIa respectively. Only seven of the patients possessed serum antibodies detectable by MAIPA. PAICA was also able to detect platelet-associated c7E3 (the chimeric form of Fab fragments of the MAb 7E3) following its infusion during antithrombotic therapy, when it proved more sensitive over a seven-day period than a MAIPA assay adapted for assessing surface-bound antibody. We propose that PAICA provides added sensitivity to the detection of platelet-associated antibodies in immune thrombocytopenias or following therapy with humanized MAbs.

Incidence of anti-mouse antibodies in thrombocytopenic patients with autoimmune disorders.

Idiopathic thrombotycopenic purpura (ITP) is an autoimmune disorder in which circulating autoantibodies react with target antigens on the platelet membrane. In order to identify the autoimmune response in ITP, two MAIPA (Monoclonal Antibody (MAb) Immobilization of Platelet Antigen) assays (MAIPA I and MAIPA II) were performed on sera from thrombocytopenic patients. In the classic MAIPA assay (MAIPA I), control platelets were incubated simultaneously with human serum and a mouse MAb to a platelet glycoprotein. In MAIPA II, the control platelets were incubated first with the human serum and then, after washing, with the selected mouse MAb. A positive MAIPA I test but a negative MAIPA II has been shown to result from the presence of serum antibodies recognizing mouse MAb to platelet glycoproteins used in the assay. We compared the frequency of such 'anti-mouse' antibodies in patients with thrombocytopenia associated or not with other autoimmune states and in healthy donors with a normal platelet count. Statistically significant differences were found in the incidence of anti-mouse antibodies between patients and healthy donors. Furthermore, the identity of the targeted mouse MAbs varied in sera from the patients. The detected anti-mouse antibodies may include anti-idiotypic antibodies produced against cross-reactive idiotypes shared by human and mouse anti-platelet antibodies.

Analysis of VH and VL genes of a monospecific human anti-myosin antibody produced by a B cell from the primary repertoire.

Epstein-Barr virus (EBV) transformation of B lymphocytes from a Glanzmann's thrombasthenia patient with a serum antibody to the integrin alpha IIb beta 3, led to the immortalization of a B cell secreting a monospecific IgM monochonal antibody (MAb), B7, reactive with platelet myosin. Analysis of B7 V genes revealed minimally mutated sequences: the immortalized B cell issued from the primary repertoire, with no evidence of an in vivo selection by myosin. The V genes were here compared with sequences of human MAbs available on databases to more clearly understand the monospecificity of the B7 MAb. B7 V genes were closely identical to rearranged V genes in clones with self-specificities, often secreting polyreactive antibodies. In contrast, B7 is an unmutated monoreactive human MAb able to recognize myosin with a high avidity. Comparison of the CDR3H sequence with that of MAbs in databases supports a central role for the CDR3H subdomain in determining monospecificity. Our results suggest the existence of a monospecific autoreactive B cell compartment, besides the well-known polyspecific one, susceptible to be the template of pathogenic autoreactivity, characterized by antibodies of high affinity and specificity.

Rapid screening of purification strategies for the capture of a human recombinant F(ab')2 expressed in baculovirus-infected cells using a micro-plate approach and SELDI-MS.

The development of a capture step of a human recombinant F(ab')(2) produced and expressed in baculovirus-infected cells was investigated by screening three mixed-mode chromatography sorbents (HEA HyperCel, PPA HyperCel and MEP HyperCel) and two ion exchangers (Q Ceramic HyperD F, S Ceramic HyperD F) sorbents using a 96-well plate format and SELDI-MS. HEA HyperCel gave the best separation performance therefore the conditions tested in micro-plate were transferred to laboratory scale chromatographic experiments, confirming that the recombinant F(ab')(2) was effectively captured on the mixed-mode sorbent without any pre-treatment of the crude extract with a 82% recovery and a 39-fold purification.

Three-step purification of bacterially expressed human single-chain Fv antibodies for clinical applications.

We have obtained a cell line which secretes a human monoclonal IgM (B7) reacting with the myosin heavy chain of human heart. We have constructed single-chain fragments (scFv) of B7. The scFv may be useful for the imaging of myocardial necrosis after myocarditis, cardiac drug toxicosis or graft rejection. The aim of our work was to purify the scFv for immunoscintigraphy. We describe several purification steps including immobilized metal affinity chromatography (IMAC), anti-c-myc monoclonal antibody affinity chromatography, size-exclusion chromatography with Superdex 75 HR 10/30 and ion-exchange chromatography (mini Q TM 30Q).

Further evidence that heparin-dependent thrombocytopenia may result from Fc receptor-mediated interactions.

Heparin-dependent thrombocytopenia (HDT) and associated thrombotic complications, are thought to be linked to the appearance of anti-platelet antibodies. Attempts were made to characterize the antibodies in the sera from 10 such patients. Western blotting against platelet antigens was inconclusive, often revealing multiple bands but with considerable variability from patient to patient. An often seen band of approximately 83 kDa was also given by some nonimmune sera and the antigen appeared to be predominantly intracellular in origin. Antibodies to major membrane glycoproteins were not readily apparent. This was confirmed by the MAIPA ("Monoclonal Antibody Immobilization of Platelet Antigens") test, performed to detect antibodies to GP Ib-IX and GP IIb-IIIa. Only one weak activity to GP Ib-IX and one weak activity to GP IIb-IIIa were detected. In contrast, an antibody to the PlA1 alloantigen was readily detected in the serum of an additional patient. The ability of HDT serum to induce the aggregation of control platelets was studied in detail. Aggregation was inhibited by EDU-3, a monoclonal antibody to GP IIb-IIIa complexes, and by the synthetic peptide RGDS, suggesting an involvement of the same pathway as used by physiologic agonists. However, in agreement with Chong et al (Thromb Res 55:291, 1989), we observed that aggregation was inhibited by rabbit IgG, suggesting that it was Fc-receptor mediated. Interestingly, 2E1, a monoclonal antibody to the Fc gamma RII receptor, induced platelet aggregation with a lag phase, a characteristic of HD antibodies. Nonetheless, for different donors, there was no correlation between the length of the lag phase induced by 2E1 and HD antibodies. Fc-receptor blockade should be considered as a means for diminishing the clinical complications of HDT.

A patient with autoimmune thrombocytopenic purpura demonstrating serum antibodies reactive with mouse cross-reactive idiotypes.

Autoimmune thrombocytopenic purpura (ATP) is a syndrome of destructive thrombocytopenia due to platelet-binding antibodies. We report the case of a young woman (C.V.) who has a history of chronic ATP with severe but transient bouts of thrombocytopenia. Using the classic monoclonal antibody (MAb) immobilization of platelet antigens (MAIPA) assay to screen serum antibody specificity, results were strongly positive with MAbs to glycoproteins (GPs) Ib-IX, Ia-IIa, IV, and p24, but weakly positive or negative for GP IIb-IIIa. In contrast, a two-step incubation assay (MAIPA II), in which platelets were incubated sequentially with C.V. serum and the murine MAb, gave negative results for all GPs. Affinity chromatography performed using Bx-1, a MAb to GP Ib, showed that the patient's serum contained antibodies to determinants expressed by mouse immunoglobulins. These were present on Fab fragments on Bx-1. A survey of sera from other patients with thrombocytopenia of immune origin revealed that antibodies reactive with selected idiotypes of mouse MAbs were not infrequent and raises the question of their role in the thrombocytopenia.

The ex vivo production of human monoclonal antibodies to glycoprotein IIb-IIIa complexes of blood platelets.

The integrin alpha IIb beta 3 (GPIIb-IIIa complex) of blood platelets mediates platelet aggregation by binding adhesive proteins which form bridges between activated cells. This same process is implicated in arterial thrombosis. The goal of our research is to take B-lymphocytes from patients possessing inhibitory antibodies to GPIIb-IIIa and develop technology permitting their production ex vivo. Starting point is the peripheral blood from two patients with Glanzmann's thrombasthenia, an inherited disorder in which platelets lack these complexes, and where high titre antibodies to GPIIb-IIIa have formed following contact with normal platelets after transfusion and/or pregnancy. We describe a strategy of in vitro stimulation to overcome the following constraints: (i) peripheral blood contains a low concentration of antigen-reactive specific B-cells, and (ii) the circulating B-cells are arrested in a phase in which additional stimuli are required to induce antigen-specific clonal activation. Optimal conditions involve the use of a combination of growth factors, polyclonal activators and soluble GPIIb-IIIa prior to the fusion of activated B-cells with either (a) the murine myeloma cell line X63 Ag 8,653 or (b) the heteromyeloma cell line SPM4-0. In this way, we have obtained several cell lines secreting antibodies specific for the GPIIb-IIIa complex. Our next aim is to rescue the relevant human immunoglobulin genes from these hybridoma cells.

Characterization and application of new macroporous membrane ion exchangers.

A new ready-to-use unit for high-performance membrane chromatography has been characterized. Its dynamic capacity, resolving power and protein recovery were measured at different flow-rates. The binding capacity was 0.5-2 mg/cm2 with a 95% recovery at 10 ml/min irrespective of the protein concentration up to 10 mg/ml. For very-high flow-rates (50 and 100 ml/min) the recovery was 90% and 70%. At these flow-rates, the maximum back-pressure was about 0.1 MPa and was independent of the filtration area. By increasing the filtration area, a proportional capacity increase was obtained, indicating an easy scale-up. High flow-rates had only a slight effect on resolution. This new adsorber was able to purify IgM from supernatant of cell culture of a human hybridoma in less than 8 min with a high degree of purity (95%).

Autoantibodies and anti-mouse antibodies in thrombocytopenic patients as assessed by different MAIPA assays.

Two MAIPA (monoclonal antibody [MAb] immobilization of platelet antigen) assays were performed to determine (a) autoantibodies to platelet glycoproteins (GP) and (b) serum antibodies recognizing mouse MAbs used in the assay. In MAIPA I, control platelets were incubated simultaneously with human serum and a mouse MAb to a platelet glycoprotein (GP IIb-IIIa, Ib-IX, Ia-IIa, IV and p24). In MAIPA II, the control platelets were incubated first with the human serum and then, after washing, with the selected mouse MAb. A series of 25 patients with autoimmune thrombocytopenic purpura (ATP) associated or not with other autoimmune states were examined. Autoantibodies (both MAIPA I and MAIPA II positive) or anti-mouse Abs (MAIPA I positive and MAIPA II negative) were frequent in both groups of patients. Statistically significant differences existed in the incidence of anti-mouse Abs between patients (56.5%) and healthy donors (10%). This suggests that their production may be related to thrombocytopenias associated with autoimmune disease. We speculate that the presence of anti-mouse antibodies could reflect an abnormality in the immunological modulation of the idiotypic network.

A close spatial relationship between GP IIb-IIIa complexes and CD9 antigen as demonstrated by the MAIPA technique.

CD9 is a well-defined component of the platelet plasma membrane and has a copy number almost equivalent to that of glycoprotein (GP) IIb-IIIa complexes, the aggregation receptor on platelets. It has an apparent molecular mass of 24 kD and is otherwise known as p24. Stimulation of p24 by monoclonal antibodies (MAb) induces platelet aggregation and granule release, involves FcγRII, and is mainly mediated through the stimulation of phospholipase C. In accordance with a signalling function, p24 has been reported to associate with small GTP-binding proteins and to GP IIb-IIIa complexes upon activation. We now report further evidence of a strong relationship between p24 and GP IIb-IIIa in platelets. Using the MAIPA (monoclonal antibody immobilization of platelet antigens) assay in the screening of human antibodies to platelet glycoproteins, we found that GP IIb-IIIa-antibody complexes were almost invariably associated with p24 in the harvested detergent-soluble fraction of platelet lysates. Thus, associated human antibodies were detected following the targeting of either GP IIb-IIIa or p24 by monospecific murine monoclonal antibodies (MAbs). This is a point to bear in mind when assessing for antibodies to p24 or GP IIb-IIIa in immune thrombocytopenias.

A human monoclonal antibody obtained from EBV-transformed B cells with specificity for myosin.

We describe the preparation of a stable human lymphoblastoid cell line obtained during ex vivo studies in which peripheral blood lymphocytes of a Glanzmann's thrombasthenia patient were transformed with Epstein-Barr virus. Somatic hybrids secreted an IgM monoclonal antibody (B7) that reacted with the myosin heavy chain of human platelets by immunoblotting. Flow cytometry showed that B7 barely recognized unstimulated intact platelets, but bound abundantly after permeabilization of fixed cells with Triton X-100. The reactivity of the antibody on thin sections of human myocardium and aorta was studied by immunohistochemistry. B7 specifically stained myosin of myocytes, but there was no labelling of aortic smooth muscle cells. The epitope was conserved in cardiac or skeletal myosin prepared from pig or rabbit. Measurement of the dissociation constant in a competitive ELISA showed that B7 bound with high affinity (10(-8) M). Purified Fab fragments retained their ability to bind to myosin, suggesting that B7 may be useful in the imaging of myocardial necrosis after myocardial infarction, myocarditis, cardiac drug toxicosis or graft rejection. This work also shows that EBV transformation of B cells may uncover naturally occurring autoantibodies which under normal circumstances are inhibited by the immune surveillance system.

Autoimmune thrombocytopenic purpura (AITP) and acquired thrombasthenia due to autoantibodies to GP IIb-IIIa in a patient with an unusual platelet membrane glycoprotein composition.

The subject (E.B.) is a 63-year-old woman with autoimmune thrombocytopenic purpura (AITP) who was first examined some 6 years ago with symptoms of epistaxis and gum bleeding, severe thrombocytopenia, and large platelets. Her serum tested positively with control platelets in the MAIPA assay performed using monoclonal antibodies (MoAb) to glycoprotein (GP) IIIa (XIIF9, Y2/51), yet was negative in the presence of MoAbs to GP IIb (SZ 22) or to the GP IIb-IIIa complex (AP2, P2). The patient's platelets failed to aggregate with all agonists tested except for ristocetin. IgG isolated from the patient's serum inhibited ADP-induced aggregation of control platelets. Unexpectedly, flow cytometry showed an altered expression of membrane glycoproteins on the patient's platelets. Levels of GP Ib-IX were much higher than previously located by us in platelets. In contrast, the expression of GP IIb-IIIa was about half that seen with control subjects. When Western blotting was performed, a striking finding was a strong band of 250 kDa recognized by a series of MoAbs to GP Ib alpha in addition to the band in the normal position of GP Ib alpha. Finally, ADP-stimulated (E.B.) platelets failed to express activation-dependent epitopes on GP IIb-IIIa as recognized by PAC-1, AP6, or F26 and additionally gave a reduced P-selectin expression after thrombin addition. In conclusion, we present a novel patient with a severely perturbed platelet function where an altered membrane GP profile is associated with the presence of an autoantibody recognizing a complex-dependent determinant on GP IIb-IIIa and inhibitory of platelet aggregation.

Anti-platelet antibodies in patients with systemic lupus erythematosus and the primary antiphospholipid antibody syndrome: their relationship with the observed thrombocytopenia.

The role of antiphospholipid antibodies in the pathogenesis of the thrombocytopenia observed during primary antiphospholipid antibody syndrome (APAS) and systemic lupus erythematosus (SLE) remains controversial. We have used the MAIPA test to examine the frequency and specificity of anti-platelet antibodies directed against the major platelet membrane glycoproteins (GP IIb-IIIa, GP Ib-IX, GP Ia-IIa and GP IV) in patients where SLE and APAS were associated or not with thrombocytopenia. Results were compared with a series of 26 ITP patients, 46% of whom were shown to possess anti-platelet antibodies directed against one or more of the platelet surface glycoproteins. When APAS was associated with thrombocytopenia, 7/10 patients possessed antibodies against GP IIb-IIIa and/or GP Ib-IX. For SLE patients with thrombocytopenia, 6/10 patients were shown to have antiplatelet antibodies against GP IIb-IIIa, GP Ib-IX or GP IV. In contrast, for APAS (n=11) and SLE patients (n=11) without thrombocytopenia, only one patient had an antibody directed against GP IIb-IIIa and one patient had an antibody to GP IV. Our results suggest that antibodies directed against major platelet membrane glycoproteins may play a role in the thrombocytopenia that is seen during SLE and APAS.

Characterisation, cloning and sequencing of a conformation-dependent monoclonal antibody to the alphaIIbbeta3 integrin: interest for use in thrombus detection.

The detection of newly formed thrombi is of primary importance in clinical medicine. The activated platelet is a potential target for the localization of thrombotic lesions in arteries. The integrin alpha(IIb)beta(3) membrane changes conformation upon activation. A novel anti-alpha(IIb)beta(3) monoclonal antibody (MAb), XIIF9, is described which recognizes an epitope whose expression was enhanced by activation. Radioiodinated XIIF9 bound to a single class of sites on the beta(3) subunit, with 13600 +/- 2000 molecules bound per unstimulated platelet and a K(d) of 34.5 nM. Platelets stimulated with 0.5 U/ml of thrombin bound 66000 +/- 4000 molecules/cell (K(d) = 51.6 nM). Moreover, XIIF9 binding to unstimulated platelets could be increased 4-fold by treatment of the alpha(IIb)beta(3) complex with 5 mM EDTA. Thus, XIIF9 recognized an epitope on the beta(3) subunit whose accessibility was increased upon thrombin activation or EDTA treatment. Sequence analysis of the gene segment encoding the XIIF9 heavy chain revealed interesting motifs shared with cyclic CX6-7C anti-alpha(IIb)beta(3) peptides or with AC7, a published MAb specific for activated alpha(IIb)beta(3). In vivo experiments in atherosclerotic rabbits followed by immunohistological analysis, revealed a specific binding of XIIF9 on platelets engaged in thrombus formation, demonstrating real clinical potential for such MAbs in imaging.

Improving selection of alphaIIbbeta3-binding phage antibodies with increased reactivity derived from immunized donors.

Although many studies of the immune response in polytransfused Glanzmann thrombasthenia (GT) patients and in autoimmune thrombocytopenic purpura (AITP) have demonstrated the frequent development of Abs directed against the alphaIIbbeta3 integrin, little is known about the induced anti-alphaIIbbeta3 autoantibodies at the molecular level. Phage display is a powerful technology for selecting and engineering mAbs expressed on the surface of filamentous bacteriophage. Combinatorial libraries of single-chain IgG were constructed from splenocytes from two patients with AITP and one patient with GT. In a previous study, activated platelets or alphaIIbbeta3-expressing CHO cells selection was performed to isolate human IgG anti-alphaIIbbeta3 binding fragments using combinatorial libraries created from the B cells of a GT and an AITP patient. However, we have experienced practical problems such as enrichment of truncated antibodies during selection. We decided to test prolonged treatments with elution agents after screening on the purified form of the alphaIIbbeta3 integrin activated with the RGD peptide. We obtained a higher percentage of clones with full-size antibody fragments as well as an enrichment of more specific alphaIIbbeta3-binding phage-Abs. Some of them, recognizing the activated form of the integrin, would be interesting to further study as potential diagnostic or therapeutic agents in acute coronary syndromes. Sequencing of selected phage-Abs revealed that they used different VH and VL genes with, for the majority of them, a high level of extensive hypermutations in the complementarity determining regions, indicating the diversity of the antigen-driven immune response that occurred in GT and AITP patients.