DK2190863T3

DK2190863T3 - New albumin binding compositions, methods and uses

Info

Publication number: DK2190863T3
Application number: DK08786222.3T
Authority: DK
Inventors: Lars Abrahmsen; Andreas Jonsson; Per-Ake Nygren; Jakob Dogan
Original assignee: Affibody Ab
Priority date: 2007-07-31
Filing date: 2008-07-17
Publication date: 2015-11-30
Also published as: US8937153B2; CN101765608A; DE08786222T1; CN104710518A; US20100273979A1; EP2190863B1; EP2190863A2; JP2010534486A; CA2694139A1; US20150158916A1; CA2694139C; EP2546261A3; HK1211040A1; ES2346178T1; CN101765608B; ES2346178T3; JP5718638B2; WO2009016043A3; WO2009016043A2; US10118949B2

Claims

DESCRIPTION Field of the invention [0001] The present invention relates to a class of engineered polypeptides having a binding affinity for albumin. It also relates to new methods and uses that exploit binding by these and other compounds to albumin in different contexts, some of which have significance for the treatment of disease in mammals including humans. Background Serum albumin [0002] Serum albumin is the most abundant protein in mammalian sera (40 g/l; approximately 0.7 mM in humans), and one of its functions is to bind molecules such as lipids and bilirubin (Peters T, Advances in Protein Chemistry 37:161, 1985). The half-life of serum albumin is directly proportional to the size of the animal, where for example human serum albumin (HSA) has a half-life of 19 days and rabbit serum albumin has a half-life of about 5 days (McCurdy TR et al, J Lab Clin Med 143:115, 2004). Human serum albumin is widely distributed throughout the body, in particular in the intestinal and blood compartments, where it is mainly involved in the maintenance of osmolarity. Structurally, albumins are single-chain proteins comprising three homologous domains and totaling 584 or 585 amino acids (Dugaiczyk L et al, Proc Natl Acad Sci USA 79:71, 1982 ). Albumins contain 17 disulfide bridges and a single reactive thiol, C34, but lack N-linked and O-linked carbohydrate moieties (Peters, 1985, supra; Nicholson JP et al, Br J Anaesth 85:599, 2000). The lack of glycosylation simplifies recombinant expression of albumin. This property of albumin, together with the fact that its three-dimensional structure is known (He XM and Carter DC, Nature 358:209 1992), has made it an attractive candidate for use in recombinant fusion proteins. Such fusion proteins generally combine a therapeutic protein (which would be rapidly cleared from the body upon administration of the protein per se) and a plasma protein (which exhibits a natural slow clearance) in a single polypeptide chain (Sheffield WP, Curr Drug Targets Cardiovacs Haematol Disord 1:1, 2001 ). Such fusion proteins may provide clinical benefits in requiring less frequent injection and higher levels of therapeutic protein in vivo. Fusion or association with HSA results in increased in vivo half-life of proteins [0003] Serum albumin is devoid of any enzymatic or immunological function and, thus, should not exhibit undesired side effects upon coupling to a bioactive polypeptide. Furthermore, HSA is a natural carrier involved in the endogenous transport and delivery of numerous natural as well as therapeutic molecules (Sellers EM and Koch-Weser MD, Albumin Structure, Function and Uses, eds Rosenoer VM et al, Pergamon, Oxford, p 159, 1977). Several strategies have been reported to either covalently couple proteins directly to serum albumins or to a peptide or protein that will allow in vivo association to serum albumins. Examples of the latter approach have been described e.g. in WO91/01743, in WO01/45746 and in Dennis et al, J Biol Chem 277:35035-43 (2002). The first document describes inter alia the use of albumin binding peptides or proteins derived from streptococcal protein G (SpG) for increasing the half-life of other proteins. The idea is to fuse the bacterially derived, albumin binding peptide/protein to a therapeutically interesting peptide/protein, which has been shown to have a rapid clearance in blood. The thus generated fusion protein binds to serum albumin in vivo, and benefits from its longer half-life, which increases the net half-life of the fused therapeutically interesting peptide/protein. WO01/45746 and Dennis et al relate to the same concept, but here, the authors utilize relatively short peptides to bind serum albumin. The peptides were selected from a phage displayed peptide library. In Dennis etal, earlier work is mentioned in which the enhancement of an immunological response to a recombinant fusion of the albumin binding domain of streptococcal protein G to human complement receptor Type 1 was found. US patent application published as US2004/0001827 (Dennis) also discloses the use of constructs comprising peptide ligands, again identified by phage display technology, which bind to serum albumin and which are conjugated to bioactive compounds for tumor targeting. Association with HSA results in decreasedimmunogenicity [0004] In addition to the effect on the in vivo half-life of a biologically active protein, it has been proposed that the non-covalent association with albumin of a fusion between a biologically active protein and an albumin binding protein acts to reduce the immune response to the biologically active protein. Thus, in W02005/097202, there is described the use of this principle to reduce or eliminate the immune response to a biologically active protein. Albumin binding domains of bacterial receptor proteins [0005] Streptococcal protein G (SpG) is a bi-functional receptor present on the surface of certain strains of streptococci and is capable of binding to both IgG and serum albumin (Bjorck et al, Mol Immunol 24:1113, 1987). The structure is highly repetitive with several structurally and functionally different domains (Guss et al, EMBO J 5:1567, 1986), more precisely three lg-binding motifs and three serum albumin binding domains (Olsson et al, Eur J Biochem 168:319, 1987). The structure of one of the three serum albumin binding domains has been determined, showing a three-helix bundle domain (Kraulis et al, FEBS Lett 378:190, 1996). This motif was named ABD (albumin binding domain) and is 46 amino acid residues in size. In the literature, it has subsequently also been designated G148-GA3. [0006] Other bacterial albumin binding proteins than protein G from Streptococcus have also been identified, which contain domains similar to the albumin binding three-helix domains of protein G. Examples of such proteins are the PAB, PPL, MAG and ZAG proteins. Studies of structure and function of such albumin binding proteins have been carried out and reported e.g. by Johansson and coworkers (Johansson et al, J Mol Biol 266:859-865, 1997; Johansson et al, J Biol Chem 277:8114-8120, 2002), who introduced the designation "GA module" (protein G-related albumin binding module) for the three-helix protein domain responsible for albumin binding. Furthermore, Rozak et al have reported on the creation of artificial variants of the GA module, which were selected and studied with regard to different species specificity and stability (Rozak et al, Biochemistry 45:3263-3271, 2006). In the present disclosure, the terminology with regard to GA modules from different bacterial species established in the articles by Johansson etal and by Rozak et al will be followed. [0007] In addition to the three-helix containing proteins described above, other bacterial proteins exist that bind albumin. For example, the family of streptococcal proteins designated the "M proteins" comprises members that bind albumin (see e.g. Table 2 in Navarre & Schneewind, MMBR 63:174-229, 1999). Non-limiting examples are proteins M1/Emm1, M3/Emm3, M12/Emm12, EmmL55/Emm55, Emm49/EmmL49, and H. Neonatal Fc receptor (FcRn) mediated transcytosis ofHSA [0008] The MHC class l-related neonatal Fc receptor (FcRn) mediates cellular trafficking and recycling of albumin and IgG (Brambell et al, Nature 203:1352, 1964; Chaudhury et al, J Exp Med 197:315, 2003). The FcRn, also known as the Brambell receptor, specifically binds albumin and IgG at low endosomal pH and thus protect pinocytosed proteins from lysosomal degradation by transportation to the cell surface and release at neutral pH. The FcRn has a good affinity for both albumin and IgG at pH 5-6, while showing from poor to no affinity at neutral pH. In this manner, the concentrations and the half-lives of albumin and IgG are regulated. Furthermore, the FcRn is responsible for actively transporting albumin and IgG over cellular barriers, e.g. the epithelium of the airways and the endothelium covering the intestines and the placenta. [0009] As evident from the different sections of this background description, the provision of a selection of polypeptide molecules with a high affinity for albumin is a key factor in the development of various biomedical, biotechnological and other applications, and there is therefore a need in the art of additional such polypeptide molecules. Disclosure of the invention [0010] The first aspect of the invention meets the need for novel polypeptides with a comparably high albumin affinity, through the provision of an albumin binding polypeptide comprising an albumin binding motif, which motif consists of the amino acid sequence: GVSDX5YKX8X9I X11X12AX14TVEGVX20 ALX23X24X25I wherein, independently of each other, X5 is selected from Y and F; Χβ is selected from N, R and S; Xg is selected from V, I, L, M, F and Y; X-11 is selected from N, S, E and D; X-I2 is selected from R and K; X-I4 is selected from K and R; X20 is selected from D, N, Q, E, H, S and R; X23 is selected from K, I and T; X24 is selected from A, S, T, G, H, L and D; and X25 is selected from Η, E and D; the albumin binding polypeptide binding to albumin such that the Kq value of the interaction is at most 1 x 10"^ M. [0011] The above definition of a class of sequence related, albumin binding polypeptides according to the invention is based on a statistical analysis of a large number of albumin binding polypeptides identified and characterized as detailed in the experimental section below. The variants were selected from a large pool of random variants of a parent polypeptide sequence or "scaffold", said selection being based on an interaction with albumin in e.g. phage display or other selection experiments. The identified albumin binding motif, or "ABM", corresponds to the albumin binding region of the parent scaffold, which region constitutes two alpha helices within a three-helical bundle protein domain. While the original amino acid residues of the two ABM helices in the parent scaffold already constitute a binding surface for interaction with albumin, that binding surface is modified by the substitutions according to the invention to provide an alternative albumin binding ability. [0012] As the skilled person will realize, the function of any polypeptide, such as the albumin binding capacity of the polypeptides according to the invention, is dependent on the tertiary structure of the polypeptide. It is therefore possible to make minor changes to the sequence of amino acids in a polypeptide without affecting the function thereof. Thus, the invention encompasses modified variants of the ASM, which are such that the albumin binding characteristics are retained. For example, it is possible that an amino acid residue belonging to a certain functional grouping of amino acid residues (e.g. hydrophobic, hydrophilic, polar etc) could be exchanged for another amino acid residue from the same functional group. [0013] In one embodiment of the polypeptide according to this aspect of the invention, X5 is Y. [0014] In one embodiment of the polypeptide according to this aspect of the invention, Xe is selected from N and R, and may in particular be R. [0015] In one embodiment of the polypeptide according to this aspect of the invention, Xg is L. [0016] In one embodiment of the polypeptide according to this aspect of the invention, X-|-| is selected from N and S, and may in particular be N. [0017] In one embodiment of the polypeptide according to this aspect of the invention, X12 is selected from R and K, such as X12 being R or X-12 being K. [0018] In one embodiment of the polypeptide according to this aspect of the invention, X-14 is K. [0019] In one embodiment of the polypeptide according to this aspect of the invention, X20 is selected from D, N, Q, E, H, S and R, and may in particular be E. [0020] In one embodiment of the polypeptide according to this aspect of the invention, X23 is selected from K and I, and may in particular be K. [0021] In one embodiment of the polypeptide according to this aspect of the invention, X24 is selected from A, S, T, G, H and L. [0022] In a more specific embodiment of the polypeptide according to this aspect of the invention, X24 is L. [0023] In an even more specific embodiment of the polypeptide according to this aspect of the invention, X23X24 is KL. [0024] In another even more specific embodiment of the polypeptide according to this aspect of the invention, X23X24 is TL. [0025] In one embodiment of the polypeptide according to this aspect of the invention, X24 is selected from A, S, T, G and H. [0026] In a more specific embodiment of the polypeptide according to this aspect of the invention, X24 is selected from A, S, T, G and H and X23 is I. [0027] In one embodiment of the polypeptide according to this aspect of the invention, X25 is H. [0028] As described in detail in the experimental section to follow, the selection of albumin binding variants led to the identification of a substantial amount of individual albumin binding motif (ABM) sequences. The sequences of individual albumin binding motifs are presented in Figure 1 and as SEQ ID NO: 1-257. [0029] In a specific embodiment of this aspect of the invention, the ABM sequence is selected from from SEQ ID NO:3, SEQ ID NO:53 and SEQ ID NO:239. [0030] In embodiments of the present invention, the ABM may form part of a three-helix bundle protein domain. For example, the ABM may essentially constitute or form part of two alpha helices with an interconnecting loop, within said three-helix bundle protein domain. [0031] In particular embodiments of the invention, such a three-helix bundle protein domain is selected from the group consisting of three-helix domains of bacterial receptor proteins. Non-limiting examples of such bacterial receptor proteins may be selected from the group consisting of albumin binding receptor proteins from species of Streptococcus, Peptostreptococcus and Finegoldia, such as for example selected from the group consisting of proteins G, MAG, ZAG, PPL and PAB. In a specific embodiment of the invention, the ABM forms part of protein G, such as for example protein G from Streptococcus strain G148. In different variants of this embodiment, the three-helix bundle protein domain of which the ABM forms a part is selected from the group consisting of domain GA1, domain GA2 and domain GA3 of protein G from Streptococcus strain G148, in particular domain GA3. [0032] In alternative embodiments, the ABM forms part of one or more of the five three-helix domains of the bacterial receptor protein protein A from Staphylococcus aureus; i.e. the three-helix bundle protein domain is selected from the group consisting of protein A domains A, B, C, D and E. In other similar embodiments, the ABM forms part of protein Z, derived from domain B of protein A from Staphylococcus aureus. [0033] In embodiments of the present invention wherein the ABM "forms part of" a three-helix bundle protein domain, this is understood to mean that the sequence of the ABM is "inserted" into or "grafted" onto the sequence of the naturally occurring (or otherwise original) three-helix bundle domain, such that the ABM replaces a similar structural motif in the original domain. For example, without wishing to be bound by theory, the ABM is thought to constitute two of the three helices of a three-helix bundle, and can therefore replace such a two-helix motif within any three-helix bundle. As the skilled person will realize, the replacement of two helices of the three-helix bundle domain by the two ABM helices has to be performed so as not to affect the basic structure of the polypeptide. That is, the overall folding of the Ca backbone of the polypeptide according to this embodiment of the invention will be substantially the same as that of the three-helix bundle protein domain of which it forms a part, e.g. having the same elements of secondary structure in the same order etc. Thus, an ABM according to the invention "forms part" of a three-helix bundle domain if the polypeptide according to this embodiment of the invention has the same fold as the original domain, implying that the basic structural properties are shared, those properties e.g. resulting in similar CD spectra. The skilled person is aware of other parameters that are relevant. [0034] In one embodiment of the invention, the albumin binding polypeptide is a three-helix bundle protein domain, which comprises the albumin binding motif as defined above and additional sequences making up the remainder of the three-helix configuration. Thus, the invention provides an albumin binding polypeptide, which comprises the amino acid sequence: LAEAKXgXbAXcXd ELXeKY-[ABM]-LAALP wherein [ABM] is an albumin binding motif as defined above, and, independently of each other, Xg is selected from V and E; Xb is selected from L, E and D; Xq is selected from N, L and I; Xq is selected from R and K; and Xe is selected from D and K. [0035] In one embodiment of this polypeptide, Xg is V. [0036] In one embodiment of this polypeptide, )¾ is L. [0037] In one embodiment of this polypeptide, Xc is N. [0038] In one embodiment of this polypeptide, Xq is R. [0039] In one embodiment of this polypeptide, Xe is D. [0040] Again, as described in detail in the experimental section to follow, the selection and sequencing of a number of albumin binding variants led to the identification of individual albumin binding polypeptide sequences. The sequences of these individual albumin binding polypeptides are presented in Figure 1. Also encompassed by the present invention is an albumin binding polypeptide having an amino acid sequence with 85 % or greater identity to a sequence selected from SEQ ID NO:260, SEQ ID NO:310 and SEQ ID NO:496. [0041] As is evident from the above, in addition to a polypeptide whose amino acid sequence is selected from SEQ ID NO:260, SEQ ID NO:310 and SEQ ID NO:496, the present invention also encompasses variants thereof. The amino acid sequences of such encompassed variants exhibit small differences only in comparison with SEQ ID NO:260, SEQ ID NO:310 and SEQ ID NO:496. In some embodiments, the inventive polypeptide may have a sequence which is at least 86 %, at least 87 %, at least 88 %, at least 89 %, at least 90 %, at least 91 %, at least 92 %, at least 93 %, at least 94 %, at least 95 %, at least 96 %, at least 97 %, at least 98 % or at least 99 % identical to the sequence selected from SEQ ID NO:260, SEQ ID NO:310 and SEQ ID NO:496. The comparison may be performed over a window corresponding to the shortest of the sequences being compared, or over a window corresponding to an albumin binding motif in at least one of the sequences being compared. [0042] The terms "albumin binding" and "binding affinity for albumin" as used in this specification refer to a property of a polypeptide which may be tested for example by the use of surface plasmon resonance technology, such as in a Biacore instrument. For example as described in the examples below, albumin binding affinity may be tested in an experiment in which albumin, or a fragment thereof, is immobilized on a sensor chip of the instrument, and the sample containing the polypeptide to be tested is passed over the chip. Alternatively, the polypeptide to be tested is immobilized on a sensor chip of the instrument, and a sample containing albumin, or a fragment thereof, is passed over the chip. Albumin may, in this regard, be a serum albumin from a mammal, such as human serum albumin. The skilled person may then interpret the results obtained by such experiments to establish at least a qualitative measure of the binding affinity of the polypeptide for albumin. If a qualitative measure is desired, for example to determine a Kd value for the interaction, surface plasmon resonance methods may also be used. Binding values may for example be defined in a Biacore2000 instrument (Biacore AB). Albumin is suitably immobilized on a sensor chip of the measurement, and samples of the polypeptide whose affinity is to be determined are prepared by serial dilution and injected in random order. Kd values may then be calculated from the results using for example the 1:1 Langmuir binding model of the BIAevaluation 4.1 software provided by the instrument manufacturer (Biacore AB). [0043] The albumin binding polypeptide according to this first aspect of the present invention binds to albumin such that the Kq value of the interaction is at most 1 x 10'^ M, i.e. 1 nM. In some embodiments, the Kq value of the interaction is at most 1 x 10"”^ M, such as at most 1 x10'11 M, for example at most 1 x 10'12 M. [0044] In one embodiment of the invention, the albumin to which the albumin binding polypeptide binds is human serum albumin. [0045] The invention also encompasses an albumin binding polypeptide as described above, which further comprises one or more additional amino acid(s) positioned on one or both sides of the albumin binding motif. These additional amino acid residues may play a role in enhancing the binding of albumin by the polypeptide, but may equally well serve other purposes, related for example to one or more of the production, purification, stabilization in vivo or in vitro, coupling or detection of the polypeptide, as well as any combination thereof. Such additional amino acid residues may comprise one or more amino acid residue(s) added for purposes of chemical coupling, e.g. to a chromatographic resin to obtain an affinity matrix or to a chelating moiety for complexing with a metal radionuclide. An example of this is the addition of a cysteine residue at the very first or very last position in the polypeptide chain, i.e. at the N or C terminus. Such additional amino acid residues may also comprise a "tag" for purification or detection of the polypeptide, such as a hexahistidyl (Hisg) tag, or a "myc" ("c-Myc") tag or a "FLAG" tag for interaction with antibodies specific to the tag. The skilled person is aware of other alternatives. [0046] The "additional amino acid residues" discussed above may also constitute one or more polypeptide domain(s) with any desired function, such as the same binding function as the first, albumin binding domain, or another binding function, or a therapeutic function, or an enzymatic function, or a fluorescent function, or mixtures thereof. Linked polypeptide "units" in a such a polypeptide according to the invention may be connected by covalent coupling using known organic chemistry methods, or expressed as one or more fusion polypeptides in a system for recombinant expression of polypeptides, or joined in any other fashion, directly or mediated by a linker comprising a number of amino acids. [0047] Furthermore, this aspect of the invention also encompasses fragments of albumin binding polypeptides that retain albumin binding. The possibility of creating fragments of a wild-type three-helix domain with retained binding specificity was shown by Braisted AC et al in Proc Natl Acad Sci USA 93:5688-5692 (1996). In the experiments described in that paper, using a structure-based design and phage display methods, the binding domain of a three-helix bundle of 59 residues was reduced to a resulting two-helix derivative of 33 residues. This was achieved by stepwise selection of random mutations from different regions, which caused the stability and binding affinity to be iteratively improved. Following the same reasoning, with the polypeptides of the present invention, the skilled addressee will be able to obtain a "minimized" albumin binding polypeptide with the same binding properties as that of the "parent" albumin binding polypeptide. Thus, a polypeptide constituting a fragment of a polypeptide according to the invention and binding to human serum albumin such that the Kq value of the interaction is at most 1 x 10'9 M is within the scope of the invention. As a non-limiting example, the fragment may correspond to an albumin binding polypeptide according to the description above which has been N-terminally truncated. Such a truncation may for example be by from 1 to 3 amino acids. [0048] As outlined above, the invention also encompasses multimers of the polypeptide with affinity for albumin, i.e. polypeptide chains comprising at least two albumin binding polypeptides or fragments thereof as monomer units. It may be of interest, e.g. in a method of purification of albumin or in a therapeutic method exploiting the albumin binding function, to obtain even stronger binding of albumin than is possible with one polypeptide according to the invention. In this case, the provision of a multimer, such as a dimer, trimer or tetramer, of the polypeptide may provide the necessary avidity effects. The multimer may consist of a suitable number of polypeptides according to the invention. These polypeptide domains according to the invention, forming monomers in such a multimer, may all have the same amino acid sequence, but it is equally possible that they have different amino acid sequences. As described above, the linked polypeptide "units" in a multimer according to the invention may be connected by covalent coupling using known organic chemistry methods, or expressed as one or more fusion polypeptides in a system for recombinant expression of polypeptides, or joined in any other fashion, directly or mediated by a linker comprising a number of amino acids. [0049] Additionally, "heterogenic" fusion polypeptides or proteins, or conjugates, in which an albumin binding polypeptide according to the invention, or fragment or multimer thereof, constitutes a first domain, or first moiety, and the second and further moieties have other functions than binding albumin, are also contemplated and fall within the ambit of the present invention. The second and further moiety/moieties of the fusion polypeptide or conjugate in such a protein suitably has a desired biological activity. Non-limiting examples of such a desired biological activity comprise a therapeutic activity, a binding activity, and an enzymatic activity. In some embodiments of this aspect of the invention, the second moiety and any further moieties are selected from the group consisting of GLP-1 (glucagon-like peptide 1); HGH (human growth hormone); G-CSF (granulocyte colony-stimulating factor); IL-1 receptor agonist (interleukin 1 receptor agonist); TNF-α (tumor necrosis factor alpha); and blood clotting factors Vil, VIII, IX and X In other embodiments, said second and any further moieties are selected from binding moieties capable of selective interaction (binding) with a target molecule, typically a target molecule other than albumin even though albumin is not excluded. Such a binding moiety is suitably selected from the group consisting of antibodies and fragments and domains thereof substantially retaining antibody binding activity; microbodies, maxybodies, avimers and other small disulfide-bonded proteins; and binding proteins derived from a scaffold selected from the group consisting of staphylococcal protein A and domains thereof, lipocalins, ankyrin repeat domains, cellulose binding domains, γ crystallines, green fluorescent protein, human cytotoxic T lymphocyte-associated antigen 4, protease inhibitors, PDZ domains, peptide aptamers, staphylococcal nuclease, tendamistats, fibronectin type III domain, zinc fingers, conotoxins, and Kunitz domains. In some embodiments of the invention, the target molecule for binding of said target binding moiety is selected from the group consisting of Αβ peptide; other disease-associated amyloid peptides; toxins, such as bacterial toxins and snake venoms; blood clotting factors, such as von Willebrand factor; interleukins, such as IL-13; myostatin; pro-inflammatory factors, such as TNF-α, TNF-a receptor and IL-8; complement factors, such as C3a and C5a; hypersensitivity mediators, such as histamine and IgE; tumor-related antigens, such as CD19, CD20, CD22, CD30, CD33, CD40, CD52, CD70, cMet, HER1, HER2, HER3, HER4, CA9, CEA, IL-2 receptor, MUC1, PSMA, TAG-72. [0050] Other possibilities for the creation of fusion polypeptides or conjugates are also contemplated. Thus, an albumin binding polypeptide according to the first aspect of the invention may be covalently coupled to a second or further moiety or moieties, which in addition to or instead of target binding exhibit other functions. One example is a fusion between one or more albumin binding polypeptide(s) and an enzymatically active polypeptide serving as a reporter or effector moiety. Examples of reporter enzymes, which may be coupled to the albumin binding polypeptide to form a fusion protein, are known to the skilled person and include enzymes such as β-galactosidase, alkaline phosphatase, horseradish peroxidase, carboxypeptidase. Other options for the second and further moiety or moieties of a fusion polypeptide or conjugate according to the invention include, also without limitation, fluorescent polypeptides, such as green fluorescent protein, red fluorescent protein, luciferase and variants thereof. [0051] With regard to the description above of fusion proteins or conjugates incorporating an albumin binding polypeptide according to the invention, it is to be noted that the designation of first, second and further moieties is made for clarity reasons to distinguish between albumin binding polypeptide or polypeptides according to the invention on the one hand, and moieties exhibiting other functions on the other hand. These designations are not intended to refer to the actual order of the different domains in the polypeptide chain of the fusion protein or conjugate. Thus, for example, said first moiety may without restriction appear at the N-terminal end, in the middle, or at the C-terminal end of the fusion protein or conjugate. [0052] The invention also encompasses polypeptides in which an albumin binding polypeptide as described above has been provided with a label, such as selected from the group consisting of fluorescent dyes and metals, chromophoric dyes, chemiluminescent compounds and bioluminescent proteins, enzymes, radionuclides and particles, for example for purposes of detection of the polypeptide in vitro or in vivo. [0053] A related aspect of the present invention provides a polynucleotide encoding a polypeptide as described above. Also disclosed are an expression vector comprising the polynucleotide and a host cell comprising the expression vector. The latter three are tools for the production of a polypeptide according to the invention, and the skilled person will be able to obtain them and put them into practical use without undue burden, given the information herein concerning the polypeptide that is to be expressed and given the current level of skill in the art of recombinant expression of proteins. Thus, also disclosed are methods of producing a polypeptide according to the first aspect of the invention, comprising expressing a polynucleotide as herein described, for example via the culturing of a host cell as herein defined under conditions permitting expression of the polypeptide from the expression vector, and isolating the polypeptide. [0054] As described in the background section and as is well known to the person skilled in the art, the possible applications of a polypeptide molecule with a binding affinity for albumin are several. The albumin binding polypeptide, as well as a fragment, multimer and fusion protein or conjugate thereof, of the invention may find use in any one or more of these applications. [0055] As a non-limiting example of applications of the albumin binding polypeptides described above, the present invention provides, in another of its aspects, a fusion protein or conjugate of an albumin binding polypeptide according to the first aspect of the invention with a polypeptide having a desired biological activity (as defined above) for use as a medicament which exhibits a half-life in vivo which is longer than the half-life in vivo of the polypeptide having a desired biological activity per se. Alternatively speaking, disclosed is a method for prolonging the half-life in vivo of a polypeptide having a desired biological activity, through the fusion or conjugation of such a polypeptide to an albumin binding polypeptide according to the first aspect of the invention. For details of this application of albumin binding molecules, reference is made e.g. to the teachings of the PCT applications published as WO91/01743 and WO01/45746. [0056] As another non-limiting example of applications, the present invention provides, in another of its aspects, a fusion protein or conjugate of an albumin binding polypeptide according to the first aspect of the invention with a polypeptide having a desired biological activity (as defined above) for use as a medicament which elicits no or a reduced immune response upon administration to the mammal, as compared to the immune response elicited upon administration to the mammal of the polypeptide having a desired biological activity perse. Alternatively speaking, disclosed is a method for decreasing the immunogenicity of a polypeptide having a desired biological activity, through the fusion or conjugation of such a polypeptide to an albumin binding polypeptide according to the first aspect of the invention. For details of this application of albumin binding molecules, reference is made to the teachings of the PCT application published as W02005/097202. [0057] A slow extravasation of monoclonal antibodies from blood has been raised as one of the biological barriers that limit the efficacy of antibody mediated therapy (Wu and Senter, Nature Biotechnology 23:1137-46, 2005). Interestingly, at equilibrium, approximately 60 % of the serum albumin in a human being is found in the interstitial space, whereas only 40 % is found in the blood stream. Thus, the association with albumin as provided by the present invention is considered a superior means of obtaining a wide distribution outside of the blood stream. The affinity of the association with serum albumin is suitably characterized by an off-rate (decomposition of the complex) that is sufficiently slow, such that only a minute fraction of the complex dissociates during transition from the blood stream to the interstitium. However, the interaction does not have to be covalent, since some rebinding is possible during the transition. [0058] One possible contributing mechanism for extravasation and wide distribution is active transport following binding of serum albumin to the FcRn receptor. Consequently, there are certain requirements on the albumin binding moiety in an albumin binding fusion protein to obtain a similar distribution. For example, the affinity may be very tight also in the acidic environment encountered during receptor transport in the cell, probably down to a pH below6. Brief description of the figures [0059] Figure 1 is a listing of the amino acid sequences of examples of albumin binding motifs comprised in albumin binding polypeptides of the invention (SEQ ID NO:1-257), examples of albumin binding polypeptides according to the invention (SEQ ID NO:257-514), and the GA3 domain from protein G of Streptococcus strain G148 (SEQ ID NO:515). Figure 2 is an illustration of the main features of the coding insert in expression vector pAY1075 without (A) and with (B) a cassette encoding helices 2 and 3 of variant ABD molecules. Figure 3Ashows the strategy for amplification of DNAfragments encoding dummy, Zwt and Gill in the preparation of the coding insert of expression vector pAY1075. Figure 3B shows the overlap of these fragments for the creation of the entire coding insert. Figure 4 is a vector map of the expression vector pAY1075, prepared as described in Example 1. Figure 5 is a vector map of the expression vector pAY1075-ABD, prepared as described in Example 1. Figure 6 is a table showing the theoretical (shaded columns) and experimental (clear columns) values for the amino acid variation at each varied position in the ABD variant sub-library created using the AFFI-793 mixture of oligonucleotides, as described in Example 1. Figure 7 is a table showing the theoretical (shaded columns) and experimental (clear columns) values for the amino acid variation at each varied position in the ABD variant sub-library created using the AFFI-794 mixture of oligonucleotides, as described in Example 1. Figure 8 is a schematic illustration of the amino acid sequence of an ABD variant as expressed in the pAY442 vector according to the description in Example 2. Figures 9A-9C show ELISA titration curves for serum obtained at days 0-45 from primates injected with Z00342 as described in Example 4, when analyzed on ELISA plates coated with Z00342. Figures 10A-10C show ELISA titration curves for serum obtained at days 0-45 from primates injected with Z00342-ABD00003 as described in Example 4, when analyzed on ELISA plates coated with Z00342-ABD00003. Figure 11 shows the median concentration of IgG specific for Z variants in serum obtained at days 0-45 from primates injected with Z00342 and Z00342-ABD00003 as described in Example 4. Figure 12 shows the amount of A) Z00342 and B) Z00342-ABD00003 in blood circulation over time as analyzed by sandwich ELISA as described in Example 4. Figures 13A-13B show ELISA titration curves for serum obtained at days 0-45 from primates injected with (Z01154) 2 as described in Example 5, when analyzed on ELISA plates coated with (Z01154)2- Figures 14A-14B show ELISA titration curves for serum obtained at days 0-45 from primates injected with (Z01154)2-ABD00239 as described in Example 5, when analyzed on ELISA plates coated with (Z01154)2-ABD00239. Figure 15 shows normalized values of the A) (Z01154)2 and B) (Z01154)2-ABD00239 samples analyzed in Figures 13 and 14 respectively. The sample absorbances were normalized against the positive control at 1600xdilution. [0060] The invention will now be illustrated further through the non-limiting description of experiments conducted in accordance therewith. Unless otherwise specified, conventional chemistry and molecular biology methods were used throughout. Examples Example 1 Construction of phaae display library of variants of an albumin binding polypeptide Summary [0061] In this example, a phage display library of polypeptide variants was created, through variation of 16 positions in helices 2 and 3 of the albumin binding domain GA3 of Streptococcus strain G148 (in the following referred to as "ABD"). The wildtype sequence of ABD ("ABDwt") is provided as SEQ ID 140:491 in Figure 1 and in the appended sequence listing. A new phage display vector (pAY1075) based on the previously described pAffi 1 vector (Gronwall et al, J Biotechnol 128:162-183, 2007) was constructed for this new library. The varied ABD fragment (helices 2-3) was cloned into pAY1075 with restriction enzymes Sad and Nhe\. Ligations were purified and electroporated to E. coli RR1AM15 cells (Ruther, Nucleic Acids Res 10:5765-5772, 1982). The newly constructed library was designated LibABDmat2005 and consisted of two sub-libraries, depending on which oligonucleotides had been used for creation of the varied sequence of helices 2 and 3. One was built on the ABD molecule and the other had an extra amino acid inserted between positions 17 and 18 of ABD, which some of the proteins homologous to ABD have (see e.g. Rozak et al, Biochemistry 45:3263-3271, 2006). The size of LibABDmat2005 was 1 x 109 members (5 x 108 for each sub-library). The quality of the new library was satisfying, in that DNA sequencing showed that about 87 % of the clones were functional and in that the measured values of relative frequencies of amino acids agreed well with the theoretical values. Construction of phagemide vector pAY1075 [0062] A new phage display vector (pAY1075) was constructed for the new library. pAY1075 was based on the phagemid vector pAffil (Gronwall etal, supra). For creation of pAY1075, pAffil was digested with Xhol and Xma\ (10 units/pl; New England Biolabs), and a new insert or cloning cassette was created and ligated into the vector. The new insert contained DNA encoding helix 1 of ABDwt, a dummy sequence, a thrombin site, Zwt (an engineered IgG binding domain based on domain B of staphylococcal protein A, see Nilsson et al, Prot Eng 1:107-113, 1987), truncated Gill (residues 249-406), the termination domain TT and some additional restriction enzyme sites. For a schematic representation of the elements encoded by this insert, see Figure 2A. Figure 2B shows the insert of the expression vector when the dummy sequence has been replaced by a sequence encoding the remaining ABD variant polypeptide (see below). The sequences of the various DNA oligonucleotides used as primers and templates in the cloning experiment and library construction are provided below in Table 1. Table 1: Oligonucleotide primers and templates

[0063] In order to create the new cloning cassette for pAY1075, the dummy fragment and Gill were PCR amplified from pAffil, and Zwt was amplified from plasmid pEZZ18 (Lowenadler et al, Gene 58:87-97, 1987), using primers according to Figure 3A. The newly generated fragments had overlapping segments to each other, as shown in Figure 3B. The PCR fragments were gel purified with QIAquick gel extraction kit (Qiagen) according to the manufacturer's recommendations and thereafter assembled together with the oligonucleotide AFFI-772 (Figure 3B) in an assembly PCR. A further PCR reaction using external primers AFFI-772 and AFFI-40 was performed to amplify the entire fragment. PCR products were purified using QIAquick PCR purification kit (Qiagen) according to the manufacturer's recommendations. [0064] The plasmid pAffil was purified with QIAgen™ midi-prep kit (Qiagen), according to the manufacturer's recommendations. Thereafter, pAffil and the amplified PCR fragment for the cloning cassette were digested with Xhol and XmaI (10 units/pl; New England Biolabs) in NEB4 buffer (20 mM Tris acetate, 10 mM magnesium acetate, 50 mM potassium acetate,1 mM dithiothreitol, pH 7.9; New England Biolabs) for 1 h at 37 °C and the vector was thereafter dephosphorylated using calf intestinal alkaline phosphatase (CIAP; Fermentas). The digestions were purified on a 1 % agarose gel using QIAquick gel extraction kit according to the manufacturer's recommendations. The new fragment was ligated into Xhol andXmal cleaved pAffil for 1 h at room temperature using T4 DNAIigase (5 units/pl; Fermentas). Part of the ligation mixture was electroporated into E. coli TG1 cells (Stratagene) using 1 mm cuvettes. The cells were plated on tryptose blood agar base plates (TBAB plates; 30 g/l TBAB) supplemented with 200 pg/ml ampicillin. Clones having a correct insert were identified by PCR, using the three different primer pairs AFFI-21/AFFI-42, AFFI-47/AFFI-40 and AFFI-21/AFFI-40. PCR fragments were analyzed on a 1 % agarose gel, and positive clones were plasmid purified with QIAprep Miniprep kit (Qiagen), according to the manufacturer's recommendations, and thereafter sequenced with primers AFFI-38, 40, 71, 72, and 772 using ABI PRISM® BigDye™ Terminator Cycle Sequencing Ready Reaction kit 3.1 (Applied Biosystems). The sequence PCR reactions were purified on a Magnatrix 8000 instrument (Magnetic Biosolutions) and the nucleotide sequence was determined with an ABI PRISM® 3100 Genetic analyzer (Applied Biosystems). Sequencer™ v 4.0.5 (Gene Codes Corporation) was used to record and analyze the sequence data. Sequencing revealed that the new phagemid vector had been successfully created. Vector maps of the vector with dummy sequence (pAY1075) or with sequence encoding varied ABD helices 2 and 3 (pAY1075-ABD) are shown in Figure 4 and 5, respectively. Design of a library of variant ABD sequences [0065] A set of oligonucleotides having a randomized sequence for helices 2 and 3 of the ABD molecule were prepared as described below. These oligonucleotides were subsequently used for replacement of the dummy sequence in pAY1075, to create pAY1075-ABD (Figures 2B and 5). The pAY1075-ABD vector was subsequently used for the expression of the library of ABD variants on the surface of phages [0066] The design was based on information from alanine mutations of ABD (Linhult et al, Protein Science 11:206-213, 2002), a study of the ALB8-HSAcomplex (Lejon et al, J Biol Chem 279:42924-42928, 2004), sequence homologies with other known albumin binding domains and ease of oligonucleotide preparation. 16 amino acid positions in the ABDwt sequence represented by SEQ ID 140:491 were selected for some degree of randomization, and were grouped in 4 different groups depending on characteristic: (I) the hydrophobic core, (II) conserved positions, (III) electrostatic interactions and (IV) others: 1. (I) Positions Y20, L24, L27 and 141 are involved in creating the central hydrophobic core in the interaction with serum albumin. These positions are highly conserved among domains homologous to ABD and randomization in these positions tested whether another hydrophobic amino acid residue could improve the hydrophobic interaction. 2. (II) Positions 18', S18, T30, E32 and G33 are very conserved among the albumin binding domains. Positions S18 and T30 are involved in two intermolecular H-bonds, and the rationale for randomization was that similar polar amino acids like threonine (T) and asparagine (N) could also work. E32 and G33 do not interact with the binding surface to any great extent. However, they are likely to be important for protein structure, and it was of interest to see if another amino acid could work. The sequence of ABDwt does not comprise position 18' (i.e. 18' represents an added amino acid residue between positions 17 and 18 in ABDwt), but homologous domains have threonine or serine at that position. It was of interest to see if binding could be improved with this additional amino acid. 3. (Ill) Positions N23, N27, K29 and E40 are involved or could be involved in electrostatic interactions. The randomization at these positions was based on an interest to see whether or not it would be possible to enhance or suppress some of the attractive or repulsive interactions of these amino acid residues with albumin. 4. (IV) Positions A36, K35 and D39 were randomized due to other similar considerations. [0067] In order to create the desired mix of amino acid residues at each position, the ABDwt sequence was varied in accordance with Table 2. Variations were categorized as "randomized" or "doped". In "randomized" positions, all chosen amino acids were represented in equal proportions. In "doped" positions, the original amino acid was more frequent than the others, i.e. the position was biased towards the original amino acid. Table 2: Design strategy for variant ABD sequences

[0068] Oligonucleotide mixtures AFFI-793 and AFFI-794 corresponding to DNA encoding residues 13-46 of the ABDwt sequence as modified according to Table 2 and including restriction sites were obtained from Scandinavian Gene Synthesis AB. AFFI-794 comprises the extra amino acid represented by position 18'. AFFI-793:

AFFI-794:

[0069] Table 3 summarizes the required percentage distribution of nucleotides in the oligonucleotide mixtures necessary to achieve the library design described in Table 2. Table 3: Distribution of nucleotides in AFFI-793 and AFFI-794 oligonucleotide mixtures

Library construction [0070] The following procedure was used to create the genetic library LibABDmat2005, encoding ABD variants. In an assembly reaction, the oligonucleotide AFFI-791 and oligonucleotide mixture AFFI-793 or AFFI-794 were annealed and extended with Taq DNA polymerase. A PCR reaction using the external primers AFFI-791 and AFFI-792 was performed to amplify the fragment. PCR products were purified using QIAquick PCR purification kit (Qiagen) according to the manufacturer's recommendations. [0071] Phagemid pAY1075 was prepared from 250 ml overnight culture (tryptic soy broth, 2 % glucose, 100 pg/ml ampicillin) using Qiagen plasmid midi kit (Qiagen) following the manufacturer's recommendations. The phagemid was digested with Sad and Nhe\ (10 units/pl; New England Biolabs) in NEB4 buffer (20 mM Tris acetate, 10 mM magnesium acetate, 50 mM potassium acetate, 1 mM dithiothreitol, phi 7.9; New England Biolabs) for 3 h at 37 °C. The solution was phenol/chloroform purified and EtOFI precipitated, and the vector was then gel purified from a 1 % agarose gel using QIAquick gel extraction kit (Qiagen) according to the manufacturer's recommendations. [0072] The PCR amplified fragments from the assembly reactions between AFFI-791 and AFFI-793 or AFFI-794 were digested with Sad and Nhe\ in NEB4 buffer for 3 h at 37 °C. The DNA fragments were purified from a 1 % agarose gel using QIAquick gel extraction kit according to the manufacturer's recommendations. The resulting gene fragments encoding two sub-libraries of variants of ABD were ligated into Sad and Nhe\ cleaved pAY1075 for 1 h at room temperature using T4 DNA ligase (5 units/pl; Fermentas). [0073] The ligations were then phenol/chloroform extracted, EtOH precipitated and resolved in a smaller volume of 10mM Tris. Electrocompetent E. coli RR1AM15 cells (Ruther, 1982, supra) were transformed with 60 aliquots of ligated material of each of the two sub-libraries using 0.2 cm gap size cuvettes in an ECM 630 set (BTX) using the parameters 2.5 kV, 125 Ω and 50 pF. Cells were grown in SOC medium (47 ml TSB+YE (30 g/l tryptic soy broth, 5 g/l yeast extract) supplemented with 1 % glucose, 10 mM MgCl2, 10 mM MgSC>4, 10 mM NaCI and 2.5 mM KCI) for 50 minutes and transferred to ten Erlenmayer flasks, each containing 1 I of TSB + YE (30 g/l tryptic soy broth, 5 g/l yeast extract) supplemented with 2 % glucose and 100 pg/ml ampicillin, and grown overnight at 37 °C. The cells were then centrifuged at 6000 g and re-suspended in PBS/glycerol solution (PBS: 2.68 mM KCI, 1.47 mM KH2PO4, 137 mM NaCI, 8.1 mM Na2HPC>4, pH 7.4) to a final approximate concentration of 20 % glycerol. The cells were then aliquoted and stored at - 80 °C. The number of cells after electroporation, amplification and transfer to glycerol stocks was titrated on TBAB plates supplemented with 200 pg/ml ampicillin. [0074] The size of each sub-library was 5 x 108, i.e. the total size of the library LibABDmat2005 was 1 x 109. The library was amplified about 50000 times and the glycerol stocks had a density of about 1 x 10^ cells/ml. In this context, the "size" of the library means the total number of members comprised in the library, without any regard to the number of unique variants encoded by the library. [0075] Ninety-six colonies from each of the two sub-libraries were picked for DNA sequencing in order to verify the design and the frequency of clones with a correct reading frame. These randomly picked colonies, cultured from glycerol stocks and originating from each pool of the library were PCR amplified using oligonucleotides AFFI-21 and AFFI-22. Sequencing of the amplified fragments was performed using ABI PRISM® dGTP, Big Dye™ Terminator v3.0 Ready Reaction Cycle Sequencing Kit (Applied Biosystems) according to the manufacturer's recommendations and with the biotinylated oligonucleotide AFFI-72. The sequencing reactions were purified by binding to magnetic, streptavidin-coated beads using a Magnatrix 8000 (Magnetic Biosolutions), and analyzed on ABI PRISM® 3100 Genetic Analyser (Applied Biosystems). [0076] In the sub-library created using AFFI-793, three clones were not readable, eleven were incorrect and seven clones were contaminations from the other sub-library. The amino acid distribution in this sub-library was deducted from the sequencing data, and compared to the theoretical values, and the results are presented in Figure 6. [0077] Regarding the sub-library created using AFFI-794, three clones were not readable and 16 were incorrect. The amino acid distribution in this sub-library was deducted from the sequencing data, and compared to the theoretical values, and the results are presented in Figure 7. [0078] The frequency of each amino acid agreed well with expected value, and about 87 % of the clones had a correct reading frame. Example 2 Phaae display selection and characterization of albumin binding polypeptide variants Summary [0079] Biotinylated human serum albumin (HSA) was used as target in phage display selections using the library constructed in Example 1. Selections were carried out using a variety of conditions in order to maximize the likelihood of obtaining ABD variants having a high affinity for albumin. After elution of selected phages, the corresponding expressed proteins were tested for affinity to albumin in an ELISA setup. Positive clones were identified and sequenced, and the predicted amino acid sequences of the corresponding polypeptides and their albumin binding motifs were deduced, which yielded a large number of sequences of albumin binding polypeptides. The amino acid sequences of deduced albumin binding motifs are listed in Figure 1 and in the sequence listing as SEQ ID NO: 1-257, whereas the amino acid sequences of the corresponding full-length ABD variants are listed in Figure 1 and in the sequence listing as SEQ ID NO:258-514. Biotinylation of human serum albumin [0080] Lyophilized human serum albumin (Sigma, cat. no. A3782-5G) was dissolved in PBS (2.68 mM KCI, 1.47 mM KH2PO4, 137 mM NaCI, 8.1 mM Na2HP04, pH 7.4) to a final concentration of 10 mg/ml. EZ-link Sulfo-NHS-LC-Biotin (Pierce, cat. no. 21335) was dissolved in water to a final concentration of 1 mg/ml and a 5 and 10 fold molar excess was added to 500 mg of albumin in a total volume of 0.5 ml. The mixtures were incubated at room temperature for 30 min. Unbound biotin was removed by dialyzing against PBS using a dialysis cassette (Slide-A-Lyser, 10 kDa; Pierce). Phage display selection [0081] In total, five rounds of selection were carried out, using increasingly stringent conditions. After three initial rounds performed chiefly with a view to establish a suitable selection protocol, the resulting phage stocks were prepared from glycerol stock prepared as in Example 1. Selection was then carried out for two more cycles using the combinations of selection buffer, target concentration and solid support that are listed in Table 4. Table 4: Selection conditions for HSA selection

[0082] All tubes and beads used in the selection procedure were pre-blocked in TPBSB (5 %) (0.05 % Tween20, 5 % bovine serum albumin (BSA), 0.02 % Na azide in PBS) or gelatin (0.5 %) for 30 min under gentle agitation at room temperature and subsequently left with no agitation over night at 4 °C. [0083] Selection solutions (1 ml) contained biotinylated human serum albumin, phages, Na azide (0.02 %), Tween 20 (0.05 %) and either BSA (3 %) or gelatin (0.1 %) according to Table 4, and were prepared in PBS. The phages were incubated with biotinylated human serum albumin target at 4 °C during three days for Cycle 4 and during one day for Cycle 5, followed by 1 h incubation under agitation at room temperature. The selection samples were transferred to blocked streptavidin beads for 15 min under agitation at room temperature. The beads were washed 10 times with 1 ml of selection buffer (i.e. TPBSB (3 %) (0.05 % Tween20, 3 % bovine serum albumin (BSA), 0.02 % Na azide in PBS) or GT (0.1 %) (0.1 % gelatin, 0.1 % Tween 20 and 0.02 % Na azide in PBS)), followed by 10 washes with PBS where the second to last wash lasted for 2 hours. Phages were either eluted with 1000 ml 0.05 M Glycine-HCI, pH 2.2, for 10 min at room temperature, followed by immediate neutralization with 900 ml PBS supplemented with 100 ml 1 M Tris-HCI, pH 8.0, or eluted with 1000 pi trypsin (2 mg/ml) for 30 min at room temperature followed by addition of 1000 pi aprotinin. The eluted phages (¾ of the volume) were used to infect 50 ml log phase E. coli RR1AM15 cells (Ruther, 1982, supra) after each cycle of selection. After 30 min incubation with gentle agitation and 30 min with vigorous agitation at 37 °C, the cells were centrifuged and the pellet was dissolved in a smaller volume and spread on TSB+YE plates (30 g/l tryptic soy broth, 5 g/l yeast extract) and finally incubated over night at 37 °C. [0084] The cycles of selection resulted in a satisfying number of eluted phages. Phage stock preparation [0085] Cells from plates were re-suspended in TSB medium (30 g/l tryptic soy broth) and the cell concentration was determined by measuring the optical density at 600 nm assuming that Οϋβοο = 1 corresponds to 5 x 108 cells/ml. Cells were inoculated (approximately 100 times excess of cells compared to eluted phages) in 100 ml TSB+YE medium supplemented with 2 % glucose and 100 mg/ml ampicillin and grown at 37 °C to approximately ODgoo = 0.5-0.7. Thereafter, 10 ml were transferred to a new flask and infected by 25 times excess of M13K07 helper phage (1 x 1012 cfu/ml; New England Biolabs, cat. no. N0315S) and incubated for 30 min with low agitation. Cells were centrifuged at 2000 g for 10 min and re-suspended in 100 ml TSB+YE medium supplemented with 100 mM isopropyl β-D-l-thiogalactopyranoside (IPTG), 50 mg/ml kanamycin and 100 mg/ml ampicillin and grown over night at 100 rpm and 25 °C. A portion of the re-suspended cells was stored at - 80°C as a glycerol stock. [0086] The induced culture was centrifuged at 2500 g for 10 min and phages in the supernatant were precipitated by adding % of the volume of precipitation buffer (PEG/NaCI) and incubated on ice for 1 hour. Precipitated phages were pelleted by centrifugation at 10000 g at 4 °C for 30 min, re-suspended in 20 ml PBS and thereafter the phage solution was filtered through a 0.45 pm filter. The precipitation procedure was repeated and the phages were finally re-suspended in 1 ml PBS. [0087] Selection solution was titrated after the selection together with wash and elution solutions after each round of selection. Phage solutions were diluted in sterile water in a microtiter plate and 100 pi log phase E. coli RR1AM15 cells were added to each phage dilution. After 20 min incubation at room temperature, 5 pi from each titration were dropped on a TYE plate (15 g agar, 10 g tryptone, 5 g yeast extract, 3 g NaCI supplemented with 2 % glucose and 100 pg/ml ampicillin) and incubated over night at 37 °C. The resulting colonies were counted and the titers (cfu/ml) calculated. EUSA analysis of albumin binding [0088] Clones from each selection plus ABDwt were expressed and screened for HSA binding activity using an ELISA setup enabling detection of binders having a Kq value of from 10 nM down to lowpM against serum albumin. Randomly picked colonies were expressed in 96 deep-well plates by inoculating each colony into 1 ml TSB+YE medium supplemented with 100 mg/ml ampicillin and 1 mM IPTG and grown over night at 37 °C. Cells were pelleted by centrifugation at 3000 g for 15 min, re-suspended in 400 pi PBS-T (0.5 % Tween 20 in PBS) and frozen at -80 °C. Frozen samples were thawed in a water bath and cell debris was pelleted at 3700 g for 40 min. Supernatants containing ABD variant-Zwt fusion proteins were collected and stored at 4 °C until used in ELISA as follows. [0089] Microtiter wells (Costar) were coated over night at 4 °C with 100 pi of HSA and with the controls rat serum albumin (RSA), human serum albumin (HSA) and mouse serum albumin (MSA) in one well each, at a concentration of 0.4 pg/ml in ELISA coating buffer (0.1 M sodium carbonate, pH 9.5). The wells were blocked with blocking buffer (2 % milk in PBS) for 2 h at room temperature. A volume of 100 pi of the prepared ABD variant-Zwt fusion proteins was added to each well, and the plates were incubated for 1.5 h at room temperature. Biotinylated IgG at a concentration of 0.5 mg/ml in washing buffer (0.5 % Tween 20 in PBS) was added to the wells and incubated for 1.5 h, so that the Zwt moiety of any albumin binding fusion proteins could bind to IgG. Bound complexes were detected with horse radish peroxidase conjugated streptavidin (Dako, cat. no. P0397) diluted 1:5000 in washing buffer, and incubated for 1 h at room temperature. Developing solution was prepared by mixing an equal volume of TMB substrates A and B (ImmunoPure TMB, Pierce), and 100 pi was added to each well. After 30 min incubation in darkness, 100 pi stop solution (2 M H2SO4) was added. The plates were read at A450 in an ELISA spectrophotometer (Basic Sunrise, Tecan). Prior to addition of each new reagent, four washes were done with washing buffer. [0090] In total, 372 clones (93 clones from each selection denoted Sample A-D in Table 4) were randomly picked for analysis of their HSA binding activity using the ELISA set-up described above. The majority of the analyzed clones gave a higher signal to HSA as compared to the ABDwt interaction with rat serum albumin, which is a lowpico molar binding (70 pM; unpublished results). Based on the result of this experiment, clones were picked for sequencing as described next. Sequencing of EUSA positive clones [0091] PCR fragments from selected colonies were amplified using oligonucleotides AFFI-69 (5'-gtgagcggataacaattcccctc-3') and AFFI-70 (5'-cagcaaaaaacccctcaagaccc-3'). Sequencing of amplified fragments was performed using ABI PRISM® dGTP, BigDye™ Terminator v3.0 Ready Reaction Cycle Sequencing Kit (Applied Biosystems) according to the manufacturer's recommendations and with the biotinylated oligonucleotide AFFI-202 (5'-i>/of/7i-gtgagcggataacaattcccctc-3'). The sequencing reactions were purified by binding to magnetic streptavidin-coated beads using a Magnatrix 8000 instrument (Magnetic Biosolutions), and finally analyzed on ABI PRISM® 3100 Genetic Analyser (Applied Biosystems). [0092] The clones exhibiting the highest A450 value in the ELISA screening were subjected to sequencing of their ABD variant insert. 257 different identified ABD variants were given the designation ABD////////#, wherein ##### is a five-digit unique label for the variant in question. The sequences of these identified ABD variants are listed in Figure 1 as SEQ ID 140:258-514. Based on the existing knowledge of the albumin binding properties of the wild-type or "parent" ABD, the albumin binding motifs of the identified ABD variants were deduced to reside in the two helices 2 and 3, corresponding to the stretch from amino acid position G16 to 141. The albumin binding motifs of the identified ABD variants were given the designation ABM#####, where #//////// is a five-digit unique label for the motif in question. The sequences of the identified albumin binding motifs are listed in Figure 1 as SEQ ID NO: 1-257. Interestingly, a subset of the sequences identified comprised a spontaneous mutation at the position corresponding to position 38 in ABDwt, despite the fact that this position had not been randomized in the creation of the library of variants. Sub-cloning of A8D variants into plasmidpAY442 [0093] DNA encoding ABDwt (SEQ ID NO:515) and twelve clones from the selection were selected for sub-cloning into the expression vector pAY442 (Gronwall et al, supra). With reference to Figure 1, the selected ABD variant clones were ABD00002, ABD00003, ABD00009, ABD00015, ABD00025, ABD00027, ABD00046, ABD00049, ABD00053, ABD00054, ABD00055 and ABD00245. Plasmids containing inserts encoding these ABD variant molecules were purified from 2 ml over night cultures (tryptic soy broth medium (30 g/l) supplemented with 2 % glucose and 100 pg/ml ampicillin) of E. coli RR1AM15 cells (Ruther, 1982, supra) using Qiagen Mini Kit (Qiagen) according to the manufacturer's recommendations. [0094] DNA for ABDwt and ABD variant molecules was sub-cloned into the expression vector pAY442 by Acc\-Not\ PCR sticky end cloning (10 units/pl of each enzyme; New England Biolabs) using the primer pairs AFFI-780, -898 and AFFI-782, -899 as listed in Table 5: Table 5: Oligonucleotide primers

[0095] Two overlapping PCR products for each ABD variant molecule were generated from the library vector pAY1075, resulting in approximately 25 % correct fragments with an Acc\-Not\ site. The expression vector pAY442 was digested in two steps at 37°C for 4 h using Accl and Not\ in NEB4 buffer (20 mM Tris acetate, 10 mM magnesium acetate, 50 mM potassium acetate, 1 mM dithiothreitol, pH 7.9; New England Biolabs) and NEB3 buffer (50 mM Tris-HCI, 10 mM MgCl2, 100 mM NaCI, 1 mM dithiothreitol, pH 7.9; New England Biolabs), respectively, and dephosphorylated with calf intestinal alkaline phosphatase (CIAP; Fermentas) for 1 h at 37 °C. The cleaved plasmid and fragments were purified by QIAquick PCR purification kit (Qiagen) according to the manufacturer's recommendations. [0096] The PCR products were hybridized and ligated into Acc\-Not\ digested and dephosphorylated pAY442 for 1 h at room temperature using T4 DNAIigase (5 units/pl; Fermentas). Part of the ligations were electroporated into E. coli BL21 (DE3) cells (F" ompT hsdSs(rQ (tib") gal dem (DE3)) using a 1 mm cuvette and an ECM 630 set (BTX) using the parameters 1700 V, 200 Ω and 25 pF. The cells were plated on tryptose blood agar base (TBAB) plates supplemented with 50 pg/ml kanamycin and incubated over night at 37 °C. Positive clones were first verified on agarose gel of bacterial PCR products and finally with DNA sequence analysis. [0097] Clones of pAY442 containing a successfully sub-cloned ABD variant encode a construct that is schematically described in Figure 8, i.e. essentially a Hisg tagged ABD variant. Expression and purification of Hisg tagged ABD variants [0098] ABDwt and the twelve ABD variants, all sub-cloned in pAY442 as described above, were expressed in E. coli BL21 (DE3) as fusions to an N-terminal Hisg-tag and purified by IMAC. A colony of each ABD variant was used to inoculate 5 ml TSB medium supplemented with 50 pg/ml kanamycin. The cultures were grown over night at 37 °C. The following day, 50 pi of each culture were inoculated separately to 100 ml TSB+YE medium supplemented with 50 pg/ml kanamycin in a 1 liter flask. The cultures were grown at 100 rpm at 37 °C to an ODgoo of 0.7-1, after which IPTG was added to a final concentration of 0.5 mM and cells were incubated at room temperature over night at 100 rpm. Cultures were harvested by centrifugation at 8000 g for 5 minutes and pellets were stored in a freezer until protein preparation. [0099] The Hisg-tagged proteins were IMAC purified under denatured conditions using Ni-NTA Superflow columns and QIAsoft 4.1, protein/Ni-NTA Superflow 96 denaturing large scale 2 Vac4-24 samples, on a Biorobot 3000 (Qiagen). The buffer was exchanged to PBS using a dialysis cassette (Slide-A-Lyser, 3.5 kDa; Pierce cat. no. 66330) by dialyzing against 5 I PBS for 2 h followed by an additional dialysis over night. [0100] Protein concentration was determined using A28O and the BCA Protein Assay Reagent Kit (Pierce) as recommended by the manufacturer. The purity of the proteins was analyzed by SDS-PAGE on 4-12 % Novex gels and stained with Coomassie Blue R, and this analysis showed that only small amounts of impurities were present. Biosensor analysis of ABD variants’ affinity for HSA and MSA [0101] Biosensor analysis on a Biacore2000 instrument (Biacore) was performed with MSA, HSA and RSA immobilized by amine coupling onto the carboxylated dextran layer on the surfaces of CM-5 chips (research grade; Biacore) according to the manufacturer's recommendations. Surface 1 on the chip was activated and deactivated and used as reference cell during injections, whereas surface 2 comprised MSA immobilized with 350 RU (resonance units), surface 3 comprised HSA immobilized with 360 RU and surface 4 comprised RSA immobilized with 340 RU. The ABD variants and ABDwt expressed and purified as described above were diluted in HBS-EP (Biacore) to 25 nM and injected at a constant flow-rate of 25 pl/min for 10 minutes, followed by injection of HBS-EP for 30 minutes. The surfaces were regenerated with twe injections of 20 pi 15 mM HCI followed by 0.05 % SDS and one more injection of 20 pi HCI. [0102] The Biacore study was not carried out with a view to determine the exact parameters for the variants' affinity for human and mouse serum albumin, but the results provide a qualitative measure of the relative affinities of these molecules for albumin. Results for binding to MSA and HSAare presented in Table 6. Table 6: Biosensor analysis of ABD variants' binding to serum albumin from mouse and human

[0103] As is evident from this table, all tested ABD variants had a substantially higher affinity for human serum albumin than the wild type ABD molecule, as evidenced by Kq values at least one order of magnitude lower, frequently approaching two orders of magnitude lower. Furthermore, comparable and/or higher affinities towards mouse serum albumin was also exhibited by all variants. Example 3 Additional biosensor characterization of selected ABD variants Summary [0104] In this example, selected ABD variants ABD00003, ABD00053 and ABD00239 plus ABDwt were all sub-cloned in pAY442 as described in Example 2 above, and expressed in a larger scale and purified with His Gravitrap™ kit. The expressed molecules were characterized for affinity to HSA using a Biacore instrument. Protein expression and purification of Hisg tagged ABD variants [0105] ABD00003, ABD00053, ABD00239 and ABDwt were expressed in E. coli BL21(DE3) cells as fusions to an N-terminal Hisg-tag using constructs as described in Example 2, and purified by IMAC. A colony of each ABD variant was used to inoculate 10 ml TSB medium supplemented with 50 pg/ml kanamycin. The cultures were grown over night at 37 °C. The following day, 500 pi of each culture were inoculated separately to 500 ml TSB+YE medium supplemented with 50 pg/ml kanamycin in a 5 liter flask. The cultures were grown at 100 rpm and 37 °C to an ODgoo of 0.7-1, which was followed by addition of IPTG to a final concentration of 0.5 mM and incubated at room temperature over night. Cultures were harvested by centrifugation at 8000 g for 5 minutes and pellets were stored at -20 °C until protein preparation. [0106] The Hisg-tagged proteins were IMAC purified under denaturing conditions using His-Gravitrap™ kit (GE Healthcare). The pellets were re-suspended (vortexed) in 20 ml of the denaturation buffer B-7M (100 mM NaH2P04, 10 mM Tris-CI, 7 M urea, pH 8) and 8 pi benzonase was added. The solutions were incubated for 30 minutes at room temperature and 200 rpm. An additional 20 ml of buffer B- 7M was added and the solutions were transferred to 50 ml Falcon tubes and sonicated on ice as follows: 3 s on/off during 3 min and with 40 % amplitude. Cell debris was removed by centrifugation at 25000 g for 40 min. The Gravitrap™ columns were equilibrated with buffer B-7M and the samples were applied. The columns were then washed with 10 ml buffer B-7M, 20 ml binding buffer (20 mM NaPC>4, 500 mM NaCI, 20 mM imidazole) and finally with 10 ml wash buffer (20 mN NaPC>4, 500 mM NaCI, 60 mM imidazole). The ABD molecules were eluted with 3 ml elution buffer (20 mN NaPC>4, 500 mM NaCI, 500 mM imidazole). [0107] A buffer exchange to PBS pH 7.2 using a Slide-A-Lyser dialysis cassette (3.5 kDa; Pierce, cat. no. 66330) was made by dialyzing against 5 I PBS pH 7.2 for 2 hours followed by an additional dialysis over night and finally a buffer exchange to PBS pH 5 using PD 10 columns (GE Healthcare) was performed according to manufacturers' recommendations. Protein concentration was determined using AbS280- The purity of the proteins were analyzed by SDS-PAGE on 4-12 % Novexgels and stained with Coomassie Blue R. [0108] The proteins were successfully expressed and purified in an acceptable yield. The analysis with gel electrophoresis showed that no impurities were present (not shown). Biosensor analysis for binding kinetics to human serum albumin [0109] Biosensor analysis on a Biacore2000 instrument (Biacore) was performed with HSA (SIGMA, cat. no. A3782-5G) immobilized by amine coupling onto the carboxylated dextran layer on surfaces of a CM-5 chip (research grade; Biacore), according to the manufacturer's recommendations. The immobilization of HSA resulted in a signal of 450 resonance units. One cell surface on the chip was activated and deactivated and used as reference cell during injections. The purified Hisg-ABD samples were diluted in HBS-EP (Biacore) to 4, 10, 40, 100 and 400 nM for ABDwt and to 0.2, 0.8, 2, 5 and 20 nM for the selected ABD variants. The samples were injected at a constant flow-rate of 25 pl/min for 10 min, followed by injection of HBS-EP for 3 hours. The surfaces were regenerated with two injections of 20 pi of 5 and 10 mM HCI. The Kd, ka and kq values were estimated and are given in Table 7, confirming the result of Example 2 that molecules exhibiting very high affinities to HSA had been obtained. Table 7: Kinetic parameters (ka, kq and Kd) to HSA of purified ABD molecules

Example 4 Primate immunoaenicitv and pharmacokinetics of a Zvariant polypeptide fused to a first ABD variant Summary [0110] Previous studies in mouse and rat have shown that various Z variant molecules fused to ABDwt generate a lower antibody response compared to the Z variant alone. The aim of this study was 1) to confirm these results in primates and expand it to a mutated variant of ABD exhibiting a 103-fold higher binding affinity for albumin compared to ABDwt, and 2) to compare the serum half-lives of ABD-fused and naked Z variants. A Z variant with affinity for the HER2 receptor was administered to primates with or without the ABD variant as fusion partner. Repeated immunization and bleeding proceeded over a 45-day period. The specific antibody responses against, and the serum half-lives of, the Z variant molecules were analyzed by ELISA assays. Molecules studied [0111] Z00342: a variant of protein Z, in turn derived from the B domain of staphylococcal protein A, with an affinity for the HER2 receptor. This variant was produced by recombinant DNA technology. Purification was performed using anion exchange and reverse phase chromatography methods followed by endotoxin removal on a Detoxi-Gel™ AffinityPak™ Pre-packed Column (Pierce, cat no 2034) according to the manufacturer's instructions. A detailed description of the Z00342 molecule is given in Orlova et at, Cancer Res 66:8, 4339-48 (2006), where it is denoted Z|-|er2:342· [0112] Z00342-ABD00003: a fusion protein between the Z variant Z00342 and the variant ABD molecule ABD00003 selected in Example 2. This fusion protein was produced by recombinant DNA technology. Purification was performed using affinity capture on HSA-sepharose and reverse phase chromatography followed by endotoxin removal as above. Methods [0113] Administration and sampling schemes: The animal study was performed at SMI (Smittskyddsinsitutet) in Solna, Sweden, with permission from the local ethical animal committee (N196/06). The primates were sedated before administration of the test molecules and blood sampling, by an intramuscular administration of ketamine (Ketalar®). 10 individual cynomolgus primates, Macaca fascicularis, divided into two groups were intravenously injected with the test molecules according to the scheme in Table 8. Table 8: Administration of test molecules

[0114] Time points for administration and bleeding are summarized in Table 9. PK refers to samples taken for the pharmacokinetic study. Blood was stored at 4 °C over night and sera were subsequently kept at -20 °C. Table 9: Time points for administration of test molecule and blood sampling

[0115] General ELISA method: In general, a volume of 50 pi per well was used for all incubation steps except for blocking where a volume of 100 pi was used. Plates were coated over night at 4 °C in coating buffer (15 mM Na2CC>3, 35 mM NaHCC>3, pH 9.6) and washed with tap water. Blocking and dilutions were done in PBS with 0.5 % casein. Incubation times at room temperature were 1-2 hours for blocking and serum, 1 hour for secondary antibody and 10 min for substrate solution (ImmunoPure® TMB, Pierce, cat no 34021). Washing with 4x250 pi PBS-T (PBS with 0.05 % Tween 20) per well was carried out between all steps, using an automated ELISA SkanWasher 300 (Skatron). The color reaction was stopped by addition of 50 pi 2 M H2SO4 and plates were read at 450 nm using an Ultra384 plate reader (Tecan) equipped with the Magellan software v3.11 (Tecan). [0116] Anti-Z00342 laG specific ELISA: Plates were coated with 0.3 pg/ml of Z00342 diluted in coating buffer and incubated over night at 4 °C. After washing, plates were blocked as described above. Sera from primates were added in two-fold dilution series starting from 1/100. Purified serum from hyperimmunized primate was used as positive control and added in a two-fold dilution series starting from 8 pg/ml. Following incubation, plates were washed and a secondary, HRP conjugated anti-human IgG antibody (Southern Biotech cat. no. 2040-05) (diluted 1/10000) was added. After the final incubation, plates were washed and developed as described above. [0117] Serum-Z specific ELISA for PK analysis: Plates were coated with 2 pg/ml affinity purified goat anti-Z Ig (produced in-house and specific for an epitope common to all Z variants) and incubated over night at 4 °C. After washing, plates were blocked as described above. Sera from primates injected with Z00342 or Z00342-ABD00003 were added in twe-fold dilution series starting from 1/40 (for Z00342) or 1/80 (for Z00342-ABD00003). Standards of each molecule were added in a two-fold dilution series starting from 20 ng/ml. After incubation, plates were washed and the second step antibody was added (2 pg/ml of a rabbit IgG against Z (produced in-house) for Z00342; 1/5000 of a rabbit IgG against Z-ABD (produced in-house) for Z00342-ABD00003). Following incubation, plates were washed and HRP conjugated anti-rabbit Ig (Dako cat. no. 0448) diluted 1:5000 was added. After the final incubation, plates were washed and developed as described above. Results [0118] laG specific for Z in primates injected with Z00342: The serum from each bleeding was analyzed by ELISA for the presence of IgG specific for Z variants (Figures 9A-9C). Low levels of IgG were detected at day 0 except for one primate (9039) that had moderate levels of pre-formed antibodies. After day 14, the antibody titer increased steadily and reached a maximum at day 28-35 in three of the animals (9039, 10025, 11019), whereas two showed a lowantibody response (9023 and 10105) throughout the 45 day period. [0119] InG snecific for Z in nrimates injected with Z00342-ABD00003: The serum from each bleeding was analyzed by ELISA for the presence of IgG specific for the Z-ABD00003 molecule (Figures 10A-10C). No antibody response was observed in two of the primates (12047 and 12065), whereas two primates (12041 and 12061) showed a high response. The fifth primate (12031) had high pre-serum levels of antibodies that were barely altered during the 45-day period. [0120] Concentration of antibodies specific for Z variants: The concentration of IgG specific for Z variants in the sera was calculated by linear regression using the positive control as standard (Figure 11). Individual variations were seen within each group of primates. At day 45, the median concentration of specific IgG in groups 1 and 2 was 2 and 0.1 units/ml respectively, indicating that the fusion to ABD00003 decreases the antibody response against Z00342. [0121] Pharmacokinetics of Z00342 in serum: The circulation times of Z00342 and Z00342-ABD00003 were compared in a pharmacokinetic analysis. The concentrations of the molecules over time were calculated from standard curves generated from dilution series of known amounts of Z00342 and Z00342-ABD00003, respectively. The results show that ABD-fused Z00342 fused to ABD00003 remains longer in the blood circulation compared to Z00342 alone (Figure 12). Z00342 disappeared from the circulation within 4 hours whereas the ABD00003-fused molecules were still detectable after 7 days. Summary [0122] The results of this study indicate that the ABD-fused Z variant molecule generates a lower immune response as well as exhibiting an extended elimination half-life in comparison with the Z variant molecule without an albumin binding fusion partner. Example 5 Primate immunoaenicitv of a Z variant polypeptide fused to a second ABD variant Molecules studied [0123] In this extension of Example 4, a second variant of ABD with even higher affinity for albumin (Kq= 10'13 M), was fused to a dimeric Z variant and used for an immunogenicity study in primates. [0124] (Z01154)2: Adimeric variant of protein Z, in turn derived from the B domain of staphylococcal protein A. This dimer variant was produced by recombinant DNAtechnology. Purification was performed using anion exchange, reverse phase chromatography and cation exchange methods before endotoxin removal on a Detoxi-Gel™ AffinityPak™ Pre-packed Column (Pierce cat no 2034). A detailed description of the monomeric Z01154 molecule is given in Gunneriusson E et al. Protein Eng 12:10, 873-878 (1999), where it is denoted ZTaq4:1- [0125] (Z01154)2-ABD00239: a fusion protein between the (Z01154)2 dimer and the variant ABD molecule ABD00239 selected in Example 2. This fusion protein was produced by recombinant DNA technology. Purification was performed using affinity capture on HSA-sepharose and reverse phase chromatography followed by endotoxin removal as above. Methods [0126] Administration and sampling schemes: The animal study was performed at SMI (Smittskyddsinsitutet) in Solna, Sweden, with permission from the local ethical animal committee (N196/06). The primates were sedated before administration of test molecule and blood sampling, by an intramuscular administration of ketamine (Ketalar®). 7 individual cynomolgus primates, Macaca fascicularis, divided into two groups of three and four respectively, were intravenously injected with the test molecules as outlined in Table 10. Table 10: Administration of test molecules

[0127] Time points for administration and blood sampling are summarized in Table 11. Table 11: Time points for administration of test molecule and blood sampling

[0128] Analysis of plasma samples bv specific ELISA: Plates were coated with (Z01154)2 or (Z01154)2-ABD00239 diluted in coating buffer (15 mM Na2CC>3, 35 mM NaHCC>3, pH 9.6) to a final concentration of 2 pg/ml. 50 pi solution was used per well and plates were incubated at 4 °C over night. Blocking with PBS + 0.5 % casein was performed at room temperature for 1-2 h. Serum was added in twofold dilution series starting from a 1/100 dilution in blocking buffer. Purified serum from hyperimmunized primate was used as a positive control. Incubation was performed at room temperature for 1-2 hours and washing with 4x175 pi PBS-T (PBS with 0.05 % Tween 20) per well was carried out using an automated ELISASkanWasher 300 (Skatron). Incubation with secondary HRP-conjugated goat anti-human IgG F(ab)2 (Jackson cat. no. 109-035-097), diluted 1:5000 in blocking buffer, was performed at room temperature for 1-2 h. Automated washing was carried out as above. Detection was enabled using ImmunoPure® TMB (Pierce cat. no. 34021) where the reaction was quenched after 12 min by addition of 2 M H2SO4. Plates were read at 450 nm using an Ultra384 plate reader (Tecan) equipped with the Magellan software v3.11 (Tecan). Results [0129] Immunization with 7Z01154f^> and 1Ζ01154ΐ2-ΑΒΡ00239: Sera from primates immunized with naked or ABD00239-fused (Z01154)2 were titrated on (Z01154)2 and (Z01154)2-ABD00239 coated plates respectively, and the titration curves are shown in Figures 13A-13B and Figures 14A-14B. Serum from hyperimmunized monkey was included in the titration as positive control. The absorbance of the positive control at 1600 x dilution was set to 100 % and used for normalization. Figure 15 shows normalized values of the individual responses. The results show that all three animals responded against (Z01154)2, although the response in one primate (E89) was of a lower magnitude. In contrast, only one out of four animals responded against (Z01154)2-ABD00239. REFERENCES CITED IN THE DESCRIPTION This list of references cited by the applicant is for the reader's convenience only. It does not form part of the European patent document. Even though great care has been taken in compiling the references, errors or omissions cannot be excluded and the EPO disclaims all liability in this regard. Patent documents cited in the description • W09101743Af0003] [0055j • WQ0145746Af0003] [0003| [0055| • US20040001327A [0083; • W02005097202A 108041 100581 Non-patent literature cited in the description • PETERS TAdvances in Protein Chemistry, 1985, vol. 37, 161- 100031 . DUGAICZYK L et al.Proc Natl Acad Sci USA, 1982, vol. 79, 71- [0002] • NICHOLSON JP et al.Br J Anaesth, 2000, vol. 85, 599- :00021 • HE XMCARTER DCNature, 1992, vol. 358, 209- £000¾ • SHEFFIELD WPCurr Drug Targets Cardiovacs Haematol Disord, 2001, vol. 1, 1 - [00G2J • SELLERS EMKOCH-WESER MD et al.Albumin Structure, Function and UsesPergamon19770000159- [0003| • DENNIS et al.J Biol Chem, 2002, vol. 277, 35035-43 [0003[ • BJORCK et al.Mol Immunol, 1987, vol. 24, 1113- Γ03051 . GUSS et al.EMBO J, 1986, vol. 5, 1567- [0885; . OLSSON et al.Eur J Biochem, 1987, vol. 168, 319- :0005) • KRAULIS et al.FEBS Lett, 1996, vol. 378, 190- [8805[ • JOHANSSON et al.J Mol Biol, 1997, vol. 266, 859-865 100061 . JOHANSSON et al.J Biol Chem, 2002, vol. 277, 8114-8120 100061 . ROZAK et al.Biochemistry, 2006, vol. 45, 3263-3271 100081 :00511 . NAVARRESCHNEEWINDMMBR, 1999, vol. 63, 174-229 100071 • BRAM BELL et al.Nature, 1964, vol. 203, 1352-£00081 • CHAUDHURY et al.J Exp Med, 2003, vol. 197, 315-£00081 . BRAISTED AC et al.Proc Natl Acad Sci USA, 1996, vol. 93, 5688-5692 10047) . WUSENTERNature Biotechnology, 2005, vol. 23, 1137-46 100571 • GRONWALL et al.J Biotechnol, 2007, vol. 128, 162-183 [000¾ • RljTHERNucleic Acids Res, 1982, vol. 10, 5765-5772 £0061], . NILSSON etal.Prot Eng, 1987, vol. 1, 107-113 [000¾ . LOWENADLER et al.Gene, 1987, vol. 58, 87-97 [0063| • LINHULT et al.Protein Science, 2002, vol. 11,206-213 £0066; . LEJON et al.J Biol Chem, 2004, vol. 279, 42924-42928 [8068] 1. 1. Albuminbindende polypeptid omfattende et albuminbindende motiv, hvilket motiv består af aminosyresekvensen: GVSDX5YKX8X9I XuXi2AX14 TVEGVX20 ALX23X24X25l hvori, uafhængigt af hinanden, X5 er udvalgt blandt Y og F; X8 er udvalgt blandt N, R og S; X9 er udvalgt blandt V, I, L, M, F og Y; Xn er udvalgt blandt N, S, E og D; X12 er udvalgt blandt R og K; Xi4 er udvalgt blandt K og R; X20 er udvalgt blandt D, N, Q, E, H, S og R; X23 er udvalgt blandt K, I og T; X24 er udvalgt blandt A, S, T, G, H, L og D; og X25 er udvalgt blandt Η, E og D; idet det albuminbindende polypeptid binder sig til humant serumalbumin, således at Ko-værdien for interaktionen højst er 1 x 10"9 M.
2. Albuminbindende polypeptid ifølge krav 1, hvor det albuminbindende motiv består af en aminosyresekvens udvalgt blandt SEQ ID NO: 1-4, SEQ ID NO: 15, SEQ ID NO: 18, SEQ ID NO:21, SEQ ID NO:33, SEQ ID NO:37, SEQ ID NO:41, SEQ ID NO:45-46, SEQ ID NO:49-50, SEQ ID NO:53-59, SEQ ID NO:68, SEQ ID NO:71, SEQ ID NO:74, SEQ ID NO:87, SEQ ID NO:89, SEQ ID NO:93, SEQ ID NO:97, SEQ ID N0:100, SEQ ID N0:103-10 SEQ ID NO: 118, SEQ ID NO:121, SEQ ID NO:124, SEQ ID NO:135, SEQ ID NO:137, SEQ ID NO: 139, SEQ ID NO:143, SEQ ID NO:147-148, SEQ ID NO:151-152, SEQ ID NO:155-161, SEQ ID NO: 170, SEQ ID NO: 173, SEQ ID NO: 176, SEQ ID NO: 188, SEQ ID NO: 190, SEQ ID NO: 192, SEQ ID NO:196, SEQ ID N0:200-201, SEQ ID N0:204-205, SEQ ID NO:208-211, SEQ ID NO:221, SEQ ID NO:228, SEQ ID NO:230, SEQ ID NO:234, SEQ ID NO:236, SEQ ID NO:239-246 og SEQ ID NO:248.
3. Albuminbindende polypeptid ifølge krav 2, hvor det albuminbindende motiv består af en aminosyresekvens udvalgt blandt SEQ ID NO:3, SEQ ID NO:53 og SEQ ID NO:239.
4. Albuminbindende polypeptid ifølge et hvilket som helst af de foregående krav, hvor albuminbindingsmotivet indgår i et trehelixbundt-proteindomæne.
5. Albuminbindende polypeptid ifølge krav 4, hvor trehelixbundt-proteindomænet er domæne GA3 i protein G fra Streptococcus-stamme G148.
6. Albuminbindende polypeptid ifølge krav 4, som omfatter følgende aminosyresekvens: LAEAKXaXbAXcXd ELXeKY- [ABM] - LAALP hvori [ABM] er et albuminbindende motiv som defineret i et hvilket som helst af kravene 1-3, og hvor, uafhængigt af hinanden, Xa er udvalgt blandt V og E; Xb er udvalgt blandt L, E og D; Xc er udvalgt blandt N, L og I; Xd er udvalgt blandt R og K; og Xe er udvalgt blandt D og K.
7. Albuminbindende polypeptid ifølge krav 6, hvis aminosyresekvens omfatter en sekvens, som opfylder en blandt nedenstående definitioner: i) den er udvalgt blandt SEQ ID NO:258-261, SEQ ID NO:272, SEQ ID NO:275, SEQ ID NO:278, SEQ ID NO:290, SEQ ID NO:294, SEQ ID NO:298, SEQ ID N0:302-303, SEQ ID N0:306-307, SEQ ID NO:310-316, SEQ ID NO:325, SEQ ID NO:328, SEQ ID NO:331, SEQ ID NO:344, SEQ ID NO:346, SEQ ID NO:350, SEQ ID NO:354, SEQ ID NO:357, SEQ ID NO:360-365, SEQ ID NO:375, SEQ ID NO:378, SEQ ID NO:381, SEQ ID NO:392, SEQ ID NO:394, SEQ ID NO:396, SEQ ID NO:400, SEQ ID N0:404-405, SEQ ID N0:408-409, SEQ ID NO:412-418, SEQ ID NO:427, SEQ ID NO:430, SEQ ID NO:433, SEQ ID NO:445, SEQ ID NO:447, SEQ ID NO:449, SEQ ID NO:453, SEQ ID NO:457-458, SEQ ID NO:461-462, SEQ ID NO:465-468, SEQ ID NO:478, SEQ ID NO:485, SEQ ID NO:487, SEQ ID NO:491, SEQ ID NO:493, SEQ ID NO:496-503 og SEQ ID NO:505; ii) den er en aminosyresekvens med 85 % eller større similaritet med en sekvens ifølge i).
8. Albuminbindende polypeptid ifølge krav 7, hvis aminosyresekvens omfatter en sekvens, som opfylder en blandt nedenstående definitioner: iii) den er udvalgt blandt SEQ ID NO:260, SEQ ID NO:310 og SEQ ID NO:496; iv) den er en aminosyresekvens med 85 % eller større similaritet med en sekvens ifølge iii).
9. Fragment af et albuminbindende polypeptid ifølge et hvilket som helst af de foregående krav, hvilket fragment binder sig til humant serumalbumin, således at interaktionens Ko-værdi højst er 1 x 10"9 M.
10. Multimer af albuminbindende polypeptider eller fragmenter deraf ifølge et hvilket som helst af de foregående krav, omfattende mindst to albuminbindende polypeptider eller fragmenter deraf som monomerenheder.
11. Fusionsprotein eller konjugat omfattende i) en første del bestående af et albuminbindende polypeptid, fragment eller multimer ifølge et hvilket som helst af de foregående krav; og ii) en andel del bestående af et polypeptid med en ønsket biologisk aktivitet.
12. Polypeptid ifølge et hvilket som helst af de foregående krav, yderligere omfattende et mærke.
13. Polynukleotid kodende for et polypeptid ifølge et hvilket som helst af kravene 1-11.
14. Fusionsprotein eller konjugat ifølge krav 11 til anvendelse som lægemiddel, som udviser en halveringstid in vivo, som er længere end halveringstiden in vivo for polypeptidet med ønsket biologisk aktivitet i sig selv.
15. Fusionsprotein eller konjugat ifølge krav 11 til brug som lægemiddel, som fremkalder ingen eller en nedsat immunreaktion efter indgift til pattedyret i sammenligning med den immunreaktion, der fremkaldes efter indgift til pattedyret af polypeptidet med ønsket biologisk aktivitet i sig selv.