WO2017162733A1

WO2017162733A1 - Iga antibodies with enhanced stability

Info

Publication number: WO2017162733A1
Application number: PCT/EP2017/056813
Authority: WO
Inventors: Steffen Goletz; Antje Danielczyk; Felix Hart
Original assignee: Glycotope Gmbh
Priority date: 2016-03-22
Filing date: 2017-03-22
Publication date: 2017-09-28

Abstract

The present invention pertains to IgA antibodies having an optimized glycosylation pattern. These IgA antibodies exhibit an increased stability and an extended shelf life. Compositions comprising IgA antibodies having an improved glycosylation pattern and methods for improving the shelf life of IgA antibodies are provided.

Description

"IgA antibodies with enhanced stability"

FIELD OF THE INVENTION

The present invention is directed to IgA antibodies having an optimized glycosylation pattern. The superior glycosylation pattern provides the IgA antibodies with unexpected advantages, including an increased stability and a long shelf life. The invention therefore provides a composition comprising IgA antibodies having an improved glycosylation pattern and a method for improving the shelf life of IgA antibodies.

BACKGROUND OF THE INVENTION

IgA is the predominant immunoglobulin in mucosal secretions, where it serves as first line of defense and protects against pathogens that are ingested or inhaled. In human serum, IgA is the second most prevalent antibody, with a comparable rate of synthesis like IgG. However, due to faster catabolism, the serum concentration of IgA is smaller than of IgG. Serum IgA is mainly monomeric, while secretory IgA is dimeric and can be associated with additional polypeptides, the joining (J) chain and secretory component (SC). Per day an adult produces about 66 mg IgA antibodies per kilogram, which is more than all other classes combined.

IgA monomers consist of two light chains (LC, kappa or lambda) and two heavy chains (HC). The alpha HC contains a variable region (VH) and three constant regions, Ca1 , Ca2 and Ca3. In humans there are two alpha HC genes, encoding lgA1 or lgA2 subclasses. Both, lgA1 and lgA2 HC contain an 18 amino acid C-terminal tailpiece with penultimate Cys residue, which is found on IgM antibodies as well. There is one allotype for lgA1 , whereas three allotypes for lgA2 were described: lgA2m(1 ), lgA2m(2) and lgA2(n). Of note, lgA2m(1 ) lacks disulfide bridges between HCs and LCs which is found in other isotypes, instead the two LCs are linked by disulfide bonds and HC-LC interactions are non-covalent. lgA1 possesses a 23 amino acid hinge region, while the hinge region of lgA2 is 10 amino acids in length.

The extended hinge region is rich in proline, serine and threonine and carries two to five, in some cases up to six, O-glycosylations. The most favored O-glycosylation sites are: Thr228, Ser230, Ser323. Each O-glycosylation at a serine or threonine in the hinge region can be composed of N-acetylgalactosamine, galactose and sialic acid in a heterogenic mixture. The extended T-shape structure of lgA1 , compared to the Y- shape of other Igs like lgG1 or lgA2 (lgA2 can be present in a compact T-shape as well), could be caused by the O-glycosylated extended hinge region and allow lgA1 to reach more distant antigens simultaneously. Apart from the O-glycosylation, lgA1 has two N-glycosylation sites: asparagine 263 (Asn263) in Ca2 and Asn459 in the tailpiece at the C terminus. For lgA2m(1 ) there are two additional N-glycosylation sites, Asn166 in Ca1 and Asn337 in Ca2. Asn21 1 is a fifth N-glycosylation site present in Ca1 of lgA2m(2). Most of the N-glycans are biantennary complex-type structures with a minority of triantennary and tetraantennary glycans as well as some oligomannose structures. For recombinant lgA1 , N-glycosylation sites for Ca2 and tailpiece possess mainly biantennary and triantennary glycans, respectively. In contrast to lgG1 where glycans are orientated towards the inside between the two Cy2 domains, molecular modeling for lgA1 revealed a glycan orientation towards the outside of the molecule. Compared to IgG, IgA glycosylation is characterized by more complete processing, potentially because of better accessibility of the glycans for glycosyltransferases within the Golgi. The degree of sialylation is much higher for IgA than for IgG, 90 % and 10 %, respectively (Mattu 1998).

While IgG antibodies are only present as monomers, IgA antibodies can form dimers and higher order polymers by themselves or in complex with additional components. Mucosal plasma cells co-express the J chain, a 15 kDa polypeptide which associates with IgA dimers. A disulfide bridge is formed between J chain and IgA Cys residues. The J chain has one N-glycosylation site and contains three intra-chain disulfide bonds. Two additional disulfide bonds are formed between the J chain and two IgA monomer units within the tailpiece. The same polypeptide associates with pentameric IgM antibodies. A second polypeptide associates with secretory IgA, the SC is about 80 kDa in size and the extracellular part of the polymeric immunoglobulin receptor (plgR). During epithelial cell transcytosis mediated by the plgR, the SC is cleaved off at the apical site to release SlgA, which consist of two IgA monomers, one J chain and one SC.

The major receptor binding IgA is Fc alpha receptor I (FcaRI), which is also called cluster of differentiation (CD) 89. It is expressed on neutrophils, eosinophils, monocytes, macrophages, intestinal dendritic cells and Kupffer cells. On the extracellular site, it possesses two 206-amino acids Ig-like domains (EC1 and EC2), of which N-terminal EC1 binds IgA within the Ca2/Ca3 boundary. On the IgA HC, three Leu residues in the Ca2 and 16 residues in Ca3 are involved in FcaRI binding.

FcaRI crosslinking by IgA results in inflammatory activation of effector cells such as granulocytes, monocytes or macrophages. As a second line of defense in pathological conditions, Kupffer cells in the liver can phagocyte IgA-coated pathogens present in the circulation. Inflammatory signals mediated by FcaRI induce cellular responses like phagocytosis, respiratory burst and release of cytokines. Furthermore, IgA antibodies can inhibit inflammatory responses through ITAMi and SHP-1 . Therefore, IgA antibodies have neutralizing and inflammatory functions, but at the same time also represent inflammatory regulators to prevent excessive immune responses. Taken together, these diverse functions could be harnessed to use tumor-specific IgA antibodies for cancer immunotherapy. In circulation therapeutic IgA antibodies could stay silent, while at tumor sites with high antigen expression IgA antibodies could mediate crosslinking of FcaRI on effector cells resulting in inflammatory responses.

The number of antibodies developed for clinical application in tumor therapy is increasing. Various tumor-associated surface antigens offer the potential to specifically target tumor cells using antibodies to attract components of the immune system. In addition, systemically administered antibodies can conduct direct Fab-mediated effector functions to diminish uncontrolled proliferation of cancer cells. Today, monoclonal immunoglobulin (Ig) G isotype antibodies have improved current cancer therapy regiments. Several tumor-associated targets like CD20 and epidermal growth factors have been clinically validated for the treatment of hematological and solid cancers, respectively, and more novel targets are emerging. However, these magic bullets still lack some of their magic. Epidermal growth factor receptor (EGFR) antibodies cetuximab (Erbitux) and panitumumab (Vectibix) are used for the treatment of colorectal and head and neck cancer. Clinical response rates for these mAbs are in the range of 10 to 15 % when used as single agent. Since the first approval of a mAb for cancer therapy in 1997 (anti-CD20, rituximab), research pursued to further improve current formats by for example finding the best tumor-associated target, engaging multiple targets at a time or conjugate mAbs with cytotoxic agents. In other approaches, the protein backbone is modified to improve serum half-life, effector cell receptor binding or increase avidity.

Today, different expression systems are used for the production of therapeutic antibodies. Still, in most cases rodent cells are genetically modified in order to produce recombinant proteins. For example, Mabthera, Obinutuzumab, Vectibix and Herceptin are produced in Chinese hamster ovary (CHO) cells and Erbitux is produced in murine Sp2/0 cells. One important posttranslational modification of antibodies is the glycosylation which consist of multiple branched carbohydrate units. Within the endoplasmic reticulum and Golgi of eukaryotes, many enzymes are responsible for the composition of the glycans attached to proteins and both, enzymes and the resulting glycans differ between species. While rodent and human expression systems share common characteristics, some distinct features can affect safety or efficacy. Therefore, structures which are absent in human glycans like a(1 -3)-linked galactose or a(2-3)- linked N-glycolylneuraminic acid are potentially immunogenic for patients. Moreover, CHO cells are not capable to link N-acetylneuraminic acid in the a(2-3) configuration which is present in humans. These limitations of commonly used non-human expression systems can be overcome by fully human systems. In addition to the fully human glycans, glycoengineering was used to generate cell lines which produce biologies with specific glycosylation patterns. For example, the reduction in core fucose of N-glycans within the Cy2 domain of IgG antibodies results in increased FcYRIIIa affinity and Fc-mediated effector functionality.

Today, more than twelve monoclonal antibodies are approved for cancer therapy, all of which are IgG class antibodies. Methods for IgG production and purification are well known and this isotype benefits from prolonged serum half-life through receptor- mediated recycling. However, several studies indicate better cytotoxic effects against cancer cells in vitro when engaging FcaRI on neutrophils. Using granulocytes as source of effector cells, FcaRI-engagement can induce antibody-dependent cellular cytotoxicity (ADCC) against tumor cells. Of all leukocytes in human blood, granulocytes (also called polymorphonuclear leukocytes) are the most numerous cell population and therefore attractive components of the immune system to target cancer cells. Hence, effectively recruiting these cells could potentially improve current immunotherapeutic approaches. While it is well established that recombinant IgA antibodies or simultaneous FcaRI-binding and tumor antigen-binding bispecific antibodies can recruit immune effector cells for cytotoxic effects on cancer cells, the production and purification of monoclonal IgA antibodies is not. IgA production rates as well as yields using conventional expression systems are often low and assembly is sometimes not complete. The presence of serum in growth medium hinders further development in order to produce potential biotherapeutics for human use. The high number of glycosylation sites potentially results in more heterogeneous recombinant products as compared to IgG antibodies. The IgA glycan orientation towards the outside and their role in healthy conditions as well as disease highlights the importance of correct human-like IgA production.

Hence, there is a need in the art to provide therapeutic IgA antibodies having improved properties for their use in cancer therapy. SUMMARY OF THE INVENTION

The present invention is based on the findings that IgA antibodies having an optimized glycosylation pattern exhibit improved and advantageous properties. In particular, IgA antibodies comprising complex-type N-glycans with a significant amount of bisecting GlcNAc and a high amount of sialic acids, especially 2,6-coupled sialic acid residues, and with a significant amount of hybrid-type N-glycans show desired properties. It was demonstrated that such IgA antibodies are very stable and have a very long shelf life. Furthermore, they show good antigen and downstream receptor binding. They are well suited for therapeutic applications, especially in cancer immune therapy, due to their excellent ability to inhibit cancer cell proliferation and induce ADCC against cancer cells.

In a first aspect, the present invention provides a composition comprising IgA antibodies, wherein the IgA antibodies in the composition have an N-glycosylation pattern comprising one or more of the following characteristics:

(i) a relative amount of complex-type glycans carrying bisecting N- acetylglucosamine (bisGlcNAc) of at least 3% of the total amount of complex- type glycans attached to the IgA antibodies in the composition; and/or

(ii) a relative amount of complex-type glycans carrying N-acetyl neuraminic acid (NeuNAc) of at least 40% of the total amount of complex-type glycans attached to the IgA antibodies in the composition; and/or

(iii) a relative amount of hybrid-type glycans of at least 3% of the total amount of N-glycans attached to the IgA antibodies in the composition.

In a second aspect, the present invention provides a method for increasing the shelf life of IgA antibodies, comprising providing the IgA antibodies with an N-glycosylation pattern comprising one or more of the following characteristics:

(iii) a relative amount of hybrid-type glycans of at least 3% of the total amount of N-glycans attached to the IgA antibodies in the composition. The above aspects can be combined. Other objects, features, advantages and aspects of the present invention will become apparent to those skilled in the art from the following description and appended claims. It should be understood, however, that the following description, appended claims, and specific examples, which indicate preferred embodiments of the application, are given by way of illustration only. Various changes and modifications within the spirit and scope of the disclosed invention will become readily apparent to those skilled in the art from reading the following.

DEFINITIONS

As used herein, the following expressions are generally intended to preferably have the meanings as set forth below, except to the extent that the context in which they are used indicates otherwise.

The expression "comprise", as used herein, besides its literal meaning also includes and specifically refers to the expressions "consist essentially of" and "consist of". Thus, the expression "comprise" refers to embodiments wherein the subject-matter which "comprises" specifically listed elements may and/or indeed does encompass further elements as well as embodiments wherein the subject-matter which "comprises" specifically listed elements does not comprise further elements. Likewise, the expression "have" is to be understood as the expression "comprise", also including and specifically referring to the expressions "consist essentially of" and "consist of".

The term "antibody" in particular refers to a protein comprising at least two heavy chains and two light chains connected by disulfide bonds. Each heavy chain is comprised of a heavy chain variable region (VH) and a heavy chain constant region (CH). Each light chain is comprised of a light chain variable region (VL) and a light chain constant region (CL). The heavy chain-constant region comprises three or - in the case of antibodies of the IgM- or IgE-type - four heavy chain-constant domains (CH1 , CH2, CH3 and CH4) wherein the first constant domain CH1 is adjacent to the variable region and may be connected to the second constant domain CH2 by a hinge region. The light chain-constant region consists only of one constant domain. The variable regions can be further subdivided into regions of hypervariability, termed complementarity determining regions (CDRs), interspersed with regions that are more conserved, termed framework regions (FR), wherein each variable region comprises three CDRs and four FRs. The variable regions of the heavy and light chains contain a binding domain that interacts with an antigen. The heavy chain constant regions may be of any type such as γ-, δ-, α-, μ- or ε-type heavy chains. With respect to IgA antibodies, the heavy chain of the antibody is an a-chain. Furthermore, the light chain constant region may also be of any type such as κ- or λ-type light chains. Preferably, the light chain of the antibody is a κ-chain. The constant regions of the antibodies may mediate the binding of the immunoglobulin to host tissues or factors, including various cells of the immune system (e.g., effector cells) and the first component (C1 q) of the classical complement system. The antibody can be e.g. a humanized, human or chimeric antibody. In certain embodiments, the antibody is a monoclonal antibody and/or a recombinant antibody. The antibody according to the invention is capable of inducing ADCC.

The antigen-binding portion of an antibody usually refers to full length or one or more fragments of an antibody that retain the ability to specifically bind to an antigen. It has been shown that the antigen-binding function of an antibody can be performed by fragments of a full-length antibody. Examples of binding fragments of an antibody include a Fab fragment, a monovalent fragment consisting of the VL, VH, CL and CH1 domains; a F(ab)₂ fragment, a bivalent fragment comprising two Fab fragments, each of which binds to the same antigen, linked by a disulfide bridge at the hinge region; a Fd fragment consisting of the VH and CH1 domains; a Fv fragment consisting of the VL and VH domains of a single arm of an antibody; a dAb fragment, which consists of a VH domain; and an isolated complementarity determining region (CDR). The "Fab part" of an antibody in particular refers to a part of the antibody comprising the heavy and light chain variable regions (VH and VL) and the first heavy and light chain constant regions (CH1 and CL). In cases where the antibody does not comprise all of these regions, then the term "Fab part" only refers to those of the regions VH, VL, CH1 and CL which are present in the antibody. Preferably, "Fab part" refers to that part of an antibody corresponding to the fragment obtained by digesting a natural antibody with papain which contains the antigen binding activity of the antibody. In particular, the Fab part of an antibody encompasses the antigen binding site or antigen binding ability thereof. Preferably, the Fab part comprises at least the VH region of the antibody.

The "Fc part" of an antibody in particular refers to a part of the antibody comprising the heavy chain constant regions 2, 3 and - where applicable - 4 (CH2, CH3 and CH4). In cases where the antibody does not comprise all of these regions, then the term "Fc part" only refers to those of the regions CH2, CH3 and CH4 which are present in the antibody. Preferably, the Fc part comprises at least the CH2 region of the antibody. Preferably, "Fc part" refers to that part of an antibody corresponding to the fragment obtained by digesting a natural antibody with papain which does not contain the antigen binding activity of the antibody. In particular, the Fc part of an antibody is capable of binding to the Fc receptor and thus, e.g. comprises a Fc receptor binding site or a Fc receptor binding ability.

For indicating the amino acid positions of the heavy chain and light chain, in particular the variable regions thereof, the Kabat numbering system is used herein (Kabat, E.A. et al. (1991 ) Sequences of Proteins of Immunological Interest, 5^th edition, NIH Publication No. 91 -3242). According to said system, the heavy chain variable region comprises amino acid positions from position 0 to position 1 13 including position 35A, 35B, 52A to 52C, 82A to 82C and 100A to 100K. The CDRs of the heavy chain variable region are located, according to the Kabat numbering, at positions 31 to 35B (CDR1 ), 50 to 65 (CDR2) and 95 to 102 (CDR3). The remaining amino acid positions form the framework regions FR1 to FR4. The light chain variable region comprises positions 0 to 109 including positions 27A to 27F, 95A to 95F and 106A. The CDRs are located at positions 24 to 34 (CDR1 ), 50 to 56 (CDR2) and 89 to 97 (CDR3). Depending on the initial formation of the specific gene of an antibody, not all of these positions have to be present in a given heavy chain variable region or light chain variable region. In case an amino acid position in a heavy chain or light chain variable region is mentioned herein, unless otherwise indicated it is referred to the position according to the Kabat numbering.

According to the present invention, the term "chimeric antibody" in particular refers to an antibody wherein the constant regions are derived from a human antibody or a human antibody consensus sequence, and wherein at least one and preferably both variable regions are derived from a non-human antibody, e.g. from a rodent antibody such as a mouse antibody.

According to the present invention, the term "humanized antibody" in particular refers to a non-human antibody comprising human constant regions and variable regions which amino acid sequences are modified so as to reduce the immunogenicity of the antibody when administered to the human body. An exemplary method for constructing humanized antibodies is CDR grafting, wherein the CDRs or the specificity determining residues (SDRs) of a non-human antibody are combined with human-derived framework regions. Optionally, some residues of the human framework regions may be backmutated towards the residues of the parent non-human antibody, e.g. for increasing or restoring the antigen binding affinity. Other humanization methods include, for example, resurfacing, superhumanization, and human string content optimization. In the resurfacing methods, only those residues of the non-human framework regions which are positioned at the surface of the antibody are replaced by residues present in corresponding human antibody sequences at said position. Superhumanization essentially corresponds to CDR grafting. However, while during CDR grafting the human framework regions are normally chosen based on their homology to the non-human framework regions, in superhumanization it is the similarity of the CDRs on the basis of which the human framework regions are chosen. In the human string content optimization the differences of the non-human antibody sequence to the human germline sequences is scored and then the antibody is mutated to minimize said score. Furthermore, humanized antibodies can also be obtained by empirical methods wherein large libraries of human framework regions or human antibodies are used to generate multiple antibody humanized candidates and then the most promising candidate is determined by screening methods. Also with the above- described rational approaches several humanized antibody candidates can be generated and then screened, for example for their antigen binding. Overviews of humanization processes can be found, for example, in Almagro, J.C. and Fransson, J. (2008) Frontiers in Bioscience 13, 1619-1633 and in the entire volume 36 of the Journal Methods (2005).

The term "human antibody", as used herein, is intended to include antibodies having variable regions in which both the framework and CDR regions are derived from sequences of human origin.

A target amino acid sequence is "derived" from or "corresponds" to a reference amino acid sequence if the target amino acid sequence shares a homology or identity over its entire length with a corresponding part of the reference amino acid sequence of at least 75 %, more preferably at least 80 %, at least 85 %, at least 90 %, at least 93 %, at least 95 % or at least 97 %. For example, if a framework region of a humanized antibody is derived from or corresponds to a variable region of a particular human antibody, then the amino acid of the framework region of the humanized antibody shares a homology or identity over its entire length with the corresponding framework region of the human antibody of at least 75 %, more preferably at least 80 %, at least 85 %, at least 90 %, at least 93 %, at least 95 % or at least 97 %. The "corresponding part" means that, for example, framework region 1 of a heavy chain variable region (FRH1 ) of a target antibody corresponds to framework region 1 of the heavy chain variable region of the reference antibody. In particular embodiments, a target amino acid sequence which is "derived" from or "corresponds" to a reference amino acid sequence is 100% homologous, or in particular 100 % identical, over its entire length with a corresponding part of the reference amino acid sequence. A "homology" or "identity" of an amino acid sequence or nucleotide sequence is preferably determined according to the invention over the entire length of the reference sequence or over the entire length of the corresponding part of the reference sequence which corresponds to the sequence which homology or identity is defined.

"Specific binding" preferably means that an agent such as an antibody binds stronger to a target such as an epitope for which it is specific compared to the binding to another target. An agent binds stronger to a first target compared to a second target if it binds to the first target with a dissociation constant (K_d) which is lower than the dissociation constant for the second target. Preferably the dissociation constant for the target to which the agent binds specifically is more than 100-fold, 200-fold, 500-fold or more than 1000-fold lower than the dissociation constant for the target to which the agent does not bind specifically. Furthermore, the term "specific binding" in particular indicates a binding affinity between the binding partners with a K_a of at least 10⁶ M^~ , preferably at least 10⁷ M^~ , more preferably at least 10⁸ M^" . An antibody specific for a certain antigen in particular refers to an antibody which is capable of binding to said antigen with an affinity having a K_a of at least 10⁶ M ^~1 , preferably at least 10⁷ M ^~\ more preferably at least 10⁸ M^" . For example, the term "anti-EGFR antibody" refers to an antibody specifically binding EGFR and preferably is capable of binding to EGFR with an affinity having a K_a of at least 10⁶ M ^~\ preferably at least 10⁷ M ^~\ more preferably at least 10⁸ M^" .

The term "sialic acid" in particular refers to any N- or O-substituted derivatives of neuraminic acid. It may refer to both 5-N-acetylneuraminic acid and 5-N- glycolylneuraminic acid, but preferably only refers to 5-N-acetylneuraminic acid. The sialic acid, in particular the 5-N-acetylneuraminic acid preferably is attached to a carbohydrate chain via a 2,3- or 2,6-linkage. Preferably, in the IgA antibodies described herein both 2,3- as well as 2,6-coupled sialic acids are present.

A "relative amount of glycans" according to the invention refers to a specific percentage or percentage range of the glycans attached to the IgA antibodies of an IgA preparation or in a composition comprising IgA antibodies, respectively. In particular, the relative amount of glycans refers to a specific percentage or percentage range of all glycans comprised in the IgA antibodies and thus, attached to the polypeptide chains of the IgA antibodies in an IgA preparation or in a composition comprising IgA antibodies. 100 % of the glycans refers to all glycans attached to the IgA antibodies of the IgA preparation or in a composition comprising IgA antibodies, respectively. For example, a relative amount of glycans carrying bisecting GlcNAc of 60% refers to a composition comprising IgA antibodies wherein 60% of all glycans comprised in the IgA antibodies and thus, attached to the IgA polypeptide chains in said composition comprise a bisecting GlcNAc residue while 40% of all glycans comprised in the IgA antibodies and thus, attached to the IgA polypeptide chains in said Composition do not comprise a bisecting GlcNAc residue. The corresponding reference amount of glycans representing 100% may either be all glycan structures attached to the IgA antibodies in the composition, or all N-glycans, i.e. all glycan structures attached to an asparagine residue of the IgA antibodies in the composition, or all complex-type glycans. The reference group of glycan structures generally is explicitly indicated or directly derivable from the circumstances by the skilled person.

The term "N-glycosylation" refers to all glycans attached to asparagine residues of the polypeptide chain of a protein. These asparagine residues generally are part of N- glycosylation sites having the amino acid sequence Asp - Xaa - Ser/Thr, wherein Xaa may be any amino acid except for proline. Likewise, "N-glycans" are glycans attached to asparagine residues of a polypeptide chain. The terms "glycan", "glycan structure", "carbohydrate", "carbohydrate chain" and "carbohydrate structure" are generally used synonymously herein.

N-glycans generally have a common core structure consisting of two N- acetylglucosamine (GlcNAc) residues and three mannose residues, having the structure Manal ,6-(Mana1 ,3-)Man31 ,4-GlcNAc31 ,4-GlcNAc31 -Asp with Asp being the asparagine residue of the polypeptide chain. N-glycans are subdivided into three different types, namely complex-type glycans, hybrid-type glycans and high mannose- type glycans.

The numbers given herein, in particular the relative amounts of a specific glycosylation property, are preferably to be understood as approximate numbers. In particular, the numbers preferably may be up to 10% higher and/or lower, in particular up to 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2% or 1 % higher and/or lower.

A "composition comprising IgA antibodies" may be any composition or substance comprising or consisting of IgA antibodies. Hence, the term "a composition comprising IgA antibodies" may be replaced by the term "IgA antibody" herein. It may be in solid or fluid form and may comprise further ingredients in addition to IgA antibodies. In particular, a composition comprising IgA antibodies may be a solution comprising IgA antibodies and a suitable solvent such as water and/or alcohol, or a powder obtained, for example, after lyophilization of a solution containing IgA antibodies. The composition comprising IgA antibodies preferably comprises a reasonable amount of IgA antibodies, in particular at least 1 fmol, preferably at least 1 pmol, at least 1 nmol or at least 1 μηιοΙ of the IgA antibody. A composition comprising a specific IgA antibody may additionally comprise further antibodies. However, preferably a composition comprising a specific IgA antibody does not comprise other antibodies apart from the specific antibody. In particular, at least 75%, preferably at least 80%, at least 85%, at least 90%, at least 95%, at least 97%, at least 98% or at least 99%, most preferably about 100% of the antibodies in the composition are directed to or bind to the same antigen or epitope. Suitable examples of a composition comprising IgA antibodies are compositions obtained after expression of IgA antibodies in cells, in particular after purification of the IgA antibodies, or pharmaceutical compositions comprising IgA antibodies. A composition comprising IgA antibodies may contain, in addition to IgA antibodies, for example solvents, diluents, excipients, stabilizers, preservatives, salts, adjuvants and/or surfactants.

A "relative amount of 2,6-coupled sialic acid" refers to a specific percentage or percentage range of the total amount of sialic acids being 2,6-coupled sialic acids. A relative amount of 2,6-coupled sialic acid of 100% thus means that all sialic acids are 2,6-coupled sialic acids. For example, a relative amount of 2,6-coupled sialic acids of 60% refers to a composition comprising IgA antibodies wherein 60% of all sialic acids comprised in the IgA antibodies and thus, attached to the oligosaccharide chains of the IgA antibodies in said composition are attached via a 2,6-linkage while 40% of all sialic acids comprised in the IgA antibodies and thus, attached to the oligosaccharide chains of the IgA antibodies in said composition are not attached via a 2,6-linkage, but for example via a 2,3-linkage or a 2,8-linkage. In certain embodiments, the "relative amount of 2,6-coupled sialic acid" refers to a specific percentage or percentage range of the total amount of sialic acids being 2,6-coupled sialic acids in the N-glycosylation, in particular in the complex-type glycan structure of the N-glycosylation.

The terms "cell" and "cells" and "cell line" used interchangeably, preferably refer to one or more mammalian cells, in particular human cells. The term includes progeny of a cell or cell population. Those skilled in the art will recognize that "cells" include progeny of a single cell, and the progeny can not necessarily be completely identical (in morphology or of total DNA complement) to the original parent cell due to natural, accidental, or deliberate mutation and/or change. "Cell" preferably refers to isolated cells and/or cultivated cells which are not incorporated in a living human or animal body.

The term "patient" means according to the invention a human being, a nonhuman primate or another animal, in particular a mammal such as a cow, horse, pig, sheep, goat, dog, cat or a rodent such as a mouse and rat. In a particularly preferred embodiment, the patient is a human being. In case of a human patient, the IgA antibody preferably is human, humanized or chimeric IgA antibody.

The term "pharmaceutical composition" particularly refers to a composition suitable for administering to a human or animal, i.e., a composition containing components which are pharmaceutically acceptable. Preferably, a pharmaceutical composition comprises an active compound or a salt or prodrug thereof together with a carrier, diluent or pharmaceutical excipient such as buffer, preservative and tonicity modifier.

DETAILED DESCRIPTION OF THE INVENTION

The present invention is according to one aspect based on the finding that IgA antibodies having an optimized human glycosylation pattern exhibit improved properties which are desired for therapeutic applications. In particular, the glycosylation pattern as obtained by expression in glycooptimized human blood cell derived cell lines provides IgA antibodies with a high stability. This leads to a very long shelf life of the IgA antibodies. Even after storage for 2 years, IgA antibodies of different subtype such as IgA 1 and lgA2 did not show significant aggregation or fragmentation. This was highly remarkable especially for the lgA2m(1 ) allotype. These antibodies were known for their low stability due to the lack of disulfide bonds between their heavy and light chains. Furthermore, IgA antibodies with the optimized glycosylation pattern are also well suited for therapeutic applications. In particular, they show good antigen and immune receptor binding and demonstrate high efficacy, for example in the inhibition of tumor cell proliferation and the induction of antibody-dependent cellular cytotoxicity (ADCC). The IgA antibodies having the optimized glycosylation pattern were produced in human immortalized blood cell lines of myeloid leukemia origin. These cell lines provided the IgA antibodies in unexpectedly high yield of more than 10 g in a 35 days perfusion process.

In view of these findings, the present invention provides, in a first aspect, a composition comprising IgA antibodies, wherein the IgA antibodies in the composition have an N- glycosylation pattern comprising one or more of the following characteristics:

In certain embodiments, the present invention provides a composition comprising IgA antibodies, wherein the IgA antibodies in the composition have an N-glycosylation pattern comprising one or more of the following characteristics:

(i) a relative amount of complex-type glycans carrying bisecting N- acetylglucosamine (bisGlcNAc) of at least 5%, in particular at least 7% of the total amount of complex-type glycans attached to the IgA antibodies in the composition; and/or

(ii) a relative amount of complex-type glycans carrying N-acetyl neuraminic acid (NeuNAc) of at least 50%, in particular at least 55% of the total amount of complex-type glycans attached to the IgA antibodies in the composition; and/or

(iii) a relative amount of hybrid-type glycans of at least 4%, in particular at least 5% of the total amount of N-glycans attached to the IgA antibodies in the composition.

Preferably, said IgA antibodies are monoclonal IgA antibodies, especially recombinant IgA antibodies and thus, are obtained by recombinant production in a host cell, which preferably is a human host cell. Suitable human host cells which provide a respective glycosylation pattern are described subsequently.

In certain embodiments, the IgA antibodies in the composition have at least two of these characteristics, in particular at least (i) and (ii), at least (i) and (iii), or at least (ii) and (iii). In further embodiments, the IgA antibodies in the composition have all of these characteristics.

The relative amount of complex-type glycans carrying bisGlcNAc may in particular be at least 4% of the total amount of complex-type glycans attached to the IgA antibodies in the composition. Especially, the amount is at least 5%, at least 6%, at least 7%, at least 8%, at least 9% or at least 10%. In certain embodiments, the relative amount of complex-type glycans carrying bisGlcNAc is in the range of from 3% to 35%, in particular from 4% to 30%, from 5% to 27% or from 6% to 25% of the total amount of complex-type glycans attached to the IgA antibodies in the composition.

The relative amount of complex-type glycans carrying NeuNAc may in particular be at least 45% of the total amount of complex-type glycans attached to the IgA antibodies in the composition. Especially, the amount is at least 50%, at least 53%, at least 55%, at least 57%, at least 58% or at least 60%. In certain embodiments, the relative amount of complex-type glycans carrying NeuNAc is in the range of from 40% to 95%, in particular from 45% to 90%, from 50% to 85% or from 55% to 80% of the total amount of complex-type glycans attached to the IgA antibodies in the composition.

The relative amount of hybrid-type glycans may in particular be at least 4% of the total amount of N-glycans attached to the IgA antibodies in the composition. Especially, the amount is at least 4.5%, at least 5%, at least 5.5% or at least 6%. In certain embodiments, the relative amount of hybrid-type glycans is in the range of from 3% to 20%, in particular from 4% to 15%, from 5% to 12.5% or from 5.5% to 10% of the total amount of N-glycans attached to the IgA antibodies in the composition.

In certain embodiments, the N-glycosylation pattern of the IgA antibodies in the composition comprises one or more of the following characteristics:

(i) a relative amount of complex-type glycans carrying fucose of at least 40 of the total amount of complex-type glycans attached to the IgA antibodies in the composition; and/or

(ii) a relative amount of complex-type glycans carrying at least two N-acetyl neuraminic acid (NeuNAc) residues of at least 3% of the total amount of complex-type glycans attached to the IgA antibodies in the composition; and/or (iii) a relative amount of complex-type glycans carrying at least two galactose residues of at least 40% of the total amount of complex-type glycans attached to the IgA antibodies in the composition; and/or

(iv) a relative amount of at least biantennary complex-type glycans of at least 65% of the total amount of complex-type glycans attached to the IgA antibodies in the composition; and/or

(v) a relative amount of 2,6-coupled sialic acid of at least 1 % of the total amount of sialic acids on all N-glycans attached to the IgA antibodies in the composition.

The relative amount of complex-type glycans carrying fucose may in particular be at least 45% of the total amount of complex-type glycans attached to the IgA antibodies in the composition. Especially, the amount is at least 48%, at least 50%, at least 53%, at least 55%, at least 58% or at least 60%. In certain embodiments, the relative amount of complex-type glycans carrying fucose is in the range of from 40% to 95%, in particular from 45% to 90%, from 50% to 87% or from 55% to 85% of the total amount of complex-type glycans attached to the IgA antibodies in the composition.

The relative amount of complex-type glycans carrying at least two NeuNAc residues may in particular be at least 4% of the total amount of complex-type glycans attached to the IgA antibodies in the composition. Especially, the amount is at least 4.5%, at least 5%, at least 5.5% or at least 6%. In certain embodiments, the relative amount of complex-type glycans carrying at least two NeuNAc residues is in the range of from 3% to 30%, in particular from 4% to 25%, from 4.5% to 22% or from 5% to 20% of the total amount of complex-type glycans attached to the IgA antibodies in the composition.

The relative amount of complex-type glycans carrying at least two galactose residues may in particular be at least 45% of the total amount of complex-type glycans attached to the IgA antibodies in the composition. Especially, the amount is at least 47%, at least 50%, at least 52% or at least 55%. In certain embodiments, the relative amount of complex-type glycans carrying at least two galactose residues is in the range of from 40% to 95%, in particular from 45% to 90% or from 50% to 85% of the total amount of complex-type glycans attached to the IgA antibodies in the composition.

The relative amount of at least biantennary complex-type glycans may in particular be at least 68% of the total amount of complex-type glycans attached to the IgA antibodies in the composition. Especially, the amount is at least 70%, at least 72%, at least 74% or at least 76%. In certain embodiments, the relative amount of at least biantennary complex-type glycans is in the range of from 65% to 95%, in particular from 70% to 90% or from 75% to 85% of the total amount of complex-type glycans attached to the IgA antibodies in the composition. The term "at least biantennary" refers to any complex-type glycans which have two or more antennae or branches, such as biantennary, triantennary and tetraantennary complex-type glycan structures. An antenna or branch of a complex-type glycan is any GlcNAc residue attached to one of the two terminal mannose residues of the core structure of an N-glycan. Said GlcNAc residue may carry further saccharide residues such as galactose and sialic acid residues. A bisecting GlcNAc residue which is attached to the central mannose residue of the core structure of N-glycans is not considered an antenna or branch.

The relative amount of 2,6-coupled sialic acid may in particular be at least 5% of the total amount of sialic acids on all N-glycans attached to the IgA antibodies in the composition. Especially, the amount is at least 10%, at least 20%, at least 30%, at least 40%, at least 45% or at least 50%. In certain embodiments, the relative amount of 2,6- coupled sialic acid is in the range of from 10% to 95%, in particular from 25% to 90%, from 40% to 85% or from 50% to 80% of the total amount of sialic acids on all N- glycans attached to the IgA antibodies in the composition.

In preferred embodiments, the IgA antibodies according to the invention do not comprise N-glycolyl neuraminic acids (NeuGc) or detectable amounts of NeuGc. Furthermore, the recombinant IgA antibodies preferably also do not comprise Galili epitopes (Galcd ,3-Gal structures) or detectable amounts of the Galili epitope.

The present invention in particular provides a composition comprising IgA antibodies with a human glycosylation pattern. Due to these glycosylation properties, foreign immunogenic non-human structures which induce side effects are absent which means that unwanted side effects or disadvantages known to be caused by certain foreign sugar structures such as the immunogenic non-human sialic acids (NeuGc) or the Galili epitope (Gal-Gal structures), both known for rodent production systems, or other structures like immunogenic high-mannose structures as known from e.g. yeast systems are avoided.

The IgA antibodies in the composition comprising IgA antibodies in particular are recombinant IgA antibodies. In preferred embodiments, the IgA antibodies in the composition are all directed against the same antigen, and especially have the same amino acid sequence. Especially they are monoclonal IgA antibodies.

The IgA antibodies in the composition according to the invention may be any IgA antibodies of any subtype, such as lgA1 antibodies or lgA2 antibodies. In certain embodiments, the IgA antibodies are lgA2 antibodies, in particular lgA2m(1 ) or lgA2m(2) or lgA2(n) antibodies. The IgA antibodies may be monomeric, i.e. comprise only one antibody molecule which may contain two heavy chains and two light chains. Alternatively, the IgA antibodies may be dimeric, i.e. comprise two antibody molecules linked via a J chain. The dimeric IgA antibodies may further comprise a secretory component. In particular, the IgA antibodies are monomeric or dimeric lgA2m(1 ) antibodies, the dimeric lgA2m(1 ) antibodies comprising a J chain and optionally a secretory component.

The term "IgA antibody" as used herein also includes fragments, fusion constructs and conjugates of IgA antibodies. The IgA antibody fragments preferably comprise one or more of the glycosylation sites of the whole IgA antibody. In certain embodiments, the fragments comprise at least one, especially at least two and in particular all three of the three different heavy chain constant regions of the IgA antibody. In further embodiments, the fragment exhibits the antigen binding activity of the whole IgA antibody. In preferred embodiments, the IgA antibody is a whole IgA antibody. A fusion construct of an IgA antibody comprises an IgA antibody or a fragment thereof wherein to one or more of its polypeptide chains a further polypeptide chain is fused. A conjugate of an IgA antibody is an IgA antibody or a fragment thereof to which a further molecule such as a therapeutic molecule of toxin is covalently attached.

In certain embodiments, the IgA antibody is a therapeutic or diagnostic antibody which can be used in medicine, in particular in the treatment, prophylaxis, diagnosis, prognosis and/or monitoring of a disease, in particular cancer. Suitable examples are IgA antibodies against cancer antigens which target cancer cells. Exemplary cancer antigens include EGFR, HER2, CD20, MUC1 and TF (Thomsen-Friedenreich antigen). The composition comprising IgA antibodies in particular is a pharmaceutical composition. The present invention also provides the composition comprising IgA antibodies as described herein for use in medicine, in particular for use in the treatment of cancer. Furthermore, the present invention provides a method for treatment of cancer in patients in need thereof, comprising administering a therapeutically effective amount of the composition comprising IgA antibodies as described herein to the patient.

In certain embodiments, the IgA antibody is directed against EGFR. In particular, it may be derived from the chimeric IgG antibody Cetuximab and especially may comprise one or more of the CDRs selected from the group consisting of CDRH1 having the amino acid sequence of SEQ ID NO: 1 , CDRH2 having the amino acid sequence of SEQ ID NO: 2, CDRH3 having the amino acid sequence of SEQ ID NO: 3, CDRL1 having the amino acid sequence of SEQ ID NO: 4, CDRL2 having the amino acid sequence of SEQ ID NO: 5, CDRL3 having the amino acid sequence of SEQ ID NO: 6. In particular, the antibody may be an anti-EGFR IgA antibody comprising

(i) a heavy chain variable region having the amino acid sequence of SEQ ID NO:

7 or comprising a CDRH1 having the amino acid sequence of SEQ ID NO: 1 , a CDRH2 having the amino acid sequence of SEQ ID NO: 2 and a CDRH3 having the amino acid sequence of SEQ ID NO: 3; and

(ii) optionally a light chain variable region having the amino acid sequence of SEQ ID NO: 8 or comprising a CDRL1 having the amino acid sequence of SEQ ID NO: 4, a CDRL2 having the amino acid sequence of SEQ ID NO: 5 and a CDRL3 having the amino acid sequence of SEQ ID NO: 6.

Said antibody preferably is capable of binding the same antigen, in particular the same epitope as Cetuximab.

In certain embodiments, the IgA antibody is directed against MUC1 . In particular, it may be derived from the IgG antibody Pankomab and especially may comprise one or more of the CDRs selected from the group consisting of CDRH1 having the amino acid sequence of SEQ ID NO: 9 or 10, CDRH2 having the amino acid sequence of SEQ ID NO: 1 1 or 12, CDRH3 having the amino acid sequence of SEQ ID NO: 13 or 14, CDRL1 having the amino acid sequence of SEQ ID NO: 15 or 16, CDRL2 having the amino acid sequence of SEQ ID NO: 17 or 18, CDRL3 having the amino acid sequence of SEQ ID NO: 19 or 20. In particular, the antibody may be a chimeric or humanized anti-TA-Muc1 IgA-antibody comprising

(i) a heavy chain variable region having the amino acid sequence of SEQ ID NO:

21 or comprising a CDRH1 having the amino acid sequence of SEQ ID NO: 9, a CDRH2 having the amino acid sequence of SEQ ID NO: 1 1 and a CDRH3 having the amino acid sequence of SEQ ID NO: 13; or a heavy chain variable region having the amino acid sequence of SEQ ID NO:

22 or comprising a CDRH1 having the amino acid sequence of SEQ ID NO: 10, a CDRH2 having the amino acid sequence of SEQ ID NO: 12 and a CDRH3 having the amino acid sequence of SEQ ID NO: 14; and

(ii) optionally a light chain variable region having the amino acid sequence of SEQ ID NO: 23 or comprising a CDRL1 having the amino acid sequence of SEQ ID NO: 15, a CDRL2 having the amino acid sequence of SEQ ID NO: 17 and a CDRL3 having the amino acid sequence of SEQ ID NO: 19; or optionally a light chain variable region having the amino acid sequence of SEQ ID NO: 24 or comprising a CDRL1 having the amino acid sequence of SEQ ID NO: 16, a CDRL2 having the amino acid sequence of SEQ ID NO: 18 and a CDRL3 having the amino acid sequence of SEQ ID NO: 20.

Said antibody preferably is capable of binding the same antigen, in particular the same epitope as Pankomab. ln certain embodiments, the IgA antibody is directed against HER2. In particular, it may be derived from the IgG antibody Trastuzumab and especially may comprise one or more of the CDRs selected from the group consisting of CDRH1 having the amino acid sequence of SEQ ID NO: 25, CDRH2 having the amino acid sequence of SEQ ID NO: 26, CDRH3 having the amino acid sequence of SEQ ID NO: 27, CDRL1 having the amino acid sequence of SEQ ID NO: 28, CDRL2 having the amino acid sequence of SEQ ID NO: 29, CDRL3 having the amino acid sequence of SEQ ID NO: 30. In particular, the antibody may be a chimeric or humanized anti-HER2 IgA-antibody comprising

(i) a heavy chain variable region having the amino acid sequence of SEQ ID NO:

31 or comprising a CDRH1 having the amino acid sequence of SEQ ID NO: 25, a CDRH2 having the amino acid sequence of SEQ ID NO: 26 and a CDRH3 having the amino acid sequence of SEQ ID NO: 27; and

(ii) optionally a light chain variable region having the amino acid sequence of SEQ ID NO: 32 or comprising a CDRL1 having the amino acid sequence of SEQ ID NO: 28, a CDRL2 having the amino acid sequence of SEQ ID NO: 29 and a CDRL3 having the amino acid sequence of SEQ ID NO: 30.

Said antibody preferably is capable of binding the same antigen, in particular the same epitope as Trastuzumab.

In certain embodiments, the IgA antibody is directed against TF. In particular, it may be derived from the antibody Karomab and especially may comprise one or more of the CDRs selected from the group consisting of CDRH1 having the amino acid sequence of SEQ ID NO: 33, CDRH2 having the amino acid sequence of SEQ ID NO: 34 or 35, CDRH3 having the amino acid sequence of SEQ ID NO: 36, 37 or 38, CDRL1 having the amino acid sequence of SEQ ID NO: 39, 40 or 41 , CDRL2 having the amino acid sequence of SEQ ID NO: 42 or 43, CDRL3 having the amino acid sequence of SEQ ID NO: 44 or 45. In particular, the antibody may be a chimeric or humanized anti-TF IgA- antibody comprising

(i) a heavy chain variable region having the amino acid sequence of SEQ ID NO:

46 or comprising a CDRH1 having the amino acid sequence of SEQ ID NO: 33, a CDRH2 having the amino acid sequence of SEQ ID NO: 35 and a CDRH3 having the amino acid sequence of SEQ ID NO: 38; or a heavy chain variable region having the amino acid sequence of SEQ ID NO:

47 or comprising a CDRH1 having the amino acid sequence of SEQ ID NO: 33, a CDRH2 having the amino acid sequence of SEQ ID NO: 34 and a CDRH3 having the amino acid sequence of SEQ ID NO: 36; or a heavy chain variable region having the amino acid sequence of SEQ ID NO: 48 or comprising a CDRH1 having the amino acid sequence of SEQ ID NO: 33, a CDRH2 having the amino acid sequence of SEQ ID NO: 34 and a CDRH3 having the amino acid sequence of SEQ ID NO: 37; and

(ii) optionally a light chain variable region having the amino acid sequence of SEQ ID NO: 49 or comprising a CDRL1 having the amino acid sequence of SEQ ID NO: 39, a CDRL2 having the amino acid sequence of SEQ ID NO: 42 and a CDRL3 having the amino acid sequence of SEQ ID NO: 44; or optionally a light chain variable region having the amino acid sequence of SEQ ID NO: 50 or comprising a CDRL1 having the amino acid sequence of SEQ ID NO: 40, a CDRL2 having the amino acid sequence of SEQ ID NO: 42 and a CDRL3 having the amino acid sequence of SEQ ID NO: 45; or optionally a light chain variable region having the amino acid sequence of SEQ ID NO: 51 or comprising a CDRL1 having the amino acid sequence of SEQ ID NO: 40, a CDRL2 having the amino acid sequence of SEQ ID NO: 43 and a CDRL3 having the amino acid sequence of SEQ ID NO: 45.

Said antibody preferably is capable of binding the same antigen, in particular the same epitope as Karomab.

In certain embodiments, the IgA antibody is directed against CD20. In particular, it may be derived from the antibody Obinutuzumab and especially may comprise one or more of the CDRs selected from the group consisting of CDRH1 having the amino acid sequence of SEQ ID NO: 52, CDRH2 having the amino acid sequence of SEQ ID NO: 53, CDRH3 having the amino acid sequence of SEQ ID NO: 54, CDRL1 having the amino acid sequence of SEQ ID NO: 55, CDRL2 having the amino acid sequence of SEQ ID NO: 56, CDRL3 having the amino acid sequence of SEQ ID NO: 57. In particular, the antibody may be a chimeric or humanized anti-CD20 IgA-antibody comprising

(i) a heavy chain variable region having the amino acid sequence of SEQ ID NO:

58 or comprising a CDRH1 having the amino acid sequence of SEQ ID NO: 52, a CDRH2 having the amino acid sequence of SEQ ID NO: 53 and a CDRH3 having the amino acid sequence of SEQ ID NO: 54; and

(ii) optionally a light chain variable region having the amino acid sequence of SEQ ID NO: 59 or comprising a CDRL1 having the amino acid sequence of SEQ ID NO: 55, a CDRL2 having the amino acid sequence of SEQ ID NO: 56 and a CDRL3 having the amino acid sequence of SEQ ID NO: 57. Said antibody preferably is capable of binding the same antigen, in particular the same epitope as Obinutuzumab.

In specific embodiments, the IgA antibodies have an enhanced biological activity. The biological activity of antibodies in this respect includes, for example, antibody- dependent cellular cytotoxicity (ADCC) and antibody-dependent cellular phagocytosis (ADCP). The enhanced biological activity is mainly achieved by the optimized glycosylation pattern, in particular the optimized glycosylation pattern at the Fc part of the antibodies. For example, the ADCC activity of antibodies of the IgA type is mediated by binding of the antibody to Fca-receptors, in particular FcaRI, via its Fc part. FcaRI is expressed on many different immune cells including neutrophils and macrophages and upon activation by an IgA antibody induces an immune response which results in apoptosis or phagocytosis of the target cell bound by the antibody. The binding affinity of the antibody to the Fca-receptor is influenced by the carbohydrates attached to the glycosylation sites at the Fc part of the antibody. Therefore, optimization of the glycosylation pattern on the Fc part of an antibody will result in a stronger FcaRI-binding and thus, in an enhanced ADCC and ADCP activity.

The therapeutic efficacy of antibodies in many cases depends on the induction of cytotoxic effects, in particular ADCC, against the target cells bound by the antibody. Therefore, increasing the ADCC activity of an antibody increases the therapeutic value thereof. For example, the same amount of antibodies administered to a patient will achieve a much higher therapeutic benefit when using antibodies optimized for their ADCC activity. Furthermore, for achieving the same therapeutic effect, a much lower amount of such antibodies has to be administered.

The composition comprising IgA antibodies may be obtainable or obtained by production in a human blood cell line, in particular a human cell line of myeloid leukemia origin. Examples of suitable cell lines are NM-H9D8, NM-H9D8-E6, NM- H9D8-E6Q12 and GT-5s and cell lines derived therefrom or cell lines homologous thereto. The cell lines were deposited under the accession numbers DSM ACC2806 (NM-H9D8; deposited on September 15, 2006), DSM ACC2807 (NM-H9D8-E6; deposited on October 5, 2006), DSM ACC2856 (NM-H9D8-E6Q12; deposited on August 8, 2007), and DSM ACC3078 (GT-5s; deposited on July 28, 2010) according to the requirements of the Budapest Treaty at the Deutsche Sammlung von Mikroorganismen und Zellkulturen (DSMZ), InhoffenstraBe 7B, 38124 Braunschweig (DE) by Glycotope GmbH, Robert-Rossle-Str. 10, 13125 Berlin (DE). Furthermore, these cell lines were derived from K562 cells and were selected for their high sialylation activity. K562 is a human myeloid leukemia cell line present in the American Type Culture Collection (ATCC CCL-243). The above cell lines and cell lines derived therefrom can be cultivated and maintained under the well-known conditions suitable for K562.

A cell line which is derived from NM-H9D8, NM-H9D8-E6, NM-H9D8-E6Q12 or GT-5s 5s can be for example obtained by randomly or specifically selecting a single clone or a group of cells from a culture of these cell lines, optionally after treating the cells in order to enhance their mutation rate, or by genetically altering said cell line. The selected clone or group of cells may further be treated as described above and/or further rounds of selection may be performed. A cell line which is homologous to NM-H9D8, NM- H9D8-E6, NM-H9D8-E6Q12 or GT-5s in particular is an immortalized human myeloid cell line. Cell lines derived from or homologous to NM-H9D8, NM-H9D8-E6, NM-H9D8- E6Q12 or GT-5s in particular have similar glycosylation features to these cell lines and/or are capable of providing IgA antibodies having a glycosylation pattern similar to that obtained from these cell lines. For example, the glycosylation characteristics of IgA antibodies produced in the homologous cell lines do not differ from those of IgA antibodies produced by NM-H9D8, NM-H9D8-E6, NM-H9D8-E6Q12 or GT-5s by more than 25%, in particular not more than 20%, not more than 15% or not more than 10%. This percentage difference refers to the average amount of specific glycosylation features such as those described herein, including the amount of specific monosaccharides such as N-acteylglucosamine, mannose, galactose, fucose and sialic acid in the total glycosylation and/or N-glycosylation, the amount of complex-type, hybrid-type and high mannose-type glycan structures in the N-glycosylation, and in particular the glycosylation features defined in the claims.

It was found that IgA antibodies produced in said cell lines exhibit a glycosylation pattern as described above and in particular exhibit the advantageous therapeutic and pharmacological properties described herein. Thus, the present invention also pertains to a composition comprising IgA antibodies, obtainable or obtained by production in a human blood cell line, in particular a human cell line of myeloid leukemia origin such as a cell of any one of the cell lines NM-H9D8, NM-H9D8-E6, NM-H9D8-E6Q12 and GT- 5s and a cell line derived therefrom or a cell line homologous thereto. Furthermore, the present invention pertains to a method for producing a composition comprising IgA antibodies by recombinantly expressing the IgA antibodies in a human blood cell line, in particular a human cell line of myeloid leukemia origin. The IgA antibodies respectively produced can be isolated and optionally purified.

Thus, the composition comprising IgA antibodies preferably is obtainable by a process comprising the steps of:

(i) cultivating a human host cell of a human blood cell line, in particular a human cell line of myeloid leukemia origin, comprising nucleic acids coding for the IgA antibody under conditions suitable for expression of the IgA antibody; and (ii) isolating the IgA antibody. The isolation of IgA antibodies preferably comprises the further steps of:

(a) obtaining the culture supernatant where the IgA antibodies are secreted by the human host cells, or lysing the human host cells where the IgA antibodies are not secreted; and

(b) isolating the IgA antibodies from the culture supernatant or cell lysate using chromatographic steps such as affinity chromatography, in particular using affinity ligands directed against the light chain of the IgA antibodies.

Preferably, the nucleic acid coding for the heavy chain of the IgA antibody and the nucleic acid coding for the light chain of the IgA antibody are comprised in expression cassettes comprised in a suitable expression vector that allows the expression in a human host cell. The nucleic acid coding for the heavy chain and the nucleic acid coding for the light chain may be comprised in the same vector, but preferably are comprised in separate vectors. Furthermore, they may also be expressed from one expression cassette using appropriate elements such as an IRES element. Preferably, the IgA antibody is secreted by the host cells. In preferred embodiments, cultivation of the host cells is performed in a fermenter and/or under serum-free conditions.

The composition comprising IgA antibodies obtainable by production in a human blood cell line, in particular a human cell line of myeloid leukemia origin, preferably exhibits the features described herein with respect to the composition comprising IgA antibodies according to the present invention. In particular, its glycosylation pattern comprises one or more of the characteristics described above, preferably at least one glycosylation pattern as described in the claims.

In another aspect, the present invention provides a method for increasing the shelf life of IgA antibodies, comprising providing the IgA antibodies with an N-glycosylation pattern comprising one or more of the following characteristics:

(iii) a relative amount of hybrid-type glycans of at least 3% of the total amount of N-glycans attached to the IgA antibodies in the composition. The IgA antibodies in particular have a glycosylation pattern as defined for the composition of IgA antibodies described herein. Furthermore, the embodiments, features and examples described above with respect to the composition comprising IgA antibodies likewise apply to the method for increasing the shelf life of IgA antibodies.

The step of providing the IgA antibodies with an N-glycosylation pattern as defined above in certain embodiments includes producing the IgA antibodies in a cell line which provides a composition comprising the IgA antibodies with said glycosylation pattern. In specific embodiments, the cell line is an immortalized human blood cell line, in particular a human blood cell line of myeloid leukemia origin, such as the cell lines NM- H9D8, NM-H9D8-E6, NM-H9D8-E6Q12 and GT-5s or a cell line derived therefrom.

In certain embodiments, the shelf life of the IgA antibody is increased in comparison to a reference IgA antibody with an N-glycosylation pattern which - in contrast to the IgA antibody with increased shelf life - does not comprising one or more of the characteristic (i), (ii) and (iii). In particular, the N-glycosylation pattern of the reference IgA antibody does not have any one of the characteristic (i), (ii) and (iii). Preferably, the IgA antibody with increased shelf life and the reference antibody have the same amino acid sequence.

An increase in shelf life in particular means that after a specific time of storage the IgA antibody composition with increased shelf life comprises less degradation products of the IgA antibody than a reference antibody composition stored under the same conditions. Degradation products of IgA antibodies include, for example, aggregated IgA antibody and fragments of the initial IgA antibody. A suitable time of storage after which the amount of degradation products is determined is for example, 6 months, 1 year, 1 .5 years, 2 years, 3 years, 4 years or 5 years. Suitable storage conditions include storage at 4^<Ό and storage at room temperature. The composition may comprise, in addition to the IgA antibody, PBS buffer or any other suitable buffer. The shelf life in particular may be determined as described in example 4.

In certain embodiments, the shelf life is increased so that the amount of degradation products of IgA antibodies in the composition comprising IgA antibodies is reduced by at least 5%, preferably at least 10%, at least 15%, at least 20%, at least 25%, at least 33%, or at least 50% after a specific time of storage, in particular after 1 year of storage at 4°C.

Numeric ranges described herein are inclusive of the numbers defining the range. The headings provided herein are not limitations of the various aspects or embodiments of this invention which can be read by reference to the specification as a whole. According to one embodiment, subject matter described herein as comprising certain steps in the case of methods or as comprising certain ingredients in the case of compositions refers to subject matter consisting of the respective steps or ingredients. It is preferred to select and combine specific aspects and embodiments described herein and the specific subject-matter arising from a respective combination of specific embodiments also belongs to the present disclosure.

FIGURES

Figure 1 : IgA antibodies are stable during long-time storage. Long-time stability of monomers and non-separated IgA was investigated by size exclusion chromatography (SEC), (a) hPM lgA2 monomers were generated and after 1 year storage no aggregation or fragmentation was observed. Fractions pooled for monomer generation are indicated (vertical dotted line). After 2 and 1 .5 years storage, comparable SEC profiles were obtained for hPM lgA1 (b) and hTM lgA2 (c), respectively. Slightly different elution volumes were caused by different chromatography systems used for analysis. Qualitatively no differences were observed and monomer contents did not change. For hPM lgA1 t=0 days, the program was paused at about 16.5 ml_ resulting in the artificial spike in the chromatogram.

Figure 2: Inhibition of cancer cell line proliferation by hTM and CM lgA2 and IgG antibodies. Proliferation within 4 to 5 days relative to a control grown in medium without antibody was calculated. Mean values of duplicates are shown, error bars indicate standard deviation, (a) hTM lgA2 and IgG antibodies were comparable potent in proliferation inhibition of SK-BR-3 target cells, (b) CM lgA2 monomers, lgA2 dimers and IgG antibodies were comparable in proliferation inhibition of A-431 target cells. For SK-BR-3 and A-431 , hOM lgA2 and IgG antibodies served as irrelevant matched isotype negative controls.

Figure 3: IgA antibody-mediated activation of granulocytes and monocytes in the presence target cells. Mean values of duplicates are shown, error bars indicate standard deviation. Antibodies and target cells were added to whole blood to investigate the potential of IgA antibodies to mediate effector cell activation. Activation was investigated by measuring the induction of reactive oxygen species (ROS) by flow cytometry. Addition of hPM lgA1 antibody and ZR-75-1 target cells in whole blood resulted in activation of granulocytes and monocytes. With increasing target cell concentrations (indicated by triangles) more granulocytes and monocytes were activated. In the absence of antibody or at low target cell concentrations granulocytes and monocytes were not activated. Granulocytes and monocytes were distinguished by corresponding gates in forward versus sideward scatter plots. Phorbol 12-myristate 13- acetate (PMA) served as positive control for granulocyte and monocyte activation (+). Figure 4: Antibody-dependent cellular cytotoxicity mediated by hPM IgA antibodies against cancer cell lines. Granulocytes isolated from peripheral blood were used as effector cells. IgA antibody concentration-dependent specific lysis of ZR- 75-1 target cells was observed. Mean values of triplicates are shown, error bars indicate standard deviation, (a) Compared to hPM lgA1 , hPM lgA2 resulted in significantly higher maximal lysis. The difference was much more pronounced for IgA antibodies co-expressed with the J chain. hPM lgA2J resulted in about two-times higher maximal lysis compared to hPM lgA1 J. (b) No significant difference in maximal lysis was observed for hPM lgA2 antibodies derived from H9D8 or Fuc^" expression systems.

Figure 5: Antibody-dependent cellular cytotoxicity mediated by hTM lgA2 and IgG antibodies against cancer cell line. Granulocytes isolated from peripheral blood were used as effector cells. Mean values of triplicates are shown, error bars indicate standard deviation. hTM lgA2 and IgG antibodies mediated specific lysis of SK-BR-3 target cells. Maximal lysis was higher for hTM lgA2 than for hTM IgG antibody. Irrelevant serum IgA control antibody served as negative control and did not mediate lysis of SK-BR-3 target cells.

Figure 6: Antibody-dependent cellular cytotoxicity mediated by CM lgA2 and IgG antibodies against cancer cell line. Granulocytes isolated from peripheral blood were used as effector cells. Mean values of triplicates are shown, error bars indicate standard deviation. CM lgA2 mediated lysis of SK-BR-3 target cells while CM IgG did not mediate lysis of SK-BR-3 target cells.

Figure 7: Antibody-dependent cellular cytotoxicity mediated by hKM lgA2 and IgG antibodies against Panc-1 target cells. Granulocytes isolated from peripheral blood were used as effector cells. hKM lgA2 mediated lysis of Panc-1 target cells while hKM IgG did not mediate lysis of Panc-1 target cells. Mean values of triplicates are shown, error bars indicate standard deviation.

Figure 8: Antibody-dependent cellular phagocytosis (ADCP) mediated by hPM and hTM lgA2 antibodies against cancer cell lines. Monocyte-derived macrophages (MDM) were used as effector cells for ADCP assays. Mean values of duplicates are shown, error bars indicate standard deviation, (a) hPM lgA2 mediated phagocytosis of ZR-75-1 and T47D target cell lines, (b) hTM lgA2 mediated phagocytosis of BT-474 target cell line. Irrelevant matched isotype control antibodies did not result in phagocytosis of target cells. EXAMPLES Example 1 : Production of IgA antibodies

A panel of IgA antibodies against five different tumor antigens (anti-HER2 antibody trastuzumab (hTM), anti-EGFR antibody cetuximab (CM), anti-MUC1 antibody pankomab (hPM), anti-CD176 antibody karomab (hKM), and anti-CD20 antibody obinutuzumab (hOM) was successfully expressed in the human myeloid leukemia cell line NM-H9D8. For PM lgA1 , TM lgA2 and CM lgA2, high producing clones were used for subsequent transfection with a plasmid encoding the J chain. In all cases, production yields using serum-free media were sufficient for subsequent purification, biochemical assessment and biofunctional evaluation of IgA antibodies mechanisms of action against cancer cell lines. Moreover, to evaluate feasibility to produce large amounts of monoclonal IgA antibodies, humanized trastuzumab lgA2 was cultivated in a bioreactor under serum-free conditions.

Example 2: High-yield production of hTM lgA2 by cultivation in a bioreactor

One major challenge of therapeutic IgA antibodies is the production of sufficient amounts of recombinant protein. To investigate feasibility of high-yield production under conditions compatible with large-scale good manufacturing practice (GMP) production, humanized trastuzumab lgA2 producing cells were cultivated in a 2 L bioreactor. In the controlled environment using a continuous perfusion process with serum-free medium, high cell densities were reached. Maximal cell concentrations were 5x10⁷ viable cells/mL which is about ten-times higher than maximal cell density in spinner cultures. More importantly, once the culture had reached the maximal perfusion rate, IgA titers reached up to 282 μg mL and the mean titer for the duration of the perfusion process was 200 μg/mL. Therefore, at a perfusion rate of 2 LJd, the process yielded on average 400 mg IgA per day. The culture was stable for 35 days and not affected by a three hour power cut. Viability of cells was higher than 90 % and productivity exceeded 360 mg per day and 1 L working volume for the duration of the cultivation. Cells were removed from the bioreactor by bleeding to avoid cell concentrations exceeding 5x10⁷ viable cells/mL and ensure sufficient nutrient supply for the culture.

This represents the first report of continuous cultivation of an IgA-producing culture in a bioreactor. Feasibility of high yield production of recombinant monoclonal IgA antibodies was confirmed. At maximal productivity, up to 564 mg IgA were produced per day. The total IgA antibody yield was 1 1 g within 35 days cultivation. While this represents an exemplary IgA antibody production in a bioreactor, for standard IgA antibody production 1 L spinner cultures yielded sufficient material to confirm biochemical integrity and elucidate biofunctionality against cancer cell lines. Example 3: IgA antibodies form monomers and multimers

The produced IgA antibodies formed mainly monomers consisting of two heavy chains (HCs) and two light chains (LC). Nevertheless, in the absence of J chain co- expression, IgA antibodies associated to dimers and higher order structures as shown by SEC under native conditions. Among all IgA antibodies without J chain, similar SEC profiles were obtained with a major peak at about 15 ml_ elution volume representing IgA monomers. Between 13 and 10 mL elution volume, dimers and higher order multimers were eluted, respectively. Absolute elution volumes depend on the FPLC system used, qualitatively no differences were observed though.

Clones with high J chain expression showed reduced monomer content in SEC profiles as compared to clones with lower J chain expression. In summary, J chain co- expression promotes the formation of dimers and higher order multimers of IgA antibodies.

Example 4: Purified IgA antibodies are stable during long-time storage

Stability of IgA antibodies is a concern and especially true for lgA2m(1 ) allotype due to the lack of disulfide bonds between LC and HC. Using SEC, stability of IgA antibodies was investigated after long-time storage for up to 2 years. As shown for hPM lgA2 after 1 year storage at 4 ^<C in PBS, monomer preparations were not prone to aggregation or fragmentation, monomer content exceeded 95 % (Figure 1 a). Affinity-purified hPM lgA1 and hTM lgA2 showed comparable SEC profiles after 2 and 1 .5 years storage, respectively (Figure 1 b and c). Long-time stability during storage at 4 ^<C in PBS of both, IgA monomers and non-separated IgA antibodies, was confirmed. This is in contrast to studies which showed dissociation and fragmentation of lgA2J dimers after 1 year storage (Lohse et al. (2012) The Journal of Biological Chemistry 287(30): 25139-25150).

Example 5: Human N-glycosylation profiles of IgA antibodies

Glycosylation impacts stability, biological function and clearance of antibodies. IgA antibodies possess a more complex N-glycosylation than IgG antibodies and IgA N- glycans point towards to outside of the molecule. lgA1 antibodies have 2 N- glycosylation sites and lgA2m(1 ) antibodies have 4 N-glycosylation sites per heavy chain as compared to 1 N-glycosylation site for IgG antibodies. Depending on the amino acid sequence, IgA or IgG antibodies might have additional N-glycosylation sites within the variable domains as in the case of hPM and CM antibodies. Here the N-glycan profile for hPM lgA1 and hTM lgA2 was to evaluate predominant glycan structures and the degree of sialylation on recombinant IgA antibodies produced by a human expression system.

For the two IgA antibodies analyzed, complex N-glycosylation profiles were obtained by hydrophilic-interaction liquid chromatography (HILIC) of liberated and labeled N- glycans. All three N-glycan types which can be attached to proteins were identified: complex, oligomannose and hybrid type.

Table 1 : Three types of N-glycans were detected on IgA antibodies.

Relative molar abundance of glycan types [%]

Antibody Complex type Oligomannose Hybrid type hPM IgAU 69 16 12 hTM lgA2 64 25 8

N-glycans did not add up to 100 % because some N-glycans could not be annotated for hPM lgA1 and hTM lgA2, respectively. Percentages represent fractions of total peak area of in HLIC chromatograms of liberated and labeled N-glycans.

Table 2: Relative abundance of monosaccharide units in biantennary complex type N-glycans.

Relative molar abundance of glycan structures [%]

Antibody F B SO S>0 GO G>0 hPM IgAU 49 27 19 77 2 94 hTM lgA2 64 15 32 62 3 91

F: fucose, B: bisecting N-acetylglucosamine, S: sialic acid, G: galactose, numbers indicate monomer units per N-glycan

Furthermore, also the glycosylation pattern of other lgA2 antibodies were determined, using the same analysis technology as described above. In the following tables the results of the different lgA2 antibodies are summarized: Table 3: Three types of N-glycans were detected on IgA antibodies.

Relative molar abundance of glycan types [%]

Complex type Oligomannose Hybrid type

Range 64-74 17-26 6-9

Average 67 22 7

Percentages represent fractions of total peak area of in HLIC chromatograms of liberated and labeled N-glycans.

Table 4: Relative abundance of monosaccharide units in biantennary complex type N-glycans.

Relative molar abundance of glycan structures [%]

F B S>0 S>1 G>1 A>1

64-84 7-15 62-68 6-15 56-82 85-94

76 1 1 64 1 1 71 90

F: fucose, B: bisecting N-acetylglucosamine, S: sialic acid, G: galactose, A: antennarity, numbers indicate monomer units per N-glycan

In summary, human N-glycosylation was confirmed for different lgA1 and lgA2 antibodies. Recombinant IgA antibodies showed complex, oligomannose and hybrid type N-glycans. As expected for a human expression system, non-human glycan structures like a(1 -3)-linked galactose or a(2-3)-linked N-glycolylneuraminic acid were not detected. This is in contrast to other expression systems used for the production of therapeutic antibodies.

Example 6: IgA antibodies bind their corresponding antigen and target cell lines

Target binding of IgA antibodies was investigated by ELISA, immunofluorescence, flow cytometry and surface plasmon resonance. Antigen and target cell binding was a prerequisite for subsequent evaluation of IgA antibodies in biofunctional assays. Different cancer cell lines were used depending on the antigens present on the surface of a target cell and the antibody to be tested. Table 5: Overview of cancer cell lines used in different assays.

Antibodies binding cell line

Cell line Origin (antigen binding sites) Assays using cell line

Epidermoid

A-431 CM (++) Proliferation inhibition carcinoma

Proliferation inhibition,

BT-474 Breast cancer hTM (+++)

ADCP

Ls174T Colon cancer hTM (+), CM (+) ADCC hTM (+), CM (++), hKM (+),

Panc-1 Pancreas cancer ADCC

hPM (++)

Raji Burkitt's lymphoma hKM (+), hOM (n.d.) Flow cytometry

SK-BR-3 Breast cancer hTM (+++), CM ADCC

T47D Breast cancer hPM (+++) ADCP

Flow cytometry, IF,

ZR-75-1 Breast cancer hPM (+++)

ADCC, ADCP

IF: immunofluorescence; ADCC: antibody-dependent cellular cytotoxicity; ADCP: antibody-dependent cellular cytotoxicity; antigen binding sites were determinded using IgG antibodies; antigen binding sites, +: 1 x10⁴-1 x10⁵; ++ 1 x10⁵-5x10⁵; +++ >5x10⁵; n.d.: not determined

All antibodies retained antigen binding when switching to IgA isotype. Moreover, specific binding of hPM lgA1 to antigen-positive cells was confirmed by immunofluorescence microscopy. For hPM, hKM and hOM IgA antibodies, binding to target cell was shown by flow cytometry. For hTM and CM lgA2 antigen binding kinetic was investigated by surface plasmon resonance and target cell binding was indirectly confirmed by biofunctional assays (e.g. inhibition of target cell proliferation).

Example 7: IgA dimers show increased antigen binding avidity compared to IgG antibodies

Surface plasmon resonance was used to compare antigen binding kinetics of IgG and IgA antibodies. Her2 or EGFR were immobilized on CM5 chips, to evaluate binding of hTM or CM antibodies, respectively. The EGFR chip was also used to compare binding kinetics of CM lgA2 monomers and lgA2J dimers.

Binding of hTM lgA2 monomers to Her2 immobilized on a CM5 chip was shown by kinetic analysis. Injections of multiple antibody concentrations revealed comparable binding kinetics for hTM lgA2 monomers and IgG antibodies. Both isotypes showed rapid association within 5 minutes, followed by slow dissociation rates within 15 minutes. Equilibrium dissociation constants for IgA and IgG were 0.67 nM and 0.16 nM, respectively. Both, hTM lgA2 and IgG showed high avidity sub-nanomolar dissociation constants within a comparable range.

As for hTM, CM lgA2 monomers bound their antigen in a similar concentration range as CM IgG. Injections of multiple antibody concentrations revealed comparable binding kinetics for CM lgA2 monomers and IgG antibodies. Equilibrium dissociation constants for lgA2 and IgG were 47 nM and 20 nM, respectively, indicating high avidity binding in low nanomolar ranges. CM lgA2J dimers had an about ten to 23 times lower equilibrium dissociation constants compared to CM lgA2 monomers, 2-6 nM compared to 47 nM, respectively. It was shown that binding of monomeric antibodies, either lgA2 or IgG, results in higher dissociation constants compared to dimeric lgA2 antibodies. This indicates an increase in avidity which could be attributed to higher valence of dimeric antibodies.

Example 8: Recombinant IgA antibodies bind Fc alpha receptor I

Binding to FcaRI is a prerequisite for Fc-mediated effector functions of IgA antibodies. Hence, binding of IgA antibodies to commercially available, recombinant FcaRI was investigated by ELISA. Concentration-dependent binding was shown for both, monomers and multimer-containing IgA preparations. Higher signals were obtained for IgA multimers than for monomers using same mass concentrations (data not shown). This could be caused by increased Fc avidity or more binding sites for secondary detection antibody on multimeric IgA antibodies. Binding of CM lgA2 monomers to FcaRI was shown. Controls lacking FcaRI or IgA antibody were negative. In summary, antibody concentration-dependent binding to the main IgA Fc receptor was confirmed for the recombinant monoclonal IgA antibodies.

Example 9: IgA antibodies inhibit cancer cell line proliferation

Inhibition of target cell proliferation in the presence of hTM and CM IgA or IgG antibodies targeting EGFR and Her2, respectively, was investigated. hTM lgA2 and hTM IgG showed comparable proliferation inhibition of SK-BR-3 target cells within 5 d relative to a control grown in medium without antibody. hTM lgA2 and IgG resulted in minimal proliferation of 59 % and 51 %, respectively (Figure 2a). The antimitotic agent aphidicolin which inhibits DNA synthesis served as positive control for proliferation inhibition. Irrelevant hOM lgA2 and IgG matched isotype negative control antibodies did not affect proliferation of SK-BR-3 target cells.

Neither hTM lgA2 or IgG inhibited proliferation of slow-growing BT-474 target cells. Since proliferation inhibition is measured relative to a control grown in medium without antibody, differences in proliferation are more pronounced for faster growing cell lines (data not shown).

CM lgA2 and IgG antibodies were comparable in proliferation inhibition of A-431 target cells within 4 d relative to a control grown in medium without antibody. Minimal proliferation was about 56 % for CM lgA2, CM lgA2J dimers, CM lgA2J multimers or CM IgG antibodies. There was no difference in proliferation inhibition between CM lgA2 monomers and lgA2J dimers using same mass concentrations. Irrelevant hOM lgA2 and IgG matched isotype negative control antibodies did not affect proliferation of A- 431 target cells. In the presence of 10 μΜ aphidicolin, 5 % proliferation of target cells was observed (Figure 2b).

In summary, as shown for both isotypes, lgA2 and IgG, targeting Her2 and EGFR inhibited proliferation of corresponding target cells. IgG and lgA2 construct were comparable effective in this Fab-mediated mechanism of action against cancer cell lines. As shown for CM lgA2 antibodies, proliferation inhibition was not affected by monomer, multimer or dimer contents.

Example 10: IgA antibody-mediated activation of effector cells in human blood

Neutrophils, basophils and eosinophils constitute different cell populations in circulation belonging to the group of granulocytes. Upon activation granulocytes release cytotoxic granules which include reactive oxygen species (ROS). Like granulocytes, monocytes express the IgA Fc-binding receptor FcaRI which, upon activation, mediates cellular cytotoxicity. The potential of IgA antibodies to activate granulocytes and monocytes as effector cells in whole blood was investigated by measuring ROS using flow cytometry.

In the presence of ZR-75-1 target cells, hPM lgA1 induced reactive oxygen production in granulocytes and monocytes. For 5 x 10⁶ target cells per milliliter, 100 μg mL hPM lgA1 resulted in activation of up to 17 % and 21 % for granulocytes and monocytes, respectively. In the absence of antibody or when using target cell concentrations of 5 x 10⁴ or below, activation was <4 % and <8 % for granulocytes and monocytes, respectively. Phorbol 12-myristate 13-acetate (PMA) was used as positive control for induction of ROS. Treatment of effector cells with PMA resulted in activation of 100 % granulocytes and monocytes. Antibody concentration-dependent granulocyte activation was indicated, 100 and 10 μ9/ηιί hPM lgA1 resulted in 17 % and 14 % activation, respectively. Results for monocyte activation were qualitatively comparable (Figure 3).

As shown for hPM lgA1 in whole blood, IgA antibodies mediated the activation of granulocytes and monocytes in the presence of target cells.

Example 11 : lgA2 antibodies are more potent than lgA1 antibodies in antibody- dependent cellular cytotoxicity

While proliferation inhibition is mediated by the interaction of the Fab domain of an antibody with its target on the surface of target cell, antibody-dependent cellular cytotoxicity (ADCC) is mediated by the Fc part of an antibody while the antibody is bound to a target cell. hPM IgA antibody-mediated cytotoxicity against ZR-75-1 target cells was investigated in granulocyte ADCC experiments. IgA antibody concentration-dependent specific lysis of target cells was confirmed. For IgA antibodies without J chain, hPM lgA2 resulted in higher maximal lysis than hPM lgA1 . However, the difference was not as pronounced as for hPM lgA2J and hPM IgAU which were generated by co-expression of the J chain. hPM lgA2J resulted in more than twice as high maximal lysis as compared to hPM IgAU, 30 and 1 1 %, respectively (Figure 4a). As shown for hPM lgA2 antibodies, the production cell line did not affect ADCC potency. Maximal lysis was comparable for H9D8-derived or Fuc -derived hPM lgA2 (Figure 4b). Irrelevant serum IgA antibody negative control did not mediate lysis of ZR-75-1 target cells.

Example 12: lgA2 antibodies mediate antibody-dependent cellular cytotoxicity against cancer cell lines

Apart from hPM IgA antibodies targeting TA-MUC1 , three additional IgA antibodies targeting different tumor-associated antigens were tested in ADCC assays using granulocytes as effector cells. Biofunctionality of hTM, CM and hKM lgA2 antibodies was investigated using corresponding target cell lines. hTM lgA2 was effective in mediating cytotoxicity against SK-BR-3 target cells. Higher concentrations of hTM lgA2 were needed to reach comparable specific lysis as for hTM IgG. However, maximal specific lysis was higher for hTM lgA2 than for hTM IgG. Specificity was confirmed using serum IgA as irrelevant matched isotype negative control which did not mediate lysis of SK-BR-3 target cells (Figure 5).

CM lgA2 was tested in granulocyte ADCC using SK-BR-3 as target cells. CM lgA2 mediated lysis of SK-BR-3 target cells at concentrations exceeding 10 μg mL while CM IgG did not mediate lysis of SK-BR-3 target cells (Figure 6). ADCC activity of hKM lgA2 antibody was confirmed for one cancer cell line. hKM lgA2 effectively mediated lysis of Panc-1 target cells at antibody concentrations exceeding 1 μg mL and hKM IgG mediated marginal target cell lysis, 22 and 5 %, respectively (Figure 7).

Similarly, hPM lgA2 mediated lysis of Panc-1 target cells while no lysis was observed for hPM IgG (data not shown).

In summary, the potential of one lgA1 antibody and four lgA2 antibodies with different target cells to mediate cellular cytotoxicity was shown. In conclusion, recombinant monoclonal IgA antibodies recruited granulocytes to elicit cytotoxic effects against cancer cell lines. Due to the high abundance of granulocytes in circulation, this represents a promising strategy for cancer immunotherapy.

Example 13: lgA2 antibodies mediate antibody-dependent cellular phagocytosis of cancer cell lines

Antibodies targeting tumor antigens mediate phagocytosis of cancer cell lines by different effector cells including granulocytes, dendritic cells and macrophages. In addition to granulocytes, macrophages represent an interesting effector cell population which is found in large numbers within the tumor microenvironment. Therefore, as a second Fc-mediated effector function for IgA antibodies, antibody-dependent cellular phagocytosis (ADCP) of cancer cell lines was investigated using monocyte-derived macrophages (MDM) as effector cells.

For ADCP assays, CD45 was used as marker for MDM detection and target cells were stained with PKH26. Using flow cytometry, phagocytosis was calculated based on CD45/PKH26 double-positive cells relative to total CD45-positive cells. Unspecific phagocytosis for controls with irrelevant matched isotype control antibody or without antibody in the range of 5 to 20 % was observed in ADCP assays. Therefore, phagocytosis relative to controls without antibody or irrelevant matched isotype control antibodies was evaluated. Unless otherwise noted, M1 phenotype macrophages were used for ADCP assays. ADCP was shown for hPM lgA2 and hTM lgA2 using corresponding target cells and MDM as effector cells. hPM lgA2 mediated phagocytosis of ZT-75-1 and T47D target cells. For hPM lgA2 antibodies, phagocytosis was observed at concentrations exceeding 25 μg mL for ZR- 75-1 and T47D. Maximal phagocytosis of ZR-75-1 target cells by hPM lgA2 was 10 %. Maximal phagocytosis of T47D target cells by hPM lgA2 was 29 % (Figure 8a). hTM lgA2 mediated phagocytosis of BT-474 target cells. Maximal phagocytosis for hTM lgA2 monomers was 46 %. hTM lgA2 preparations were fractionated by SEC to obtain monomers and multimers. Higher concentrations were needed to mediate phagocytosis using multimers compared to monomers. Moreover, maximal lysis was higher for hTM lgA2 monomer preparations compared to multimers (Figure 8b).

In conclusion, hPM and hTM lgA2 antibodies mediated phagocytosis of corresponding target cells. Apart from proliferation inhibition and ADCC, phagocytosis was shown to be another mechanism of action of tumor antigen-targeting IgA antibodies.

Identification of the deposited biological material

The cell lines DSM ACC2806, DSM ACC2807, DSM ACC2856 and DSM ACC3078 were deposited at the DSMZ - Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH, InhoffenstraBe 7B, 38124 Braunschweig (DE) by Glycotope GmbH, Robert-Rossle-Str. 10, 13125 Berlin (DE) on the dates indicated in the following table.

Name of the Accession Depositor Date of Deposition Cell Line Number

NM-H9D8 DSM ACC2806 Glycotope GmbH September s, 2006

NM-H9D8-E6 DSM ACC2807 Glycotope GmbH October 5, 2006

NM-H9D8- DSM ACC2856 Glycotope GmbH August 8, 2007

E6Q12

GT-5S DSM ACC3078 Glycotope GmbH July 28, 2010

Claims

1 . A composition comprising IgA antibodies, wherein the IgA antibodies in the composition have an N-glycosylation pattern comprising one or more of the following characteristics:

2. The composition according to claim 1 , wherein the N-glycosylation pattern comprises at least two of the features (i), (ii) and (iii), and preferably all of the features (i), (ii) and (iii).

3. The composition according to claim 1 or 2, wherein the N-glycosylation pattern of the IgA antibodies in the composition comprises a relative amount of complex-type glycans carrying bisecting N-acetylglucosamine (bisGlcNAc) of at least 5%, in particular at least 7% of the total amount of complex-type glycans attached to the IgA antibodies in the composition.

4. The composition according to any one of claims 1 to 3, wherein the N-glycosylation pattern of the IgA antibodies in the composition comprises a relative amount of complex-type glycans carrying N-acetyl neuraminic acid (NeuNAc) of at least 50%, in particular at least 55% of the total amount of complex-type glycans attached to the IgA antibodies in the composition.

5. The composition according to any one of claims 1 to 4, wherein the N-glycosylation pattern of the IgA antibodies in the composition comprises a relative amount of hybrid-type glycans of at least 4%, in particular at least 5% of the total amount of N-glycans attached to the IgA antibodies in the composition.

6. The composition according to any one of claims 1 to 5, wherein the N-glycosylation pattern of the IgA antibodies in the composition comprises one or more of the following characteristics: (i) a relative amount of complex-type glycans carrying fucose of at least 40%, in particular at least 50% or at least 55% of the total amount of complex-type glycans attached to the IgA antibodies in the composition; and/or

(ii) a relative amount of complex-type glycans carrying at least two N-acetyl neuraminic acid (NeuNAc) residues of at least 3%, in particular at least 5% or at least 6% of the total amount of complex-type glycans attached to the IgA antibodies in the composition; and/or

(iii) a relative amount of complex-type glycans carrying at least two galactose residues of at least 40%, in particular at least 45% or at least 50% of the total amount of complex-type glycans attached to the IgA antibodies in the composition; and/or

(iv) a relative amount of at least biantennary complex-type glycans of at least 70%, in particular at least 75% or at least 80% of the total amount of complex- type glycans attached to the IgA antibodies in the composition; and/or

(v) a relative amount of 2,6-coupled sialic acid of at least 5% of the total amount of sialic acids on all N-glycans attached to the IgA antibodies in the composition.

7. The composition according to any one of claims 1 to 6, wherein the IgA antibodies are lgA2 antibodies, in particular lgA2m(1 ) or lgA2m(2) antibodies.

8. The composition according to claim 7, wherein the IgA antibodies are monomeric or dimeric lgA2m(1 ) antibodies.

9. A method for increasing the shelf life of IgA antibodies, comprising providing the IgA antibodies with an N-glycosylation pattern comprising one or more of the following characteristics:

10. The method of claim 9, comprising producing the IgA antibodies in a cell line which provides a composition comprising the IgA antibodies with said glycosylation pattern.

1 1 . The method of claim 10, wherein the cell line is an immortalized human blood cell line, in particular a human blood cell line of myeloid leukemia origin, such as the cell lines NM-H9D8, NM-H9D8-E6, NM-H9D8-E6Q12 and GT-5s or a cell line derived therefrom.