Composition of antibodies with high ADCC

The present invention relates to a composition comprising a monoclonal antibody, characterized in that the antibody has, on its glycosylation site (Asn 297) of Fcγ, glycan structures selected from the forms:<chemistry id="chem0001" num="0001"><img file="EP1518864B1_D0001.tif" /></chemistry>wherein the content of G0 + G1 + G0F + G1F forms is greater than 60% and the content of G0F + G1F forms is less than 50%.

The present description also describes a process for obtaining and selecting monoclonal antibodies by an ADCC type test, said antibodies being able to activate type III Fcγ receptors. The present description also discloses monoclonal antibodies having a particular glycan structure, the cells producing said antibodies, the methods for preparing the producing cells, as well as pharmaceutical compositions or diagnostic tests comprising said antibodies.

The anti-D antibody compositions according to the invention can be used for the prevention of Rhesus isoimmunization of Rh negative individuals, in particular of hemolytic disease of the newborn (MHNN) or in applications such as Idiopathic Thrombocytopenic Purpura (PTI).

Passive immunotherapy using polyclonal antibodies has been carried out since the 1970s. However, the production of polyclonal immunoglobulins poses several problems:<ul id="ul0001" list-style="none" compact="compact"><li>The immunization of voluntary subjects was interrupted in France in 1997, due to the ethical problems presented by such acts. In France, as in Europe, the number of immunized donors is too low to ensure a sufficient supply of certain antibodies so that it proves necessary to import hyperimmunized plasma from the United States for example.</li></ul>

Thus, this shortage of immunoglobulin does not make it possible to envisage antenatal administration for the prevention of MHNN.

Various studies have resulted in the production of human monoclonal antibodies in order to replace the polyclonal antibodies obtained from the fractionation of plasmas from voluntary donors.

Monoclonal antibodies have several advantages: they can be obtained in large quantities at reasonable prices, each batch of antibodies is homogeneous and the quality of the various batches is reproducible because they are produced by the same cell line which is cryopreserved in nitrogen liquid. Product safety can be assured as to the absence of viral contamination.

Several publications describe the production of cell lines producing anti-Rh D IgG class monoclonal human antibodies, from B cells of immune donors. Boylston et al. 1980; Koskimies 1980; Crawford et al. 1983; Doyle et al. 1985; Goossens et al. 1987; Kumpel et al. 1989 (a) and Mc Cann-Carter et al. 1993 describe the production of lines of B lymphocytes transformed by the EBV virus. Melamed et al. 1985; Thompson et al. 1986 and Mc Cann-Carter et al. 1993 relate to heterohybrids resulting from the fusion of B lymphocytes (transformed by EBV) x murine myeloma. Goossens et al. 1987 concerns heterohybrids resulting from the fusion of B lymphocytes (transformed by EBV) x human myeloma. Bron et al. 1984 and Foung et al. 1987 describe heterohybrids resulting from the fusion of B lymphocytes (transformed by EBV) x human-mouse heteromyeloma and finally, Edelman et al. 1997 relates to insect cells transfected with the gene coding for an anti-Rh (D) using the baculovirus system.

Among the patents and patent applications concerning such monoclonal antibodies and their secretory lines, there may be mentioned:<ul id="ul0002" list-style="none" compact="compact"><li><patcit id="pcit0001" dnum="EP576093A"><text>EP 576093</text></patcit> (AETS (FR), Biotest Pharma GmbH (Germany); Composition for prophylaxis of the haemolytic disease of the new-born understood two human monoclonal antibodies of sub-class IgG1 and IgG3, which are active against the Rhesus D antigen), <patcit id="pcit0002" dnum="RU2094462"><text>RU 2094462</text></patcit>, <patcit id="pcit0003" dnum="WO8502413A"><text>WO 85/02413</text></patcit> (Board of Trustees of the Leland Stanford Jr. University, Human Monoclonal Antibody against Rh (D) Antigen and its Uses), <patcit id="pcit0004" dnum="GB8610106A"><text>GB 86-10106</text></patcit> (Central Blood Laboratories Authority, Production of heterohybridomas for manufacture of human monoclonal antibodies to Rhesus D antigen), <patcit id="pcit0005" dnum="EP0251440A"><text>EP 0 251 440</text></patcit> (Central Blood Laboratories Authority, Human Anti-Rhesus D Producing Heterohybridomas), <patcit id="pcit0006" dnum="WO8902442A"><text>WO 89/02442</text></patcit>, <patcit id="pcit0007" dnum="WO8902600A"><text>WO 89/02600</text></patcit> and <patcit id="pcit0008" dnum="WO89024443A"><text>WO 89/024443</text></patcit> (Central Blood Laboratories Authority, Human Anti-Rh (D) Monoclonal Antibodies), <patcit id="pcit0009" dnum="WO8607740A"><text>WO 8607740</text></patcit> (Institut Pasteur, Protein Performance SA, Paris, FR, Obtaining a recombinant monoclonal antibody from a human anti-rhesus D monoclonal antibody, its production in insect cells and its uses), <patcit id="pcit0010" dnum="JP63050710A"><text>JP 88-50710</text></patcit> (International Reagents Corp., Japan, Reagents for Determination of Blood Group Substance Rh (D) Factor), <patcit id="pcit0011" dnum="JP58248865A"><text>JP 83-248865</text></patcit> (Mitsubishi Chemical Industries Co., Ltd., Japan, Preparation of Monoclonal Antibody to Rh (D) positive Antigen), CA 82-406033 (Queens University at Kingston, Human Monoclonal Antibodies) and <patcit id="pcit0012" dnum="GB8226513A"><text>GB 8226513</text></patcit> (University College London, Human Monoclonal Antibody against Rhesus D Antigen).</li></ul>

While the use of monoclonal antibodies has many advantages over the use of pools of polyclonal antibodies, it may, however, be difficult to obtain an effective monoclonal antibody. In fact, it has been found in the context of the invention that the Fcγ fragment of the immunoglobulin obtained must have very specific properties in order to be able to interact and activate the receptors of the effector cells (macrophage, TH and NK lymphocyte).

The biological activity of certain immunoglobulins G is dependent on the structure of the oligosaccharides present on the molecule, and in particular on its Fc part. IgG molecules of all human and murine subclasses have an N-oligosaccharide attached to the CH domain₂ of each heavy chain (at residue Asn 297 for human IgG). The influence of this glycan residue on the ability of the antibody to interact with effector molecules (Fc receptors and complement) has been demonstrated. The inhibition of glycosylation of a human IgG1, by culture in the presence of Tunicamycin, causes for example a 50-fold decrease in the affinity of this antibody for the FcγRI receptor present on monocytes and macrophages (Leatherbarrow et al, 1985) . Binding to the FcγRIII receptor is also affected by the loss of carbohydrates on IgG, since it has been described that an unglycosylated IgG3 is incapable of inducing ADCC-type lysis via the FcγRIII receptor of NK cells. (Lund et al, 1990).

But, beyond the necessary presence of these glycan residues, it is more precisely the heterogeneity of their structure which can lead to differences in the capacity to engage effector functions. Profiles of galactosylation variable according to individuals (human serum IgG1) were observed. These differences probably reflect disparities in the activity of galactosyltransferases and other enzymes between the cell clones of these individuals (Jefferis et al, 1990). While this normal heterogeneity of post-translational processes generates different glycoforms (even in the case of monoclonal antibodies), it can lead to atypical structures associated with certain medical conditions such as rheumatoid arthritis, Crohn's disease, for which a significant proportion of agalactosylated residues has been demonstrated (Parekh et al, 1985).

The glycosylation profile of the purified molecule is the consequence of multiple effects, certain parameters of which have already been studied. The protein skeleton of IgG and in particular the amino acids in contact with the N-acetylglucosamine (GlcNAc) and galactose residues of the mannose arm α 1-6 (aa 246 and 258 IgG) can explain the existence of preferential structures (galactosylation) as shown by the study carried out on murine and chimeric IgGs of different isotypes (Lund et al, 1993).

The differences observed also highlight specificities linked to the species and cell type used for the production of the molecule. Thus, the conventional structure of N-glycans of human IgG reveals a significant proportion of biantennate types with a GIcNAc residue in bisecting position, structure absent at the level of the antibodies produced by murine cells. Similarly, the sialic acid residues synthesized by the CHO (Chinese Hamster Ovary) line are exclusively of the α 2-3 type whereas they are of the α 2-3 and α 2-6 type with murine and human cells ( Yu Ip et al, 1994). The production of immunoglobulins in expression systems other than those derived from mammals can introduce much more significant modifications, such as the presence of xylose residues produced by insect cells or by plants (Ma et al, 1995).

Other factors such as cell culture conditions (including the composition of the culture medium, cell density, pH, oxygenation) seem to intervene on the activity of the glycosyltransferases of the cell and consequently on the glycan structure of the cell. molecule (Monica et al, 1993; Kumpel et al, 1994 b).

Wright &amp; Morrison, 1997 is a review of the effects of glycosylation on the functions of antibodies. The authors mention the results of Rothman et al, 1989 showing that antibodies produced by clones treated with Castanospermine (Cs), a glycosylation inhibitor leading to obtaining forms of oligosaccharides of oligomannose type, were capable of inducing a good ADCC response by NK cells.

<patcit id="pcit0013" dnum="WO9954342A1"><text>WO99 / ​​54342A1 and Umana et al, 1999</text></patcit> compare the glycosylations and the ADCC responses induced by different antibody compositions (chCE7). A correlation between the content of oligosaccharide structures with an intermediate GlcNAc and the level of ADCC observed is demonstrated. The authors attribute the induction of a good ADCC to the intercalated GlcNAc.

Lifely-1995 compares the glycosylations and the ADCC responses induced by antibody compositions (CAMPATH-1H) produced by different cell lines. The authors note that the compositions with a high proportion of oligosaccharide structures with an intercalated GlcNAc induce good ADCC, and attribute this good ADCC to the intercalated GlcNAc.

However, it has been found in the context of the present invention that a structure of biantennary type, with short chains, low sialylation, terminal mannoses and / or non-intermediate terminal GlcNAc is the common denominator of the glycan structures conferring a strong activity. ADCC to monoclonal antibodies. We have also developed a process for the preparation of such antibodies capable of activating effector cells via FcγRIII, in particular anti-Rh (D) antibodies.

Blood group antigens are classified into several systems depending on the nature of the membrane molecules expressed on the surface of red blood cells. The Rhesus (Rh) system includes 5 molecules or antigens: D, C, c, E and e (ISSITT, 1988). The D antigen is the most important of these molecules because it is the most immunogenic, that is to say that it can induce the production of anti-D antibodies if Rh D positive red blood cells are transfused to Rh negative subjects.

The D antigen is normally expressed in 85% of Caucasoid subjects, these people are called Rh positive; 25% of these subjects are therefore Rh negative, that is to say that their red cells do not present a D antigen. The expression of the D antigens presents certain variants which can be linked either to a low antigenic density, we will then speak weak D antigens, either with a different or partial antigenicity, we will then speak of partial D antigens. The weak D character is characterized by the fact that it is a normal antigen but whose number of sites per red cell is more or less reduced; this character is transmissible according to Mendelian laws. The partial D phenotypes were discovered in Rh D positive subjects who had serum anti-D antibodies; these partial D antigens can therefore be characterized as having only part of the mosaic. Studies carried out with polyclonal and monoclonal antibodies have made it possible to define 7 categories of partial D antigens with the description of at least 8 epitopes constituting the D antigen (LOMAS et al. 1989; TIPETT 1988).

The importance of anti-Rh D antibodies appeared with the discovery of the mechanisms leading to hemolytic disease of the newborn (MHNN). This corresponds to the various pathological conditions observed in certain fetuses or in certain newborns when there is a fetal-maternal incompatibility of blood group which is responsible for the formation of maternal anti-Rh D antibodies capable of crossing the barrier. placental. Indeed, the passage of Rh positive fetal red cells in an Rh negative mother can lead to the formation of anti-D antibodies.

After immunization of the Rh negative mother, anti-D antibodies of the IgG class are capable of crossing the placental barrier and of fixing themselves on Rh positive fetal red cells. This fixation leads to the activation of immunocompetent cells via their surface Fc receptors, thus inducing hemolysis of sensitized fetal red cells. Depending on the intensity of the reaction, several degrees of severity of MHNN can be observed.

A diagnosis of MHNN can be made before and after birth. The prenatal diagnosis is based on the evolution of the level of anti-D antibodies in the mother using several immunohematological techniques. Postpartum diagnosis can be made from a cord blood sample by analyzing the following parameters: determination of the blood groups of the fetus and father; anti-D antibody testing; hemoglobin and bilirubin assays.

Prophylaxis of MHNN is currently systematically performed in all women with Rh negative blood groups who have given birth to an Rh positive child with injections of human anti-D immunoglobulins. The first real immunoprophylaxis trials began in 1964. For prevention to be effective, the immunoglobulins must be injected before immunization, that is to say within 72 hours of childbirth and that the doses of antibodies are sufficient (10 µg of anti -D for 0.5 ml Rh + red cells).

Several monoclonal anti-D antibodies have been the subject of a therapeutic evaluation: BROSSARD / FNTS 1990 (unpublished); THOMSON / IBGRL 1990; KUMPEL / IBGRL 1994; BELKINA / Moscow Institute of Hematology 1996; BIOTEST / LFB 1997 (unpublished). The clinical efficacy of antibodies to induce Rh (D) positive red cell clearance was evaluated in Rh (D) negative volunteers.

A single IgG1 antibody has shown efficacy equivalent to that of anti-D polyclonal immunoglobulins, but only in a few patients (KUMPEL et al, 1995).

The invention proposes to provide monoclonal antibodies responding to the abovementioned problems, that is to say antibodies selected by an ADCC type test specific for and / or antibodies having a glycan structure necessary for obtaining good efficiency.

Description

Thus, the present invention relates to a composition comprising a monoclonal antibody, characterized in that the antibody has on its glycosylation site (Asn 297) of Fcγ glycan structures selected from the forms:<chemistry id="chem0002" num="0002"><img file="EP1518864B1_D0002.tif" /></chemistry>wherein the content of G0 + G1 + G0F + G1F forms is greater than 60% and the content of G0F + G1F forms is less than 50%.

The present description also relates to a process for the preparation of a monoclonal antibody capable of activating the effector cells expressing FcγRIII, characterized in that it comprises the following steps:<ol id="ol0001" compact="compact" ol-style=""><li>a) purification of monoclonal antibodies obtained from different clones originating from cell lines selected from hybridomas, in particular heterohybridomas, and animal or human cell lines transfected using a vector comprising the gene coding for said antibody;</li><li>b) addition of each antibody obtained in step a) in a separate reaction mixture comprising:<ul id="ul0003" list-style="dash" compact="compact"><li>the target cells of said antibodies,</li><li>effector cells comprising cells expressing FcγRIII,</li><li>polyvalent IgG,</li></ul></li><li>c) determination of the percentage of lysis of the target cells and selection of the monoclonal antibodies which activate the effector cells causing significant lysis of the target cells (ADCC activity of FcγRIII type).</li></ol>

The clones can come from heterohybrid cell lines obtained by fusion of human B lymphocytes (coming from immunized subjects) with murine, human or heterohybrid myeloma cells, in particular myeloma K6H6-B5 (ATCC n ° CRL 1823); or also animal or human cell lines transfected using a vector containing the gene coding for a human immunoglobulin of the IgG type, said lines being able to be selected in particular from the CHO-K, CHO-Lec10, CHO Lec-1 lines. , CHO Pro-5, CHO dhfr-, Wil-2, Jurkat, Vero, Molt-4, COS-7, 293-HEK, YB2 / 0, BHK, K6H6, NSO, SP2 / 0-Ag 14 and P3X63Ag8.653 .

Multipurpose IgGs are used to inhibit the lysis mechanism of effector cells via FcγRIII.

In this process, the antibodies are selected having an ADCC level of FcγRIII type greater than 60%, 70%, 80% or preferably greater than 90%.

Target cells can be red cells treated with papain. In this case, we deposit by well:<ul id="ul0004" list-style="dash" compact="compact"><li>100 µl of monoclonal antibodies purified to around 200 ng / ml,</li><li>25 µl papain red cells, approximately 1x10⁶ cells,</li><li>25 µl of effector cells, approximately 2x10⁶ cells,</li><li>and 50µl of polyvalent IgG, in particular TEGELINE^™ (LFB, France), at a concentration between 1 and 20 mg / ml.</li></ul>

We can thus compare the amount of lysis of target cells to two positive controls consisting of a chemical compound such as NH₄Cl and an active reference antibody <i>in vivo</i> and a negative control consisting of an inactive antibody <i>in vivo.</i>

Polyclonal antibodies of commercial origin can also be used as positive controls and a monoclonal antibody incapable of inducing clearance <i>in vivo</i> as a negative control.

Advantageously, this method makes it possible to prepare monoclonal anti-Rh (D) antibodies, as indicated above. Rhesus D red cells are then used as target cells.

This description therefore describes the development of a biological activity test <i>in vitro</i> in which the activities measured are correlated with biological activity <i>in vivo</i> monoclonal or polyclonal antibodies already evaluated from the clinical point of view as to their potential to induce clearance of Rh (D) positive red cells in Rh (D) negative volunteers. This test makes it possible to evaluate the lytic activity dependent on the antibody = ADCC (Antibody Dependent Cellular Cytotoxicity) essentially induced by Fcγ type III receptors (CD 16), Fcγ type I receptors (CD64) being saturated by l addition of human IgG immunoglobulins (in the form of polyvalent therapeutic IgGs). The FcγRIII specificity of this ADCC test was confirmed by inhibition in the presence of anti-FcγRIII monoclonal antibody (see<figref idref="f0006">figure 6</figref>). Mononuclear cells from healthy subjects are used as effector cells in an effector / target ratio (W / C) close to physiological conditions<i>in vivo.</i> Under these conditions the lytic activities of polyclonal immunoglobulins and ineffective anti-D monoclonal antibodies <i>in vivo</i> (antibody DF5 Goossens et al, 1987 and antibodies AD1 + AD3, <patcit id="pcit0014" dnum="FR9207893"><text>FR 9207893</text></patcit> LFB / Biotest and FOG-1, <patcit id="pcit0015" dnum="GB2189506A"><text>GB 2189506</text></patcit>) are respectively strong and weak.

The selection of the antibodies described in the present description was therefore carried out by evaluation of their biological activity in this ADCC type test (see example 1).

The antibodies capable of being obtained from the method described above have ADCC levels of FcγRIII type greater than 60%, 70%, 80% or preferably greater than 90% relative to the reference polyclonal. The monoclonal antibodies, directed against a given antigen, activate the effector cells expressing FcγRIII causing lysis greater than 60%, 70%, 80%, preferably greater than 90% of the lysis caused by polyclonal antibodies directed against said antigen. Advantageously, said monoclonal antibodies are directed against rhesus D.

They can preferably be produced by clones derived from the Vero (ATCC n ° CCL 81), or YB2 / 0 (ATCC n ° CRL 1662) lines and can belong to the IgG1 or IgG3 class.

The present description also relates to antibodies having a particular glycan structure conferring an effector activity dependent on FcγRIII.

Such antibodies can be obtained from a process explained above and have on their glycosylation site (Asn 297) of Fcγ glycan structures of biantennary type, with short chains and low sialylation. Preferably, their glycan structure has terminal mannoses and / or non-intermediate terminal GlcNAc.

Such antibodies are more particularly selected from the forms:<chemistry id="chem0003" num="0003"><img file="EP1518864B1_D0003.tif" /></chemistry>

Thus, the present description relates to a monoclonal antibody characterized in that it has on its glycosylation site (Asn 297) of Fcγ of biantennary type glycan structures, with short chains, weak sialylation, mannoses and GlcNAc from the point d '' non-intermediate terminal attachment. Said antibodies, directed against a given antigen, activate the effector cells expressing FcγRIII causing lysis greater than 60%, 70%, 80%, preferably greater than 90% of the lysis caused by polyclonal antibodies directed against said antigen.

The invention relates to a composition comprising a monoclonal antibody, characterized in that the antibody has, on its glycosylation site (Asn 297) of Fcγ, glycan structures selected from the forms:<chemistry id="chem0004" num="0004"><img file="EP1518864B1_D0004.tif" /></chemistry>wherein the content of G0 + G1 + G0F + G1F forms is greater than 60%, preferably greater than 80%, and the content of G0F + G1F forms is less than 50%, preferably less than 30%.

More particularly, in the compositions according to the invention comprising the antibodies as defined above, the sialic acid content can be less than 25%, 20%, 15%, or 10%, preferably 5%, 4% 3 % or 2%.

Likewise, in the compositions according to the invention comprising the antibodies as defined above, the fucose content can be less than 65%, 60%, 50%, 40%, or 30%. Preferably, the fucose content is between 20% and 45% or even between 25% and 40%.<chemistry id="chem0005" num="0005"><img file="EP1518864B1_D0005.tif" /></chemistry>

An alternative to specifically target FcγRIII consists in the preparation of “high mannose” type antibodies.

In another aspect, the present description relates to a cell producing an antibody mentioned above. It may be a hybridoma, in particular a heterohybridoma obtained with the fusion partner K6H6-B5 (ATCC n ° CRL 1823); or of an animal or human cell transfected using a vector comprising the gene coding for the said antibody, in particular a cell derived from the Vero lines (ATCC n ° CCL 81), or YB2 / 0 (ATCC n ° CRL 1662 ). These cells correspond to the cell lines selected by the method according to the invention, said cells producing antibodies having the characteristics mentioned above.

A preferred antibody in the context of the invention shows significant biological activity (greater than or equal to that of the reference anti-Rh (D) polyclonal antibody) in the ADCC test using FcγRIII positive effector cells.

Its ability to activate FcγRIII receptors (after fixation) is confirmed on models <i>in vitro</i> which demonstrate the modification of the intracellular calcium flow, the phosphorylation of molecules for transduction of the activation signal or the release of chemical mediators.

These properties are associated with a particular structure of the oligosaccharides of the N-glycosylation site of the Fc part of the antibody: presence of short chains, weakly galactosylated, slightly sialylated, possessing terminal mannoses and / or terminal GIcNAc, for example not intercalated.

This antibody has therapeutic applications: prevention of MHNN, treatment of ITP in Rh (D) positive individuals, and any other application concerned with the use of anti-D polyclonal immunoglobulins.

A preferred antibody in the context of the invention may also have a specificity other than anti-Rh (D) (anti-cancer cell for example). It can have the properties described above (functional activity dependent on a binding / activation mechanism at the level of FcγRIII receptors, particular structure of oligosaccharides) and can be used in the immunotherapy of cancers or any other pathology for which a curative or preventive treatment using a monoclonal antibody, the mechanism of action of which corresponds to functional activity via the FcγRIII receptor.

Another aspect relates to a pharmaceutical composition comprising an antibody composition according to the invention and to the use of said composition for the manufacture of a medicament.

Preferably, the invention relates to the use of a composition of an anti-Rh (D) antibody described above for the manufacture of a medicament intended for the prevention of Rhesus alloimmunization of Rh individuals negative. The mode of action of anti-D immunoglobulins<i>in vivo</i> is a specific binding of antibodies to the D antigen of Rh (D) positive red blood cells, followed by elimination of these red blood cells essentially from the spleen. This clearance is associated with a dynamic mechanism of suppression of the primary immune response in individuals and therefore prevents immunization.

Thus, one can use an antibody composition of the invention prophylactically for the prevention of alloimmunization of Rhesus negative women, immediately after the birth of a Rhesus positive child, and to prevent, during subsequent pregnancies hemolytic disease of the newborn (MHNN); during abortions, ectopic pregnancies in a Rhesus D incompatibility situation or during transplacental hemorrhages resulting from amniocentesis, chorionic biopsies, or traumatic obstetric manipulations in a Rhesus D incompatibility situation

In addition, an antibody composition of the invention can be used in the case of Rh transfusions incompatible with blood or labile blood derivatives.

The invention also relates to the use of an antibody composition of the invention for the manufacture of a medicament for therapeutic use in Idiopathic Thrombocytopenic Purpura (ITP).

The antibody compositions of the invention are also useful for the manufacture of a medicament for the treatment of cancers by immunotherapy or for the treatment of infections caused by viral or bacterial pathogens.

An additional aspect of the present description relates to the use of said antibodies, in particular for diagnosis. The present description therefore relates to a kit comprising an antibody described above.

For the rest of the description, reference is made to the legends of the figures presented below.

Legends

<ul id="ul0005" list-style="none" compact="compact"><li><figref idref="f0001"><b>Figure 1</b></figref><b>: ADCC evaluation of F60 and T125 YB2 / 0 (R270)</b>This figure represents the percentage of lysis obtained as a function of the antibody concentration in the presence of 100 and 500 μg / well of TEGELINE^™ (LFB, France). A high lysis percentage is obtained for the antibody compositions according to the invention F60 and T125.</li><li><figref idref="f0002"><b>Figure 2</b></figref><b>: attachment of anti-D to the receptor (FcγRIII)</b>A high binding index is obtained for the antibody compositions according to the invention F60 and T125.</li><li><figref idref="f0003"><b>Figure 3</b></figref><b>: construction of the expression vector T125-H26 for the expression of the H chain of T125.</b></li><li><figref idref="f0004"><b>Figure 4</b></figref><b>: construction of the expression vector T125-K47 for the expression of the L chain of T125</b></li><li><figref idref="f0005"><b>Figure 5</b></figref><b>: construction of the expression vector T125-IG24 for the expression of the whole antibody T125.</b></li><li><figref idref="f0006"><b>Figure 6</b></figref><b>: ADCC inhibition in the presence of anti Fc RIII (CD16)</b>The ADCC test is established according to the procedure described in §3.3 in the presence of the commercial anti-CD16 3G8 (TEBU) whose action is to block the FcRIII receptors present on the effector cells. The final concentration of 3G8 is 5 µg / well (25 µg / ml). A witness is made in parallel in the absence of 3G8. The three antibodies tested are Poly-D WinRho, antibody F60 (Pf 155 99/47) obtained according to the method described in Example I and R297 (Pf 210 01/76) obtained according to the method described in Example II.<b>Results:</b> inhibition is observed in the presence of 3G8, which demonstrates that the ADCC induced by the three antibodies tested is mainly dependent on FcRIII. Slightly stronger inhibition is observed in the presence of Poly-D WinRho (83% compared to 68% and 61% inhibition for F60 and R297 respectively). This difference may be due to the presence in Poly-D of non-anti-D human IgG which will inhibit type 1 receptors (FCRI or CD64) and therefore act in synergy with anti-CD16.</li><li><figref idref="f0007"><b>Figure 7</b></figref><b>: characterization of Anti-D glycans by Mass Spectrometry (MS).</b></li><li><figref idref="f0008"><b>Figure 8</b></figref><b>: Comparison of the MS spectra of R 290 and DF5.</b></li><li><figref idref="f0009"><b>Figure 9</b></figref><b>: study of the glycosylation of Anti-D D31DMM by MS.</b></li></ul>

EXAMPLE 1 ESTABLISHMENT OF A HETEROHYBRID CELL LINE PRODUCING ANTI-RH ANTIBODIES (D)

1- <u>Obtaining lymphoblastoid and heterohybrid clones:</u>

<i>1.1- Source of lymphocytes:</i>

The donor of B lymphocytes is selected from the donors of anti-Rh (D) in plasmapheresis, on the activity of its serum anti-Rh (D) antibodies in the ADCC activity test described in §33. After a donation of whole blood in 1998, the “buffy coat” fraction (leukocyte concentrate) is recovered.

<i>1.2-Immortalization of donor B cells</i>

The peripheral blood mononuclear cells are separated from the other elements by centrifugation on Ficoll Plus (Pharmacia). They are then diluted to 10⁶ cells / ml in IMDM containing 20% ​​(v / v) of fetal calf serum (SVF), to which are added 20% of culture supernatant of line B95-8 (ATCC-CRL1612), 0.1 μg / ml of cyclosporine A (Sandoz), 50 µg / ml gentamycin sulfate (Life Technologies), and distributed in 24-well plates (P24 Greiner) or in 96-well round-bottom plates. They are then placed in an incubator at 37 ° C, 7% CO2. After 3 weeks, the presence of anti-Rh (D) antibodies is sought by ADCC<ul id="ul0006" list-style="none" compact="compact"><li>∧ Each of the 16 microwells from a P24 positive plate well is transferred to a new P24 well. This enrichment is repeated after 10 to 15 days of culture and each microwell is amplified in P96 then in P24.</li><li>∧ The P96 positive wells are taken up and amplified in P24 with flat bottom (Nunc). After a few days of culture, the presence of anti-Rh (D) antibodies is sought by ADCC.</li></ul>

<i>1.3- Enrichment with immune rosettes (RI):</i>

Cells from one or more P24 wells are enriched in specific cells by formation and separation of rosettes with Rh (D) positive red cells from papain: 1 volume of red cells washed in 0.9% NaCl is incubated for 10 minutes at 37 ° C with 1 volume of papain solution (Merck) at 1/1000^th (m / v), then washed 3 times in 0.9% NaCl. The cells are then washed once in Hanks solution, suspended in SVF and mixed with papain red cells in the ratio 1 cell for 33 red cells. The mixture is placed in a conical bottom centrifuge tube, centrifuged for 5 minutes at 80 g and incubated for one hour in melting ice. The mixture is then gently stirred and the Ficoll deposited at the bottom of the tube for separation 20 minutes at 900g. The pellet containing the rosettes is hemolyzed in NH solution₄Cl for 5 minutes and the cells re-cultured in P24 containing irradiated human mononuclear cells. After approximately 1 week, the supernatants are evaluated in CELA (paragraph 3.2) and ADCC tests for the presence of anti-Rh (D) antibodies having good activity. A new enrichment cycle is carried out if the percentage of rosette-forming cells increases significantly compared to the previous cycle.

<i>1.4-Cloning of lymphoblastoid cells:</i>

The cells enriched by RI are distributed at 5 and 0.5 cells per well in 96-well round bottom plates containing irradiated human mononuclear cells.

After approximately 4 weeks of culture, the supernatants of the wells containing cell clusters are evaluated by ADCC test.

<i>1.5- Heterofusion:</i>

The cloning wells of the cells transformed by EBV having an interesting ADCC activity are amplified in culture and then fused with the heteromyeloma K6H6-B5 (ATCC CRL-1823) according to the standard PEG technique. After fusion, the cells are distributed at a rate of 2 10⁴ cells / wells in flat-bottomed P96 containing murine intraperitoneal macrophages and in a selective medium containing aminopterin and ouabain (Sigma).

After 3 to 4 weeks of culture, the supernatants of the wells containing cell clusters are evaluated by ADCC test.

<i>1.6- cloning of heterohybridomas:</i>

Cloning by limiting dilution is carried out at 4, 2 and 1 cells / well in P96 with flat bottom. After 2 weeks, the microscopic appearance of the wells is examined to identify the unique clones, then the medium is renewed. After approximately 2 weeks, the supernatants from the wells containing cell clusters are evaluated by ADCC test.

2- <u>History of the clones retained:</u>

<i>2.1- clone producer of an IgG1</i>

The transformation by EBV of the cells of the donor d13 allowed the selection of a well, designated T125 2A2 on which were carried out successively: 2 enrichments, 3 cycles of RI, and cloning at 5 cells / well to give 2 clones:<ol id="ol0002" compact="compact" ol-style=""><li>1) T125 2A2 (5/1) A2 from which the DNA was extracted for the preparation of the recombinant vector;</li><li>2) T125 (5/1) A2 which has been fused with K6H6-B5 to give F60 2F6 then after 5 cloning F60 2F6 (5) 4C4, clone retained for the constitution of a cell stock prior to the preparation of banks.</li></ol>

It is an IgG1 having a Kappa light chain.<chemistry id="chem0006" num="0006"><img file="EP1518864B1_D0006.tif" /></chemistry>

<i>2.2- clone producer of an IgG3</i>

According to the same method as that used for the preparation of the antibody of isotype IgG1, a line producing an IgG3 was prepared. The original cells come from a whole blood donation, from another designated donor, whose “buffy coat” fraction (leukocyte concentrate) has been recovered.

It is an IgG3 having a Kappa light chain.<chemistry id="chem0007" num="0007"><img file="EP1518864B1_D0007.tif" /></chemistry>

3- <u>Methods of evaluation of anti-Rh (D) antibodies:</u>

After purification by affinity chromatography on protein A Sepharose (Pharmacia) and dialysis in 25 mM Tris buffer, 150 mM NaCl, pH 7.4, the concentration of the T125 antibody is determined by ELISA technique. The biological activity in vitro is then measured by the ADCC technique.

<i>3.1- Determination of the IgG level and isotypes by ELISA technique:</i>

•

Total IgG

Coating: anti-IgG (Calbiochem) at 2µg / ml in 0.05M carbonate buffer pH 9.5, 1 night at 4 ° C. Saturation: dilution buffer (PBS + 1% BSA + 0.05% Tween 20, pH 7.2) 1 h at room temperature. Washing (to be renewed at each step): H2O + 150mM NaCl + 0.05% Tween 20. Dilution of the samples in dilution buffer to approximately 100ng / ml and of the control range constituted from polyvalent human IgG LFB prediluted to 100ng / ml. 2h incubation at room temperature. Conjugate: anti-IgG (Pasteur Diagnosis) diluted to 1/5000, 2 hours at room temperature. Substrate: OPD at 0.5 mg / ml (sigma) in citrate phosphate buffer, Na perborate (Sigma), 10 minutes in the dark. Stop the reaction with 1N HCL, and read at 492 nm.

•

Dosage chain Kappa.

Coating: anti-Kappa (Caltag Lab) at 5µg / ml in 0.05M carbonate buffer pH 9.5.1 overnight at 4 ° C. Saturation: dilution buffer (PBS + 1% BSA + 0.05% Tween 20, pH 7.2) 1 h at room temperature. Washing (to be renewed at each step): H2O + 150mM NaCl + 0.05% Tween 20. Dilution of the samples in dilution buffer to approximately 100ng / ml and of the control range constituted from the monoclonal antibody AD3T1 LFB (Kappa / gamma 3) prediluted to 100ng / ml. Incubation 2 h at room temperature Conjugate: biotinylated anti-kappa (Pierce) diluted 1/1000 in the presence of Streptavidin-peroxidase (Pierce) diluted 1/1500, 2 hours at room temperature. Substrate: 0.5 mg / ml OPD (sigma) in citrate phosphate buffer, Na perborate (Sigma), 10 minutes in the dark. Stop the reaction with 1N HCL, and read at 492 nm.

<i>3.2- Specific anti-D assay by CELA technique (Cellular Enzyme Linked Assay):</i>

This method is used for the specific assay of anti-D antibodies, especially when it is a question of culture supernatant at culture stages where other non-anti-D immunoglobulins are present in the solution (early stages after EBV transformation). .

<u>Principle</u> : the anti-D antibody is incubated with Rhesus positive red cells and then revealed by a human anti-Ig labeled with alkaline phosphatase.

100µl 10% Rh + red cells diluted in 1% Liss-BSA dilution buffer. Dilution of the samples in dilution buffer to approximately 500 ng / ml and of the control range constituted from a purified monoclonal anti-D human IgG (DF5, LFB) prediluted to 500 ng / ml. 45 min incubation at room temperature. Washing (to be renewed at each stage): H2O + 150mM NaCl. Conjugate: anti-IgG alkaline phosphatase (Jackson) diluted to 1/4000 in PBS + 1% BSA, 1:30 at room temperature. Substrate: PNPP at 1 mg / ml (sigma) in 1 M diethanolamine, 0.5 mM MgCl 2; PH 9.8. Stop the reaction with 1N NaOH and read at 405 nm.

<i>3.3- ADCC technique</i>

The ADCC (Antibody-Dependent Cellular Cytotoxicity) technique assesses the ability of antibodies (anti-D) to induce lysis of Rh positive red cells, in the presence of effector cells (mononuclear cells or lymphocytes).

Briefly, the red cells of a Rh positive globular concentrate are treated with papain (1 mg / ml, 10 min at 37 ° C.) and then washed in 0.9% NaCL. The effector cells are isolated from a pool of at least 3 buffy coats, by centrifugation on Ficoll (Pharmacia), followed by an adhesion step in the presence of 25% of FCS, so as to obtain a ratio lymphocytes / monocytes of the order of 9. In a microtiter plate (96 wells), each well is deposited: 100 μl of purified anti-D antibody at 200 ng / ml, 25 μl of Rh + papain red blood cells (i.e. 1x10⁶), 25µl of effector cells (i.e. 2x10⁶) and 50 μl of polyvalent IgG (Tegeline LFB for example) at the usual concentrations of 10 and 2 mg / ml. Dilutions are made in IMDM 0.25% SVF. After incubation for 1 night at 37 ° C., the plates are centrifuged, then the hemoglobin released in the supernatant is measured in the presence of a substrate specific for peroxidase activity (2.7 diaminofluorene, DAF). The results are expressed as a percentage of lysis, 100% corresponding to the total lysis of the red cells in NH4Cl (control 100%), and 0% to the reaction mixture without antibody (control 0%).

The specific lysis is calculated as a percentage according to the following formula: <maths id="math0001" num=""><math display="block"><mfrac><mrow><mfenced separators=""><mi>DO sample</mi><mo>-</mo><mi>DO witness</mi><mspace width="1em" /><mn>0</mn><mo>%</mo></mfenced><mspace width="1em" /><mi>x</mi><mspace width="1em" /><mn>100</mn></mrow><mrow><mi>DO witness</mi><mspace width="1em" /><mn>100</mn><mo>%</mo><mo>-</mo><mi>DO witness</mi><mspace width="1em" /><mn>0</mn><mo>%</mo></mrow></mfrac><mo>=</mo><mo>%</mo><mi>ADCC</mi></math><img file="EP1518864B1_D0008.tif" /></maths>

The results presented to the <figref idref="f0001">figure 1</figref> show the activity of the antibody produced by the F60 heterohybrid compared to that of the reference antibodies:<ul id="ul0007" list-style="dash" compact="compact"><li>polyclonal anti-Rh (D) antibodies POLY-D LFB 51 and WinRhO W03 (Cangene) = positive controls</li><li>the monoclonal antibody DF5 (inactive <i>in vivo</i> on the clearance of Rh (D) positive red cells (BROSSARD / FNTS, 1990, unpublished)) = negative control</li><li>purified IgG 1 (separated from IgG3) from the polyclonal WinRhO W03.</li></ul>

Two concentrations of human IgG (Tegeline LFB) are used to show that the inhibition of activity of the negative control is linked to the binding of competitive IgGs on type I Fcγ receptors.

<i>3.4- FcγRIII fixation technique (CD16):</i>

This test makes it possible to assess the binding of anti-Rh (D) antibodies of the IgG1 isotype to FcγRIII and in particular to differentiate IgG3 antibodies. In view of the low affinity of this receptor for monomeric IgGs, the prior fixing of the antibodies to the D antigen is necessary.

Principle: on Rh + red blood cell membranes coated with a microtiter plate, the antibody to be tested (anti-D) is added, then transfected Jurkat cells expressing the FcγRIII receptor on their surface. After centrifugation, the interaction “Rh + membrane / anti-D / Jurkat CD16” is visualized by a homogeneous spreading of Jurkat CD16 in the well. On the contrary, the cells gather in the center of the well in the absence of interaction. The intensity of the reaction is expressed in numbers of +.

<u>Method</u> : 1) Incubation for 1 hour at 37 ° C. of the anti-D antibody (50 μl to 1 μg / ml in IMDM) on Capture R plate (Immunochim), then washing with water + 0.9% NaCl. Addition of Jurkat CD16 (2.10⁶ cells / ml) in IMDM + 10% SVF. Incubation 20 min at 37 ° C then centrifugation and evaluation of cell adhesion (against a control range). 2) Revelation of the anti-D fixed on the Capture R plates by ELISA type technique using human anti-IgG-Peroxidase at 1/5000 (Sanofi Diagnostics Pasteur) after having lysed the Jurkat CD16 cells with Tris -HCl 0.2M, Urea 6M, pH 5.3-5.5. OPD revelation then reading of the optical density (OD) at 492 nm. Expression of results: an arbitrary value of 0 to 3 is assigned according to the fixation and spreading of the Jurkat CD16 cells. These values ​​are assigned to each DO interval defined (from 0.1 to 0.1). We trace:<ul id="ul0008" list-style="none" compact="compact"><li>* or a curve: adhesion of Jurkat cells (Y) as a function of the amount of anti-D fixed on the red blood cell membranes (X).</li><li>* or a histogram of the “fixation indices” corresponding for each antibody, to the sum of each fixation value of the Jurkat cells (0 to 3) affected by OD interval (on a portion common to all of the antibodies tested).</li></ul>

An example of a histogram is presented in <figref idref="f0002">figure 2</figref>.

The anti-Rh (D) antibodies of isotype IgG1 (F60 and T125 YB2 / 0) show a binding index close to that of polyclonal IgG1 (WinRho), while the negative control antibodies DF5 and AD1 do not bind. Likewise, the antibody of isotype IgG3 (F41) has a good fixation index, slightly lower than that of IgG3 purified from the polyclonal Winrho and higher than that of the antibody AD3 (another IgG3 tested and ineffective in clinical trial, mixed with AD1 (Biotest / LFB, 1997, unpublished).

<u>Example 2:</u>

<u>PRODUCTION OF A RECOMBINANT ANTI-D ANTIBODY</u>

1- Isolation and amplification of cDNAs coding for the heavy and light chains of Ac

<i>1.1- RNA extraction and cDNA synthesis</i>

The total RNAs were extracted from a clone producing anti-D Ab (IgG G1 / Kappa) obtained by EBV transformation: T125 A2 (5/1) A2 (see paragraph 2, example 1) The corresponding cDNAs were synthesized by reverse transcription of total RNAs using oligo dT primers.

<i>1.2- Amplification of the variable region of the heavy chain of T125-A2: VH sequence</i>/<i>T125-A2</i>

The VH / T125-A2 sequence is obtained by amplification of the T125-A2 cDNAs using the following primers:<ul id="ul0009" list-style="dash" compact="compact"><li>primer A2VH5 located 5 ′ of the leader region of the V12 gene of T125-A2, introduces a consensus leader sequence (in bold) deduced from already published leader sequences and associated with VH genes belonging to the same VH3-30 family as the gene T125-A2 VH; this sequence also includes an Eco RI restriction site (in italics) and a Kozak sequence (underlined):<ul id="ul0010" list-style="none" compact="compact"><li>A2VH5 (SEQ ID N ° 1):<ul id="ul0011" list-style="none" compact="compact"><li>5'- CTCTCCG<i>AATT</i>VS<u>GCCGCCACC</u><b>ATGGAGTTTGGGCTGAGCTGGGT</b> -3'</li></ul></li></ul><ul id="ul0012" list-style="dash" compact="compact"><li>antisense primer GSP2ANP located 5 'from the constant region (CH) of T125-A2:<ul id="ul0013" list-style="none" compact="compact"><li>GSP2ANP (SEQ ID N ° 2): 5'- GGAAGTAGTCCTTGACCAGGCAG -3 '.</li></ul></li></ul></li></ul>

<i>1.3- Amplification of the constant region of T125-A2: CH sequence</i>/<i>T125-A2</i>

The CH / T125-A2 sequence is obtained by amplification of the T125-A2 cDNAs using the following primers:<ul id="ul0014" list-style="dash" compact="compact"><li>primer G1 located 5 'to the CH region of T125-A2:<ul id="ul0015" list-style="none" compact="compact"><li>G1 (SEQ ID N ° 3): 5'- <u>VS</u>CCTCCACCAAGGGCCCATCGGTC -3 ' The first base G of the sequence CH is here replaced by a C (underlined) in order to recreate after cloning an Eco RI site (see paragraph 2.1.1)</li></ul></li><li>antisense primer H3'Xba located 3 'to the CH of T125-A2, introduces an Xba I site (underlined) 3' to the amplified sequence:<ul id="ul0016" list-style="none" compact="compact"><li>H3'Xba (SEQ ID N ° 4):<ul id="ul0017" list-style="none" compact="compact"><li>5'- GAGAGG<u>TCTAGA</u>CTATTTACCCGGAGACAGGGAGAG -3 '</li></ul></li></ul></li></ul>

<i>1.4- Amplification of the Kappa light chain: K sequence</i>/<i>T125-A2</i>

The entire Kappa chain of T125-A2 (sequence K / T125-A2) is amplified from the cDNAs of T125-A2 using the following primers:<ul id="ul0018" list-style="dash" compact="compact"><li>primer A2VK3 located 5 ′ of the leader region of the VK gene of T125-A2, introduces a consensus sequence (in bold) deduced from the sequence of several leader regions of VK VH genes belonging to the same VK1 subgroup as the VK gene of T125-A2; this sequence also includes an Eco RI restriction site (in italics) and a Kozak sequence (underlined):<ul id="ul0019" list-style="none" compact="compact"><li>A2VK3 (SEQ ID N ° 5):<ul id="ul0020" list-style="none" compact="compact"><li>5'- CCTACC<i>GAATTC</i><u>GCCGCC</u>ACC<b>ATGGACATGAGGGTCCCCGCTCA</b> -3'</li></ul></li></ul></li><li>antisense primer KSE1 located 3 'to Kappa, introduces an Eco RI site (underlined):<ul id="ul0021" list-style="none" compact="compact"><li>KSE1 (SEQ ID N ° 6):<ul id="ul0022" list-style="none" compact="compact"><li>5'- GGTGGT<u>GAATTCC</u>TAACACTCTCCCCTGTTGAAGCTCTT -3 '.</li></ul></li></ul></li></ul>

The <figref idref="f0001">Fig. 1</figref> diagrams the strategies for amplifying the heavy and light chains of T125-A2.

2- Construction of expression vectors

<i>2.1- T125-A2 heavy chain expression vector: T125-H26</i>

The construction of T125-H26 is summarized <figref idref="f0002">Fig. 2</figref>. It is carried out in two stages: firstly, construction of the intermediate vector V51-CH / T125-A2 by insertion of the constant region of T125-A2 in the expression vector V51 derived from pCI-neo (<figref idref="f0003">Fig. 3</figref>) then cloning of the variable region in V51-CH / T125-A2.

2.1.1 cloning of the constant region of T125-A2

The amplified CH / T125-A2 sequence is inserted after phosphorylation at the Eco RI site of the vector V51 (<figref idref="f0003">Fig. 3</figref>). The ligation is carried out after treatment with Klenow polymerase of the Eco RI cohesive ends of V51 in order to make them "blunt end".

The primer G1 used for the amplification of CH / T125-A2 makes it possible to recreate, after its insertion into V51, an Eco RI site 5 ′ of CH / T125-A2.

2.1.2 cloning of variable region of T125-A2

The VH / T125-A2 sequence obtained by amplification is digested with Eco RI and Apa I and then inserted at the Eco RI and Apa I sites of the vector V51-G1 / T125-A2.

<i>2.2- Light chain vector of T125-A2: T125-K47</i>

The construction of T125-K47 is presented <figref idref="f0004">Fig. 4</figref>. The K / T125-A2 sequence obtained by PCR is digested with Eco RI and inserted at the Eco RI site of the expression vector V47 derived from pCI-neo (<figref idref="f0005">Fig. 5</figref>).

<i>2.3- Heavy and light chain vector of T125-A2: T125-IG24</i>

The construction of T125-IG24 is schematized <figref idref="f0006">Fig. 6</figref>. This vector which contains the two transcription units of the heavy and kappa chains of T125-A2 is obtained by insertion of the Sal I - Xho I fragment of T125-K47 containing the transcription unit of K / T125-A2 at the Xho sites I and SalI of T125-H26.

Thus the heavy and light chains of T125-A2 are expressed under the dependence of the CMV promoter; other promoters can be used: RSV, IgG heavy chain promoter, LTR MMLV, HIV, β actin, etc.

<i>2.4- Leader vector specific heavy and light chains of T125 A2: T125-LS4</i>

A second expression vector of T125-A2 is also constructed in which the consensus leader sequence of the Kappa chain is replaced by the real sequence of the leader region of T125-A2 previously determined by sequencing of “PCR 5'-RACE products. ”(Rapid Amplification of cDNA 5 'Ends).

The construction of this vector T125-LS4 is described <figref idref="f0007">Fig. 7</figref>. It is carried out in two stages: firstly the construction of a new expression vector of the Kappa chain of T125-A2, T125-KLS18, then the assembly of the final expression vector, T125-LS4, containing the two modified light chain and heavy chain transcription units.

2.4.1 construction of the vector T125-KLS18

The 5 ′ part of the Kappa consensus leader sequence of the vector T125-K47 is replaced by the specific leader sequence of T125 (KLS / T125-A2) during a step of amplification of the sequence K / T125-A2 carried out at using the following primers:<ul id="ul0023" list-style="dash" compact="compact"><li>primer A2VK9, modifies the 5 'part of the leader region (in bold) and introduces an Eco RI site (underlined) as well as a Kozak sequence (in italics):<img file="EP1518864B1_D0009.tif" /></li><li>primer KSE1 (described in paragraph 1.4)</li></ul>

The vector T125-KLS18 is then obtained by replacing the Eco RI fragment of T125-K47 containing the original K / T125-A2 sequence with the new KLS / T125-A2 sequence digested with Eco RI.

2.4.2 construction of the final vector T125-LS4

The Sal I - Xho I fragment of T125-KLS18 containing the modified sequence KLS / T125-A2 is inserted into T125-H26 at the Xho I and Sal I sites.

3- Production of anti-D Ab in the YB2 / 0 line

<i>3.1- Without gene amplification</i>

The two expression vectors T125-IG24 and T125-LS4 were used for the transfection of cells of the line YB2 / 0 (rat myeloma, line ATCC No. 1662).

After transfection by electroporation and selection of the transformants in the presence of G418 (neo selection) several clones were isolated. The production of recombinant anti-D Ab is approximately 0.2 µg / 10⁶ cells / 24h (<i>value obtained for R270 clone 3B2</i>). The ADCC activity of this recombinant Ab is greater than or equal to that of the poly-D controls (<figref idref="f0001"><b>figure 1</b></figref>). The Acs produced using the two expression vectors are not significantly different in terms of production level or ADCC activity.

<i>3.2- With gene amplification</i>

The gene amplification system implemented is based on the selection of methotrexate resistant transformants (MTX). It requires the prior introduction of a transcription unit coding for the enzyme DHFR (dihydrofolate reductase) into the expression vector of the recombinant Ab (SHITARI et al, 1994)

3.2.1 construction of the expression vector T125-dhfr 13

The diagram presented <figref idref="f0008">Fig. 8</figref> describes the construction of the T125-A2 expression vector containing the murine dhfr gene.

A first vector (V64) was constructed from a vector derived from pCI-neo, V43 (<figref idref="f0009">Fig. 9</figref>), by replacing, 3 'of the SV40 promoter and 5' of a synthetic polyadenylation sequence, the neo gene (Hind III-Csp 45 I fragment) with the cDNA of the murine dhfr gene (obtained by amplification from the plasmid pMT2). This vector is then modified so as to create a Cla I site 5 ′ from the dhfr transcription unit. The Cla I fragment containing the dhfr transcription unit is then inserted at the Cla I site of T125-LS4.

3.2.2 selection in the presence of MTX

◆ 1

time

strategy:

The YB2 / 0 cells transfected by electroporation with the vector T125-dhfr13 are selected in the presence of G418. The transformants producing recombinant Ab are then subjected to a selection in the presence of increasing doses of MTX (from 25 nM to 25 μM). The evolution of the production of recombinant Ab, witnessing the process of gene amplification, is followed during the selection steps in MTX. The transformants resistant to MTX are then cloned by limiting dilution. The level and the stability of the production of recombinant Ab is evaluated for each clone obtained. Anti-D antibody productivity after gene amplification is approximately 13 (+/- 7) µg / 10⁶cells / 24 h.

◆ 2

th

strategy:

The YB2 / 0 cells transfected by electroporation with vector T125-dhfr13 are selected in the presence of G418. The best transformants producing recombinant Ab are cloned by limiting dilution before selection in the presence of increasing doses of MTX. The evolution of the production of each clone, a witness to the gene amplification process, is followed during the selection stages in MTX. The level and the stability of the production of recombinant Ab is evaluated for each MTX-resistant clone obtained.

<i>4- Evaluation of the activity of the T125 antibody expressed in YB2</i>/<i>0</i>

After purification by affinity chromatography on protein A Sepharose (Pharmacia) and dialysis in 25 mM Tris buffer, 150 mM NaCl, pH 7.4, the concentration of the T125 antibody is determined by ELISA technique. The biological activity in vitro is then measured by the ADCC test described above. The results are presented to the<figref idref="f0001">figure 1</figref>.

<u>EXAMPLE 3: EVIDENCE OF THE RELATIONSHIP BETWEEN GLYCANNIC STRUCTURE AND ACTIVITY DEPENDENT ON FCγRIII:</u>

<i>1- Cell culture in the presence of Deoxymannojirimycin (DMM).</i>

Several studies describe the effect of enzyme inhibitors on the glycosylation of immunoglobulins and on their biological activity. An increase in ADCC activity is reported by ROTHMAN et al. 1989, increase not attributable to an improvement in the affinity of the antibody for its target. The modification of glycosylation caused by the addition of DMM consists of an inhibition of the α1,2 mannosidase I present in the Golgi. It leads to the production of a larger proportion of polymannosylated, non-fucosylated structures.

Different lines producing anti-Rh (D) antibodies were brought into contact with DMM and the functional activity of the monoclonal antibodies produced was evaluated in the form of culture supernatants or after purification.

The cells (heterohybrids or lymphoblastoids) are seeded between 1 and 3x10⁵ cells / ml, and cultured in an IMDM culture medium (Life Technologies) with 10% of FCS and in the presence of 20 μg / ml of DMM (Sigma, Boehringer). After 3 medium renewals, the culture supernatants are tested by human IgG ELISA and then by ADCC.<tables id="tabl0001" num="0001"><table frame="all"><title>Table 2: Effect of the culture in the presence of DMM on the ADCC activity of different anti-Rh (D)</title><tgroup cols="4"><colspec colnum="1" colname="col1" colwidth="34mm" /><colspec colnum="2" colname="col2" colwidth="29mm" /><colspec colnum="3" colname="col3" colwidth="43mm" /><colspec colnum="4" colname="col4" colwidth="61mm" /><thead><row rowsep="0"><entry morerows="2" rowsep="1" align="center" valign="middle">Samples</entry><entry namest="col2" nameend="col3" align="center" valign="top">ADCC activity</entry><entry morerows="2" rowsep="1" align="center" valign="middle">Minimum DMM dose required µg / ml</entry></row><row><entry namest="col2" nameend="col3" align="center" valign="top">As% of LFB51 poly-D activity</entry></row><row><entry align="center" valign="top">Culture without DMM</entry><entry align="center" valign="top">Culture in the presence of DMM</entry></row></thead><tbody><row><entry align="center">F60</entry><entry align="center">109</entry><entry align="center">113</entry><entry align="center">NT</entry></row><row><entry align="center">D31</entry><entry align="center">19</entry><entry align="center">87</entry><entry align="center">10</entry></row><row><entry align="center">DF5</entry><entry align="center">26</entry><entry align="center">62</entry><entry align="center">20</entry></row><row><entry align="center">T125 RI (3)</entry><entry align="center">3</entry><entry align="center">72</entry><entry align="center">20</entry></row><row><entry align="center">T125-CHO</entry><entry align="center">0</entry><entry align="center">105</entry><entry align="center">5</entry></row></tbody></tgroup><tgroup cols="4" rowsep="0"><colspec colnum="1" colname="col1" colwidth="34mm" /><colspec colnum="2" colname="col2" colwidth="29mm" /><colspec colnum="3" colname="col3" colwidth="43mm" /><colspec colnum="4" colname="col4" colwidth="61mm" /><tbody><row><entry namest="col1" nameend="col4" align="justify"><u>NT = not tested</u></entry></row></tbody></tgroup></table></tables>• Culture in the presence of Deoxymannojirimycin (DMM) brings a significant improvement in ADCC results for weakly active antibodies previously produced by:<tables id="tabl0002" num="0002"><table frame="none"><tgroup cols="2" colsep="0" rowsep="0"><colspec colnum="1" colname="col1" colwidth="57mm" /><colspec colnum="2" colname="col2" colwidth="27mm" /><tbody><row><entry>A human-mouse hybridoma</entry><entry>D31</entry></row><row><entry>A human lymphoblastoid line</entry><entry>DF5</entry></row><row><entry>A transfected murine line</entry><entry>T125 in CHO</entry></row></tbody></tgroup></table></tables>• The addition of DMM can make it possible to restore the ADCC activity of an antibody originating from the cloid T 125 = T125 RI (3) (described in example 1) and which has lost this activity by a prolonged culture. • The strong activity of the antibody produced by the F60 heterohybridoma (the production of which is described in Example 1) is not modified by culture in the presence of DMM.

<i>2- Production of recombinant anti-D antibodies by different cell lines:</i>

2.1- Preparation of an expression vector for the antibody DF5:

The nucleotide sequence of the antibody DF5, negative control in the ADCC test, is used to study the transfection of this antibody in a few lines, in parallel with the transfection of the T125 antibody.

The sequences coding for Ac DF5 are isolated and amplified according to the same techniques used for the recombinant Ac T125-A2.<ul id="ul0024" list-style="bullet" compact="compact"><li>The corresponding cDNAs are first synthesized from total RNA extracted from the clone producer of anti-D Ab (IgG G1 / Lambda) 2MDF5 obtained by EBV transformation.</li><li>The amplification of the heavy and light chains is then carried out from these cDNAs using the primers presented below.</li><li>Amplification of the variable region of the heavy chain of DF5 (VH / DF5 sequence):<ul id="ul0025" list-style="dash" compact="compact"><li>primer DFSVH1 located 5 ′ of the leader region (in bold) of the VH gene of DF5 (published sequence: Chouchane L et al.); this primer also includes an Eco RI restriction site (in italics) and a Kozak sequence (underlined):<ul id="ul0026" list-style="none" compact="compact"><li>DF5VH1 (SEQ ID N ° 8):<img file="EP1518864B1_D0010.tif" /></li></ul></li><li>antisense primer GSP2ANP located 5 'from the constant region (CH) already described in paragraph 1.2 (example 2)</li></ul></li><li>Amplification of the constant region CH of DF5 (sequence CH / DF5): primers G1 and H3'Xba already described in paragraph 1.3 (example 2).</li><li>Amplification of the Lambda light chain of DF5 (LBD / DF5 sequence):<ul id="ul0027" list-style="dash" compact="compact"><li>primer DF5VLBD1 located in the 5 'leader region of the V5 gene of DF5, introduces a consensus sequence (in bold) deduced from the sequence of several leader regions of VL genes belonging to the same VL1 subgroup as the VL gene of 2MDF5; this sequence also includes an Eco RI restriction site (in italics) and a Kozak sequence (underlined):<ul id="ul0028" list-style="none" compact="compact"><li>DF5VLBD1 (SEQ ID N ° 9):<ul id="ul0029" list-style="none" compact="compact"><li>5'CCTACC<i>GAATTC</i>G<u>CCGCCACC</u><b>ATGGCCTGGTCTCCTCTCCTCCTCAC</b></li></ul></li></ul></li><li>antisense primer LSE1 located 3 'to Lambda, introduces an Eco RI site (underlined):<ul id="ul0030" list-style="none" compact="compact"><li>LSE1 (SEQ ID N ° 10):<ul id="ul0031" list-style="none" compact="compact"><li>5'- GAGGA<u>GGAATTCAC</u>TATGAACATTCTGTAGGGGCCACTGTCTT -3 '.</li></ul></li></ul></li></ul></li><li>The construction of the heavy chain (DFS-H31), light chain (DFS-L10) and heavy and light chain (DFS-IG1) expression vectors of Ac DF5 is carried out according to a construction scheme similar to the vectors expressing the Ac T125-A2. All the original Leader sequences (introduced at the level of the amplification primers) are conserved in these different vectors.</li></ul>

2.2- Transfection of different cell lines with T125 and DF5 antibodies

The three expression vectors T125-IG24, T125-LS4 and DF5-IgG1 are used for the transfection of cells of different lines:<ul id="ul0032" list-style="none" compact="compact"><li>Stable or transient transfections are carried out by electroporation or using transfection reagent.</li></ul><tables id="tabl0003" num="0003"><table frame="all"><title>Table 3: Cell lines used for the transfection of anti-Rh (D) antibodies</title><tgroup cols="3"><colspec colnum="1" colname="col1" colwidth="32mm" /><colspec colnum="2" colname="col2" colwidth="52mm" /><colspec colnum="3" colname="col3" colwidth="83mm" /><thead><row><entry align="center" valign="top"><b>Last name</b></entry><entry align="center" valign="top"><b>Reference</b></entry><entry align="center" valign="top"><b>Cell type</b></entry></row></thead><tbody><row><entry align="center">CHO-K1</entry><entry align="center">ATCC CCL 61</entry><entry align="center">Chinese hamster ovary (epithelium like)</entry></row><row><entry align="center">CHO-Lec10</entry><entry><nplcit id="ncit0001" npl-type="s"><text>Fenouillet et al., 1996, Virology, 218, 224 - 231</text></nplcit></entry><entry align="center">Chinese hamster ovary (epithelium like)</entry></row><row><entry align="center">Jurkat</entry><entry align="center">ATCC TIB-152</entry><entry align="center">Human T lymphocyte (T leukemia)</entry></row><row><entry align="center">Molt-4</entry><entry align="center">ATCC CRL 1582</entry><entry align="center">Human T cell (Acute lymphoblastic leukemia)</entry></row><row><entry align="center">WIL2-NS</entry><entry align="center">ATCC CRL 8155</entry><entry align="center">EBV transformed human B cell</entry></row><row><entry align="center">Vero</entry><entry align="center">ATCC CCL 81</entry><entry align="center">African green monkey kidney (fibroblast like)</entry></row><row><entry align="center">COS-7</entry><entry align="center">ATCC CRL 1651</entry><entry align="center">Transformed African green monkey kidney SV40 (fibroblast like)</entry></row><row><entry align="center">293-HEK</entry><entry align="center">ATCC CRL 1573</entry><entry align="center">Primary human embryonic kidney transformed by defective adenovirus 5 DNA</entry></row><row><entry align="center">YB2 / 0</entry><entry align="center">ATCC CRL 1662</entry><entry align="center">Non-secreting rat myeloma</entry></row><row><entry align="center">BHK-21</entry><entry align="center">ATCC CCL 10</entry><entry align="center">Newborn hamster kidney (fibroblast like)</entry></row><row><entry align="center">K6H6-B5</entry><entry align="center">ATCC CRL 1823</entry><entry align="center">non-secretory human-mouse heteromyeloma</entry></row><row><entry align="center">NSO</entry><entry align="center">ECACC 85110503</entry><entry align="center">non-secreting mouse myeloma (lymphoblate like)</entry></row><row><entry align="center">SP2 / 0- Ag 14</entry><entry align="center">ECACC 85072401</entry><entry align="center">Mouse x non secretory mouse hybridoma</entry></row><row><entry align="center">CHO Lec-1</entry><entry align="center">ATCC CRL 1735</entry><entry align="center">Chinese hamster ovary</entry></row><row><entry align="center">CHO dhfr-</entry><entry align="center">ECACC 94060607</entry><entry align="center">Chinese hamster ovary</entry></row><row><entry align="center">CHO Pro-5</entry><entry align="center">ATCC CRL 1781</entry><entry align="center">Chinese hamster ovary</entry></row><row><entry align="center">P3X63 Ag8.653</entry><entry align="center">ATCC CRL 1580</entry><entry align="center">Non-secreting mouse myeloma</entry></row></tbody></tgroup></table></tables>

After selection of the transformants in the presence of G418 (neo selection) several clones were isolated.

The modification of the effector activity of a humanized monoclonal antibody as a function of the expression cell has been described by CROWE et al (1992), with the cell lines CHO, NSO, YB2 / O.

The results obtained here confirm the importance of the expression cell line with respect to the functional characteristics of the antibody to be produced. Among the cells tested, only the Vero, YB2 / 0 and CHO Lec-1 lines make it possible to express recombinant anti-Rh (D) monoclonal antibodies with strong lytic activity in the ADCC test (see Example 1 and Table 3).<tables id="tabl0004" num="0004"><img file="EP1518864B1_D0011.tif" /></tables>

3- study of glycan structures

The characterization of the glycan structures of the anti-Rh-D antibody was carried out on four purified products having ADCC activity (F60, and three recombinant proteins originating from T 125) in comparison with two purified products inactive or very weakly active in the test. ADCC described here (D31 and DF5).

In practice, the oligosaccharides are separated from the protein by a specific enzymatic deglycosylation by PNGase F at the level of Asn 297. The oligosaccharides thus released are marked by a fluorophore, separated and identified by different complementary techniques which allow:<ul id="ul0033" list-style="dash" compact="compact"><li>A fine characterization of glycan structures by mass spectrometry with matrix-assisted laser desorption (MALDI) by comparison of experimental masses with theoretical masses.</li><li>Determination of the sialylation rate by ion exchange HPLC (GlycoSep C)</li><li>Separation and quantification of oligosaccharide forms according to hydrophilicity criteria by normal phase HPLC (GlycoSep N)</li><li>Separation and quantification of oligosaccharides by capillary electrophoresis with laser induced fluorescence detection (HPCE-LIF).</li></ul>

1) Characterization of the glycans of the active forms

The different active forms studied are F60 and three recombinant antibodies, R 290, R 297 and R 270, originating from T125 and produced in YB2 / 0. The fine characterization of glycan structures by mass spectrometry (<figref idref="f0007">figure 7</figref>) shows that these forms are all of the biantennial type. In the case of R 270 the majority form is of the non-fucosylated agalactosylated type (G0, mass exp. 1459.37 Da,<figref idref="f0001">fig 1</figref>). Three other structures are identified: fucosylated agalactosylated (G0F at 1605.41 Da), non-fucosylated monogalactosylated (G1 at 1621.26 Da) and fucosylated monogalactosylated (G1F at 1767.43 Da). These same four structures are characteristic of R 290, F 60 and R 297 (<figref idref="f0001">Figure 1</figref>).

These four antibodies active in ADCC are also characterized by the absence of oligosaccharides having an N-acetylglucosamine residue as a bisector.

The quantification of glycan structures by the different techniques of HPLC and HPCE-LIF (Table I) confirms the presence of the four forms identified in mass: G0, G0F, G1 and G1F. The sialylation rate is very low, in particular for recombinant products, from 1 to 9.4%, which is confirmed by the similarity of the mass spectra obtained before and after enzymatic desialylation. The rate of fucosylation varies from 34 to 59%.

2) Inactive forms

The different inactive forms studied are D31 and DF5. The quantification of glycan structures by the different chromatographic techniques and capillary electrophoresis (Table I) reveals, for these two antibodies, a rate of sialylation close to 50%, and a rate of fucosylation of 88 and 100% for D31 and DF5, respectively. These rates of sialylation and fucosylation are much higher than those obtained from the active forms.

The characterization of the glycan structures shows that the majority form is, for the two antibodies, of the biantenal monosialylated bigalactosylated fucosylated type (G2S1F, table I). Characterization by mass spectrometry of D31 (<figref idref="f0007">figure 7</figref>) reveals that the neutral forms are mainly of monogalactosylated fucosylated type (G1F at 1767.43 Da) and bigalactosylated fucosylated type (G2F at 1929.66 Da).

The inactive antibody DF5 is characterized by the presence of oligosaccharides having an intermediate GIcNAc residue. In particular, the mass analysis (<figref idref="f0008">figure 8</figref>) reveals the presence of a majority neutral form of monogalactosylated fucosylated type Bisec-GlcNAc interlayer (G1FB at 1851.03 Da). On the other hand, these structural forms are not detectable or present in a trace state on the active antibodies studied.

The ADCC activity of D31 after action of DMM goes from 10% to 60%. The glycan structures of D31 DMM differ from those of D31 by the presence of oligomannosis forms (Man 5, Man 6 and Man 7) (see<figref idref="f0009">figure 9</figref>).

3) Conclusion

The various active antibodies are modified on Asn 297 by N-glycosylations of the biantenate and / or oligomanosidic type. For the biantenate forms, these are short structures that are very weakly sialylated, weakly fucosylated, weakly galactosylated and without interlayer GlcNAc.

REFERENCES

<ul id="ul0034" list-style="none"><li><nplcit id="ncit0002" npl-type="s"><text>Boylston, JM, Gardner, B., Anderson, RL, and Hughes-Jones, NC Production of human IgM anti-D in tissue culture by EB virus-transformed lymphocytes. Scand. J. Immunol. 12: 355-358 (1980</text></nplcit>).</li><li><nplcit id="ncit0003" npl-type="s"><text>Bron, D., Feinberg, MB, Teng, NNH and Kaplan, HS Production of Human Monoclonal IgG Antibodies against Rhesus (D) Antigen. Proc. Nat. Acad. Sci. USA 81: 3214-3217 (1984</text></nplcit>).</li><li><nplcit id="ncit0004" npl-type="s"><text>Chouchane, L., Van Spronsen, A., Breyer, J., Guglielmi, P., and Strosberg, AD. Molecular characterization of a human anti-Rh (D) antibody with a DH segment encoded by a germ-line sequence. Eur. J. Biochem. 1; 207 (3): 1115-1121 (1992</text></nplcit>).</li><li><nplcit id="ncit0005" npl-type="s"><text>Crawford, DH, Barlow, MJ, Harrison, JF, Winger, L. and Huehns, ER Production of human monoclonal antibody to rhesus D antigen. Lancet, i: 386-388 (1983</text></nplcit>).</li><li><nplcit id="ncit0006" npl-type="s"><text>Doyle, A., Jones, TJ, Bidwell, JL and Bradley, BA In vitro development of human monoclonal antibody secreting plasmacytomas. Hmm. Immunol. 13: 199-209 (1985</text></nplcit>).</li><li><nplcit id="ncit0007" npl-type="s"><text>Edelman, L., Margaritte, C., Chaabihi, H., Monchâtre, E., Blanchard, D., Cardona, A., Morin, F., Dumas, G., Petres, S. and Kaczorek, M. Obtaining a functional recombinant anti-rhesus (D) antibody using the baculovirus-insect cell expression System. Immunology, Vol. 91 (1), 13-19 (1997</text></nplcit>).</li><li><nplcit id="ncit0008" npl-type="s"><text>Foung, SKH, Blunt, JA, Wu, PS, Ahearn, P., Winn, LC, Engleman, EG and Grumet, FC Human Monoclonal Antibodies to Rho (D). Vox Sang. 53: 44-47 (1987</text></nplcit>).</li><li><nplcit id="ncit0009" npl-type="s"><text>Goossens, D., Champomier, F., Rouger, P., and Salmon, C. Human Monoclonal Antibodies against Blood Group Antigens: Preparation of a series of stable EBV immortalized B clones producing high levels of antibody of different isotypes and specificities. J. Immunol. Methods 101: 193-200 (1987</text></nplcit>).</li><li><nplcit id="ncit0010" npl-type="s"><text>Issitt, PD Genetics of the Rh Blood Group System: Some Current Concepts. Med. Lab. Sci. 45: 395-404 (1988</text></nplcit>).</li><li><nplcit id="ncit0011" npl-type="s"><text>Jefferis, R., Lund, J., Mizutani, H., Nakagawa, H., Kawazoe, Y., Arata, Y. and Takahashi, N. A comparative study of the N-linked oligosaccharides structure of human IgG Subclass proteins. Biochem. J., 268: 529-537 (1990</text></nplcit>).</li><li><nplcit id="ncit0012" npl-type="s"><text>Koskimies, S. Human Lymphoblastoid Cell Line Producing Specific Antibody against Rh-Antigen D. Scand. Immunol. 11: 73-77 (1980</text></nplcit>).</li><li><nplcit id="ncit0013" npl-type="s"><text>Kumpel, BM, Goodrick, MJ, Pamphilon, DH, Fraser, ID, Poole GD, Morse, C., Standen, GR, Chapman, GE, Thomas, DP and Anstee, DJ Human Rh D monoclonal antibodies (BRAD-3 and BRAD -5) Cause Accelerated Clearance of Rh D + Red blood Cells and Suppression of Rh D Immunization in Rh D Volunteers. Blood, Vol. 86, no.5, 1701-1709 (1995</text></nplcit>).</li><li><nplcit id="ncit0014" npl-type="s"><text>Kumpel, BM, Poole, GD and Bradley, BA Human Monoclonal Anti-D Antibodies. I. Their Production, Serology, Quantitation and Potential Use as Blood Grouping Reagents. Brit. J. Haemat. 71: 125-129 (1989a</text></nplcit>).</li><li><nplcit id="ncit0015" npl-type="s"><text>Kumpel, BM, Rademacher, TW, Rook, GAW, Williams, PJ, Wilson, IBM Galacatosylation of human IgG anti-D produced by EBV-transformed B lymphoblastoid cell lines is dependent on culture method and affects Fc receptor mediated functional activity. Hmm. Antibodies and Hybridomas, 5: 143-151 (1994</text></nplcit>). <nplcit id="ncit0016" npl-type="s"><text>Leatherbarrow, RJ, Rademacher, TW, Dwek, RA, Woof, JM, Clark, A., Burton, DR, Richardson, N. and Feinstein, A. Effector functions of monoclonal aglycosylated mouse IgG2a; binding and activation of complement component C1 and interaction with human Fc receptor. Molec. Immun. 22, 407-415 (1985</text></nplcit>).</li><li><nplcit id="ncit0017" npl-type="s"><text>Lifely MR, Hale C, Boyce S, Keen MJ, Phillips J. Glycosylation and biological activity of CAMPATH-1H expressed in different cell lines and grown under different culture conditions. Glycobiology. 1995 Dec; 5 (8): 813-22</text></nplcit>.</li><li><nplcit id="ncit0018" npl-type="s"><text>Lomas, C., Tippett, P., Thompson, KM, Melamed, MD and Hughes-Jones, NC Demonstration of seven epitopes on the Rh antigen D using human monoclonal anti-D antibodies and red cells from D categories. Vox Sang. 57: 261-264 (1989</text></nplcit>).</li><li><nplcit id="ncit0019" npl-type="s"><text>Lund, J., Takahaski, N., Nakagawa, H., Goodall, M., Bentley, T., Hindley, SA, Tyler, R. and Jefferis, R. Control of IgG / Fc glycosylation: a comparison of oligosaccharides from chimeric human / mouse and mouse subclass immunoglobulin G5. Molec. Immun. 30, no.8, 741-748 (1993</text></nplcit>).</li><li><nplcit id="ncit0020" npl-type="s"><text>Lund, J., Tanaka, T., Takahashi, N., Sarmay, G., Arata, Y. and Jefferis, R. A protein structural change in aglycosylated IgG3 correlates with loss of hu Fc □ RI and hu FcγRIII binding and / or activation. Molec. Immun. 27, 1145-1153 (1990</text></nplcit>).</li><li><nplcit id="ncit0021" npl-type="s"><text>Ma, JK and Hein, MB Immunotherapeutic potential of antibodies produced in plants.Trends Biotechnol. 13, 522-527 (1995</text></nplcit>).</li><li><nplcit id="ncit0022" npl-type="s"><text>Mc Cann-Carter, MC, Bruce, M., Shaw, EM, Thorpe, SJ, Sweeney, GM, Armstrong, SS and James, K. The production and evaluation of two human monoclonal anti-D antibodies. Transf. Med. 3: 187-194 (1993</text></nplcit>).</li><li><nplcit id="ncit0023" npl-type="s"><text>Melamed, MD, Gordon, J., Ley, SJ, Edgar, D. and Hughes-Jones, NC Senescence of a human lymphoblastoid clone producing anti-Rhesus (D). Eur. J. Immunol. 115: 742-746 (1985</text></nplcit>).</li><li><nplcit id="ncit0024" npl-type="s"><text>Parekh, RB, Dwek, RA, Sutton, BJ, Fernandes, DL, Leung, A., Stanworth, D., Rademacher, TW, Mizuochi, T., Taniguchi, T., Matsuta, K., Takeuchi, F., Nagano, Y., Miyamoto, T. and Kobata, A. Asssociation of rheumatoid arthritis and primary osteoarthritis with changes in the glycosylation pattern of total serum IgG. Nature, 316: 452-457 (1985</text></nplcit>).</li><li><nplcit id="ncit0025" npl-type="s"><text>Rothman, RJ, Perussia, B., Herlyn, D. and Warren, L. Antibody-dependent cytotoxicity mediated by natural killer cells is enhanced by castanospermine-induced alterations of IgG glycosylation. Mol. Immunol. 26 (12): 1113-1123 (1989</text></nplcit>).</li><li><nplcit id="ncit0026" npl-type="s"><text>Shitara K., Nakamura K., Tokutake-Tanaka Y., Fukushima M., and Hanai N. A new vector for the high level expression of chimeric antibodies to myeloma cells. J. Immunol. Methods 167: 271-278 (1994</text></nplcit>).</li><li><nplcit id="ncit0027" npl-type="s"><text>Thompson, KM, Hough, DW, Maddison, PJ, Melamed, MD and Hughes-Jones, NC Production of human monoclonal IgG and IgM antibodies with anti-D (rhesus) specificity using heterohybridomas. Immunology 58: 157-160 (1986</text></nplcit>)</li><li><nplcit id="ncit0028" npl-type="s"><text>Thomson, A., Contreras, M., Gorick, B., Kumpel, B., Chapman, GE Lane, RS, Teesdale, P. Hughes-Jones, NC and Mollison, PL Clearance of Rh D-positive red cells with monoclonal anti-D. Lancet 336: 1147-1150 (1990</text></nplcit>).</li><li><nplcit id="ncit0029" npl-type="s"><text>Tippett, P. Sub-divisions of the Rh (D) antigen. Med. Lab. Sci. 45: 88-93 (1988</text></nplcit>).</li><li><nplcit id="ncit0030" npl-type="s"><text>Umaña P, Jean-Mairet J, Moudry R, ​​Amstutz H, Bailey JE. Engineered glycoforms of an antineuroblastoma IgG1 with optimized antibody-dependent cellular cytotoxic activity. Nat Biotechnol. 1999 Feb; 17 (2): 176-80</text></nplcit>.</li><li><nplcit id="ncit0031" npl-type="s"><text>Ware, RE and Zimmerman, SA Anti-D: Mechanisms of action. Seminars in Hematology, vol. 35, n ° 1, supp. 1: 14-22 (1998</text></nplcit>).</li><li><patcit id="pcit0016" dnum="WO9954342A1"><text>WO99 / ​​54342A1</text></patcit></li><li><nplcit id="ncit0032" npl-type="s"><text>Wright A, Morrison SL. Effect of glycosylation on antibody function: implications for genetic engineering. Trends Biotechnol. 1997 Jan; 15 (1): 26-32</text></nplcit>.</li><li><nplcit id="ncit0033" npl-type="s"><text>Yu, IPC, Miller, WJ, Silberklang, M., Mark, GE, Ellis, RW, Huang, L., Glushka, J., Van Halbeek, H., Zhu, J. and Alhadeff, JA Structural characterization of the N -Glycans of a humanized anti-CD18 murine immunoglobulin G. Arch. Biochem. Biophys. 308, 387-399 (1994</text></nplcit>).</li><li><nplcit id="ncit0034" npl-type="s"><text>Zupanska, B., Thompson, E., Brojer, E. and Merry, AH Phagocytosis of Erythrocytes Sensitized with Known Amounts of IgG1 and IgG3 anti-Rh antibodies. Vox Sang. 53: 96-101 (1987</text></nplcit>).</li></ul>