Chemistry: natural resins or derivatives; peptides or proteins; – Proteins – i.e. – more than 100 amino acid residues – Blood proteins or globulins – e.g. – proteoglycans – platelet...
Reexamination Certificate
2000-04-27
2003-12-30
Helms, Larry R. (Department: 1642)
Chemistry: natural resins or derivatives; peptides or proteins;
Proteins, i.e., more than 100 amino acid residues
Blood proteins or globulins, e.g., proteoglycans, platelet...
C530S387100, C530S391300, C435S188000
Reexamination Certificate
active
06670453
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to multivalent and multispecific antigen binding proteins, methods for their production and uses thereof. In particular, the invention relates to antigen binding proteins comprising a polypeptide comprising in series two or more single domain binding units which are preferably variable domains of a heavy chain derived from an immunoglobulin naturally devoid of light chains.
BACKGROUND OF THE INVENTION
Antibodies are protein molecules belonging to a group of immunoglobulins generated by the immune system in response to an antigen. The structure of most antibody molecules is based on a unit comprising four polypeptides, two identical heavy chains and two identical light chains, which are covalently linked together by disulphide bonds. Each of these chains is folded in discrete domains. The C-terminal regions of both heavy and light chains are conserved in sequence and are called the constant regions, comprising one or more so-called C-domains. The N-terminal regions of the heavy and light chains, also known as V-domains, are variable in sequence and determine the specificity of the antibody. The regions in the variable domains of the light and heavy chains (V
L
and V
H
respectively) responsible for antigen binding activity are known as the hypervariable or complementarity determining regions (CDR).
Natural antibodies generally have at least two identical antigen-binding sites defined by the association of the heavy and light chain variable regions. Individual heavy or light chain domains having the capability to bind antigens have been described in the literature (Ward et al, Nature 341 (1989), 544-546) although generally most naturally occurring antibodies need both a V
H
and V
L
to form a complete antigen binding site and retain full immunoreactivity.
More recently, immunoglobulins capable of exhibiting the functional properties of the four-chain immunoglobulins described above but which comprise two heavy polypeptide chains and which furthermore are devoid of light polypeptide chains have been described (see European Patent Application EP-A-0584421, Casterman et al, 1994). Methods for the preparation of such antibodies or fragments thereof on a large scale comprising transforming a mould or yeast with an expressible DNA sequence encoding the antibody or fragment are described in patent application WO 94/25591 (Unilever).
The immunoglobulins described in EP-A-0584421, which may be isolated from the serum of Camelids, do not rely upon the association of heavy and light chain variable domains for the formation of the antigen-binding site but instead the heavy polypeptide chains alone naturally form the complete antigen binding site. These immunoglobulins, hereinafter referred to as “heavy-chain immunoglobulins” are thus quite distinct from the heavy chains obtained by the degradation of common (four-chain) immunoglobulins or by direct cloning which contribute part only of the antigen-binding site and require a light chain partner for antigen-binding, thus forming a complete antigen binding site.
As described in EP-A-0584421, heavy chain immunoglobulin V
H
regions isolated from Camelids (forming a complete antigen binding site and thus constituting a single domain binding site) differ from the V
H
regions derived from conventional four-chain immunoglobulins in a number of respects, notably in that they have no requirement for special features for facilitating interaction with corresponding light chain domains. Thus, whereas in common (four-chain) immunoglobulins the amino acid residues at the positions involved in the V
H
/V
L
interaction is highly conserved and generally apolar leucine, in Camelid derived V
H
domains this is replaced by a charged amino acid, generally arginine. It is thought that the presence of charged amino acids at this position contributes to increasing the solubility of the camelid derived V
H
. A further difference which has been noted is that one of the CDRs of the heavy chain immunoglobulins of EP-A-0584421, the CDR
3
, may contain an additional cysteine residue associated with a further additional cysteine residue elsewhere in the variable domain. It has been suggested that the establishment of a disulphide bond between the CDR
3
and the remaining regions of the variable domain could be important in binding antigens and may compensate for the absence of light chains.
In the search for multivalent and multispecific antigen binding proteins, attention has been directed towards the use of fragments, or portions, of a whole antibody which can nevertheless exhibit antigen binding activity. By comparison with the whole antibody, the smaller antibody fragment is advantageous for use in therapy, for example, as it is likely to be less immunogenic and more able to penetrate tissue.
Binding fragments of common (four-chain) antibodies which have been considered include Fab (light chain associated with the V
H
and C
H1
domains of a heavy chain), F
V
(comprising of the V-domains of the heavy and light chains associated with each other) and ScFv (comprising a V
H
domain linked to a V
L
domain by a flexible peptide linker) fragments. These fragments have only one site for antigen binding compared to the two or more sites contained in the whole antibody, however, and in an attempt to overcome this problem, recombinant fragments having two or more binding sites have been proposed.
In general, those multivalent and/or multispecific constructions which have been described in the literature either comprise two or more polypeptide chains, see for example, patent application WO 94/09131 (Scotgen Limited) and WO 97/14719 (Unilever) or are based on a ‘double ScFv’ approach, wherein the multivalency arises when two or more monovalent ScFv molecules are linked together, providing a single chain molecule comprising at least four variable domains, as described, for example, in WO 93/11161 (Enzon Inc) and WO 94/13806 (Dow Chemical Co). In all of these cases, the binding site is formed through the association of light and heavy chain variable domains. In WO 93/11161, reference is made to a single-chain protein comprising the binding portions of the variable regions of an antibody light (or heavy) chain but it is stated that as such proteins are comprised of two similar variable regions, they do not necessarily have any antigen-binding capability.
EP-A-0584421 (Casterman), referred to above, discloses fragments of heavy chain immunoglobulins devoid of light chains, including fragments corresponding to isolated V
H
domains or to V
H
dimers linked by the hinge disulphide. Further disclosed, but not exemplified, are antibodies having different specificities on each heavy polypeptide chain which could be prepared by combining two heavy chain immunoglobulins or one heavy chain of an immunoglobulin of EP-A-0584421 with a fragment of a conventional four-chain immunoglobulin. There is no suggestion that multivalent and/or multispecific constructs may be prepared by joining together individual V
H
domains. Indeed, in the absence of the inherent conformational constraints conferred on the position of the binding sites by the presence of a corresponding light chain, it might generally be expected that the binding domains in constructs of this type would sterically hinder each other, unfavourably influencing binding activity.
SUMMARY OF THE INVENTION
In a first aspect, the invention provides a multivalent antigen binding protein comprising a single polypeptide chain comprising, connected in series, two or more single domain binding units.
In another aspect, the invention provides nucleotide sequences coding for multivalent antigen binding proteins according to the invention and cloning and expression vectors comprising such nucleotide sequences. Further provided are host cells transformed with vectors comprising such nucleotide sequences and methods of producing antigen binding proteins according to the invention by expression of the nucleotide sequences in such hosts.
The invention also provides compositions comprising multi
Frenken Leon Gerardus
Howell Steven
Ledeboer Adrianus Marinus
van der Logt Cornelis P
Helms Larry R.
Kokulis Paul N.
McLeod Bonnie Weiss
Unilever Patent Holdings B.V.
LandOfFree
Multivalent antigen-binding proteins does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Multivalent antigen-binding proteins, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Multivalent antigen-binding proteins will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3096527