Albumin variants

218858/2

ALBUMIN VARIANTS

Background of the Invention

Field of the Invention

The present invention relates to variants of albumin or fragments thereof orfusion polypeptides comprising variant albumin or fragments thereof having a changein half-life compared to the albumin, fragment thereof or fusion polypeptide comprisingalbumin or a fragment thereof.

Description of the Related Art

Albumin is a protein naturally found in the blood plasma of mammals where it isthe most abundant protein. It has important roles in maintaining the desired osmoticpressure of the blood and also in transport of various substances in the blood stream.

Albumins have been characterized from many species including human, pig,mouse, rat, rabbit and goat and they share a high degree of sequence and structuralhomology.

Albumin binds in vivo to its receptor, the neonatal Fc receptor (FcRn) “Brambell”and this interaction is known to be important for the plasma half-life of albumin. FcRnis a membrane bound protein, expressed in many cell and tissue types. FcRn hasbeen found to salvage albumin from intracellular degradation (Roopenian D. C. andAkilesh, S. (2007), Nat. Rev. Immunol 7, 715-725.). FcRn is a bifunctional moleculethat contributes to maintaining a high level of IgGs and albumin in serum in mammalssuch as human beings.

Whilst the FcRn-immunoglobulin (IgG) interaction has been characterized in theprior art, the FcRn-albumin interaction is less well characterized. The major FcRnbinding site is localized within Dill (381-585). Andersen et al (2010). ClinicalBiochemistry 43,367-372. Data indicates that IgG and albumin bind non-cooperativelyto distinct sites on FcRn (Andersen et al. (2006), Eur. J. Immunol 36, 3044-3051;Chaudhury et al. (2006), Biochemistry 45, 4983-4990.).

It is known that mouse FcRn binds IgG from mice and humans whereas humanFcRn appears to be more discriminating (Ober et al. (2001) Int. Immunol 13, 1551-1559). Andersen et al. (2010). Journal of Biological Chemistry 285(7):4826-36,describes the affinity of human and mouse FcRn for each mouse and human albumin(all possible combinations). No binding of albumin from either species was observed atphysiological pH to either receptor. At acidic pH, a 100-fold 1 WO 2011/051489 PCT/EP2010/066572 difference in binding affinity was observed. In all cases, binding of albumin and IgG from eitherspecies to both receptors were additive.

Human serum albumin (HSA) has been well characterized as a polypeptide of 585 aminoacids, the sequence of which can be found in Peters, T., Jr. (1996)/4// about Albumin: Biochemistry,Genetics and Medical, Applications pp10, Academic Press, Inc., Orlando (ISBN 0-12-552110-3). Ithas a characteristic binding to its receptor FcRn, where it binds at pH 6.0 but not at pH 7.4.

The plasma half-life of HSA has been found to be approximately 19 days. A natural varianthaving lower plasma half-life has been identified (Peach, R. J. and Brennan, S. 0.,(1991) BiochimBiophys Acta. 1097:49-54) having the substitution D494N. This substitution generated an N-glycosylation site in this variant, which is not present in the wild-type albumin. It is not knownwhether the glycosylation or the amino acid change is responsible for the change in plasma half-life.

Albumin has a long plasma half-life and because of this property it has been suggested foruse in drug delivery. Albumin has been conjugated to pharmaceutically beneficial compounds (WO2000/69902A), and it was found that the conjugate - maintained the long plasma half-life ofalbumin. The resulting plasma half-life of the conjugate was generally considerably longer than theplasma half-life of the beneficial therapeutic compound alone.

Further, albumin has been fused to therapeutically beneficial peptides (WO 2001/79271 Aand WO 2003/59934 A) with the typical result that the fusion has the activity of the therapeuticallybeneficial peptide and a considerably longer plasma half-life than the plasma half-life of thetherapeutically beneficial peptides alone.

Otagiri et al (2009), Biol. Pharm, Bull. 32(4), 527-534, discloses that 77 albumin variant areknow, of these 25 are found in domain III. A natural variant lacking the last 175 amino acids at thecarboxy termini has been shown to have reduced half-life (Andersen et al (2010), ClinicalBiochemistry 43, 367-372). Iwao et al.(2007) studied the half-life of naturally accuring humanalbumin variants using a mouse model, and found that K541E and K560E had reduced half-life,E501K and E570K had increased half-life and K573E had almost no effect on half-life (Iwao, et. al.(2007) B.B.A. Proteins and Proteomics 1774, 1582-1590).

Galliano et al (1993) Biochim. Biophys.Acta 1225, 27-32 discloses a natural variant E505K.Minchiotti et a/.(1990) discloses a natural variant K536E. Minchiotti et al (1987) Biochim. Biophys.Acta 916, 411-418 discloses a natural variant K574N. Takahashi et al (1987) Proc. Natl. Acad. Sci.USA 84, 4413-4417, discloses a natural variant D550G. Carlson et al (1992).

Proc.Nat.Acad.Sci.USA 89, 8225- 8229, discloses a natural variant D550A.

Albumin has the ability to bind a number of ligands and these become associated(associates) with albumin. This property has been utilized to extend the plasma half-life of drugs2 218858/2 having the ability to noncovalently bind to albumin. This can also be achieved by binding apharmaceutical beneficial compound, which has little or no albumin binding properties, to amoiety having albumin binding properties. See review article and reference therein, Kratz(2008). Journal of Controlled Release 132, 171-183.

Albumin is used in preparations of pharmaceutically beneficial compounds, in whichsuch a preparation maybe for example, but not limited to, a nano particle or micro particleof albumin. In these examples the delivery of a pharmaceutically beneficial compound ormixture of compounds may benefit from alteration in the albumins affinity to its receptorwhere the beneficial compound has been shown to associate with albumin for the means ofdelivery.

It is not clear what determines the plasma half-life of the formed associates (forexample but not limitited to Levemir®, Kurtzhals P et al. Biochem. J. 1995; 312:725-731)conjugates or fusion polypeptides but it appears to be a result of the combination of thealbumin and the selected pharmaceutically beneficial compound/polypeptide. It would bedesirable to be able to control the plasma half-life of given albumin conjugates, associatesor albumin fusion polypeptides so that a longer or shorter plasma half-life can be achievedthan given by the components of the association, conjugation or fusion, in order to be ableto design a particular drug according to the particulars of the indication intended to betreated.

Albumin is known to accumulate and be catabolised in tumours, it has also beenshown to accumulate in inflamed joints of rheumatoid arthritis sufferers. See review articleand reference therein, Kratz (2008). Journal of Controlled Release 132, 171-183. It isenvisaged that HSA variants with increased affinity for FcRn would be advantageous forthe delivery of pharmaceutically beneficial compounds.

It may even be desirable to have variants of albumin that have little or no binding toFcRn in order to provide shorter half-lives or controlled serum pharmacokinetics asdescribed by Kenanova et al (2009) J. Nucl. Med.; 50 (Supplement 2):1582). Albuminfusions with interferon to increase its half-life were disclosed in US 2006/051859. One ofthe albumin molecules disclosed had a Q417E mutation. Sheffield et al (2000) ThrombosisResearch 99, 613-621 disclosed rabbit serum albumin (RSA) with a D494N mutation whichhad reduced half-life compared with wild-type RSA. RSA has only 74.32% sequenceidentity to human serum albumin.

Summary of the Invention

The present invention provides variants of a parent albumin with improvedproperties compared to its parent. In particular the invention provides variants of a parentalbumin having altered plasma half-life compare to its parent.

The present invention relates to isolated variants of albumin or fragments thereof,or fusion polypeptides comprising variant albumin or fragments thereof, of a parentalbumin, comprising an alteration at one or more (several) positions corresponding topositions 417, 440, 464, 490, 492, 493, 494, 495, 496, 499, 500, 501, 503, 504, 505, 506,510, 535, 536, 537, 538, 540, 541, 542, 550, 573, 574, 575, 577, 578, 579, 580, 581, 582and 584 of the mature polypeptide of SEQ ID 3 WO 2011/051489 PCT/EP2010/066572 NO: 2, wherein the variant is not the variant consisting of SEQ ID NO: 2 with the substitutionD494N, E501K, K541E, D550G.A, K573E or K574N.

The alteration at one or more position may independently be selected among substitutions,insertions and deletions, where substitution are preferred.

The present invention also relates to isolated polynucleotides encoding the variants; nucleicacid constructs, vectors, and host cells comprising the polynucleotides; and methods of producingthe variants.

The present invention also relates to conjugates or associates comprising the variantalbumin or fragment thereof according to the invention and a beneficial therapeutic moiety or to afusion polypeptide comprising a variant albumin or fragment thereof of the invention and a fusionpartner polypeptide.

The invention further relates to compositions comprising the variant albumin, fragmentthereof, fusion polypeptide comprising variant albumin or fragment thereof or conjugatescomprising the variant albumin or fragment thereof, according to the invention or associatescomprising the variant albumin or fragment thereof, according to the invention. The compositionsare preferably pharmaceutical compositions.

The invention further relates to a pharmaceutical composition comprising a variant albumin,fragment thereof, fusion polypeptide comprising variant albumin or fragment thereof or conjugatescomprising the variant albumin or fragment thereof, or associates comprising the variant albumin orfragment thereof, wherein said variant albumin, fragment thereof, fusion polypeptide comprisingvariant albumin or fragment thereof or conjugates comprising the variant albumin or fragment orassociates of variant albumin or fragment thereof has altered plasma half-life compared to thecorresponding plasma half-life of the HSA or fragment thereof, fusion polypeptide comprising HSAor fragment thereof or conjugates or associates of HSA or, fragment thereof, comprising HSA orfragment thereof.

Brief Description of the Figures

Figure 1 shows a restriction map of the expression plasmid pDB4082.

Figure 2 shows a restriction map of the expression plasmid pDB2305

Figure 3 shows a restriction map of the expression plasmid pDB4005

Figure 4 shows SPR sensorgrams 10 μΜ albumin injected over shFcRn HSA (JTA) = fattyacid free HSA obtained from Sigma-Aldrich (A3782), HSA (Novozymes) = CommercialRecombinant human serum albumin (RECOMBUMIN).

Figure 5 shows ELISA binding of shFcRn-GST to human serum albumin (HSA) variants(100-0.045 pg/ml). Binding of WT, D494N, D494Q and D494A pH 6.0 and pH 7.4. Binding of WT,4 WO 2011/051489 PCT/EP2010/066572 D494N, D494N/T496A and T496A at pH 6.0 and pH 7.4. Binding of WT, E495Q and E495A at pH6.0 and pH 7.4.

Figure 6 shows representative sensorgrams of binding of 0.2 μΜ .of HSA variants toimmobilized shFcRn (-4600 RU). WT, D494N, D494Q, D494A, D494N/T496A and T496A.

Figure 7 shows representative sensorgrams of binding of 1 μΜ of HSA variants toimmobilized shFcRn (-1400 RU). WT, D494N, D494Q, D494A, D494N/T496A and T496A.

Figure 8 shows relative binding of the HSA variants compared to WT based on twoindependent SPR experiments as shown (A) Figure 6 and (B) Figure 7.

Figure 9 shows ELISA: (A) binding of shFcRn to albumins from human, donkey, bovine,sheep, goat and rabbit at pH 6.0. (B) binding of shFcRn to albumin from guinea pig, hamster, ratand chicken at pH 6.0. (C) binding of shFcRn to albumin from human, donkey, bovine, sheep, goatand rabbit at pH 7.4. (D) binding of shFcRn to albumin from guinea pig, hamster, rat and chicken atpH 7.4. (E) relative binding of the different albumins. Relative binding of human albumin to shFcRnis defined as 1.0. The ELISA values represent the mean of duplicates.

Figure 10 shows SPR: Binding of shFcRn-GST to albumin from several species at pH 6.0and pH 7.4. Representative sensorgrams showing binding of 5.0 μΜ of albumin from differentspecies; (A) human, (B) donkey, (C) bovine, (D) goat, (E) sheep, (F) rabbit, (G) dog, (H) guinea pig, (I) hamster, (J) rat, (K) mouse and (L) chicken. The albumin variants were injected overimmobilized GST-tagged shFcRn (-2100 RU). Injections were performed at 25°C at a rate of 40pl/min.

Figure 11 shows SPR sensorgrams of selected HSA mutants compared with wild-typeHSA. 20 μΜ of (A) WT and P499A (B) WT and K500A, (C) WT and K536A, (D) WT and P537Aand (E) WT and K538A and (F) WT and K537A were injected over immobilized shFcRn at pH 6.0(-1500 RU)

Figure 12 shows SPR sensorgrams of HSA mutants compared with WT HSA. 10 μΜ of (A)WT and K573A (B) WT and K573C, (C) WT and K573F, (D) WT and K573G and (E) WT andK573L and (F) WT and K573M, (G) WT and K573Q, (H) WT and K573R and (I) WT and K573T and (J) WT and K573V injected over immobilized shFcRn at pH 5.5 and pH7.4. Injections wereperformed at 25°C at a flow rate of 80 μΙ/min.

Figure 13 shows SPR sensorgrams of HSA mutants compared with wild-type HSA. 10 μΜof (A) WT and K573D (B) WT and K573E, (C) WT and K573H, (D) WT and K573I and (E) WT andK573N and (F) WT and K573P, (G) WT and K573S, (H) WT and K573* and (I) WT and K573W and(J) WT and K573Y injected over immobilized shFcRn at pH 5.5 and pH7.4. Injections wereperformed at 25°C at a flow rate of 80 μΙ/min. 5 WO 2011/051489 PCT/EP2010/066572

Figure 14 shows SPR sensorgrams of HSA mutants compared with wild-type HSA. 20 μΜof (A) WT and E492G+K538H+K541N+E542D (B) WT and E492T+N503K+K541A, (C)WTandE492P+N503K+K541G+E542P,(D) WT and E492H+E501P+N503H+E505D+T506S+T540S+K541E and (E) WT and A490D+E492T+V493L+E501P+N503D+A504E+E505K+T506F+K541D and (F) WT and E492G+V493P+K538H+K541N+E542D injected overimmobilized shFcRn at pH 6.0. Injections were performed at 25°C at a flow rate of 80 pl/min.

Figure 15 shows SPR sensorgrams of HSA mutants compared with wild-type HSA. TwentyμΜ of (A) WT, (B) H440Q, (C) H464Q and (D) H535Q injected over immobilized shFcRn at pH 6.0.Injections were performed at 25°C at a flow rate of 80 μΙ/min.

Figure 16 shows SPR sensorgrams of HSA mutant K500E compared with wild-type HSA.Ten μΜ of HSA mutant K500E injected over immobilized shFcRn at pH 5.75. Injections wereperformed at 25°C at a flow rate of 30 μΙ/min.

Figure 17 shows a restriction map of the expression plasmid pDB3017

Figure 18 shows a restriction map of the expression plasmid pDB3021

Figure 19 shows a restriction map of the expression plasmid pDB3056

Figure 20 shows a restriction map of the expression plasmid pDB3165

Figure 21 shows a restriction map of the expression plasmid pDB4172

Figure 22 shows a restriction map of the expression plasmid pDB4267

Figure 23 shows a restriction map of the expression plasmid pDB4285

Figure 24 shows a GP-HPLC chromatogram of WT HSA and mutant K573P HRPconjugates for shFcRn analysis. Injections of 25μL were made onto a TSK G3000SWXL column(Tosoh Bioscience) as described in materials and methods.

Figure 25 shows SDS PAGE separation followed by both visual (A) and ultraviolet (B)detection of the Fluorescein conjugated albumin. HSA::F5M (Lane 1), K573P::F5M (Lane 2) andrHA standard (Lane 3).

Figure 26 shows shFcRn binding properties of HSA variants. 10μΜ of WT rHA andE492T(A), WT rHA and D494N/E495Q/T496A(B), WT rHA and N503D(C), WT rHA and N503K(D),WT rHA and E492T/N503D(E), WT rHA and E495Q/T496A(F), WT rHA and K538H(G), WT rHAand E492D(H) injected over immobilised shFcRn at pH5.5

Figure 27 shows shFcRn binding properties of HSA variants. 10μΜ of WT rHA andK541A(I) and WT rHA and K541N(J) were injected over immobilised shFcRn at pH5.5.

Figure 28 shows competitive binding of K573A and K573P measured by injecting shFcRn(100 nM) alone or pre-incubated with different amounts of HSA K573A and K573P overimmobilized HSA (-2500 RU) at pH6.0 6 WO 2011/051489 PCT/EP2010/066572

Figure 29 shows competitive binding of HSA-FLAG variants measured by injecting shFcRn(100 nM) alone or together with different amounts of HSA-FLAG variants over immobilized HSA(-2500 RU) at pH6.0.

Figure 30 shows competitive binding of HSA-IL1Ra variants measured by injecting shFcRn(100 nM) alone or together with different amounts of HSA-IL1Ra variants over immobilized HSA(-2500 RU) at pH6.0

Figure 31 shows competitive binding of scFv-fused HSA variants measured by injectingshFcRn (100 nM) alone or together with different amounts of (A) scFv-HSA-FLAG variants or (B)HSA-scFv-FLAG variants over immobilized HSA (-2500 RU) at pH6.0.

Figure 32 shows binding of HSA, single, double and triple mutant variants to shFcRn.Samples of 10 μΜ of each HSA variant were injected over immobilized shFcRn at pH 5.5 or pH 7.4.

Detailed Description of the Invention

The present invention relates to isolated variants of albumin or fragments thereof, or fusionpolypeptides comprising variant albumin or fragments thereof, of a parent albumin, comprising analteration at one or more (several) positions corresponding to positions 417, 440, 464, 490, 492,493, 494, 495, 496, 499, 500, 501, 503, 504, 505, 506, 510, 535, 536, 537, 538, 540, 541, 542,550, 573, 574, 575, 577, 578, 579, 580, 581, 582 and 584 of the mature polypeptide of SEQ IDNO: 2, wherein the variant is not the variant consisting of SEQ ID NO: 2 with the substitutionD494N, E501K, K541E, D550G,A, K573E or K574N..

The alteration at one or more position may independently be selected among substitutions,insertions and deletions, where substitution are preferred.

Definitions

Variant: The term “variant” means a polypeptide derived from a parent albumin by one ormore alteration(s), i.e., a substitution, insertion, and/or deletion, at one or more (several) positions.A substitution means a replacement of an amino acid occupying a position with a different aminoacid; a deletion means removal of an amino acid occupying a position; and an insertion meansadding 1 or more, preferably 1-3 amino acids immediately adjacent to an amino acid occupying aposition.

Mutant: The term “mutant” means a polynucleotide encoding a variant.

Wild-Type Albumin: The term “wild-type” (WT) albumin means albumin having the sameamino acid sequence as naturally found in an animal or in a human being.

Parent or Parent albumin The term “parent” or “parent albumin” means an albumin towhich an alteration is made by the hand of man to produce the albumin-variants of the present7 WO 2011/051489 PCT/EP2010/066572 invention. The parent may be a naturally occurring (wild-type) polypeptide or an allele thereof, oreven a variant thereof.

FcRn and shFcRn: The term “FcRn” means the human neonatal Fc receptor (FcRn).shFcRn is a soluble recombinant form of FcRn. smFcRn: The term “smFcRn” is a soluble recombinant form of the mouse neonatal FcReceptor.

Isolated variant: The term “isolated variant” means a variant that is modified by the hand ofman and separated completely or partially from at least one component with which it naturallyoccurs. In one aspect, the variant is at least 1% pure, e.g., at least 5% pure, at least 10% pure, atleast 20% pure, at least 40%> pure, at least 60% pure, at least 80% pure, and at least 90% pure, asdetermined by SDS-PAGE or GP-HPLC.

Substantially pure variant: The term “substantially pure variant” means a preparation thatcontains at most 10%, at most 8%, at most 6%, at most 5%>, at most 4%, at most 3%, at most 2%,at most 1%, and at most 0.5% by weight of other polypeptide material with which it is natively orrecombinantly associated. Preferably, the variant is at least 92% pure, e.g., at least 94% pure, atleast 95% pure, at least 96% pure, at least 97% pure, at least 98% pure, at least 99%, at least99.5% pure, and 100% pure by weight of the total polypeptide material present in the preparation.The variants of the present invention are preferably in a substantially pure form. This can beaccomplished, for example, by preparing the variant by well known recombinant methods and bypurification methods.

Mature polypeptide: The term “mature polypeptide” means a polypeptide in its final formfollowing translation and any post-translational modifications, such as N-terminal processing,C-terminal truncation, glycosylation, phosphorylation, etc. In one aspect, the mature polypeptide isamino acids 1 to 585 of SEQ ID NO: 2, with the inclusion of any post-translational modifications.

Mature polypeptide coding sequence: The term “mature polypeptide coding sequence”means a polynucleotide that encodes a mature albumin polypeptide. In one aspect, the maturepolypeptide coding sequence is nucleotides 1 to 1758 of SEQ ID NO: 1.

Sequence Identity: The relatedness between two amino acid sequences or between twonucleotide sequences is described by the parameter “sequence identity”.

For purposes of the present invention, the degree of sequence identity between two aminoacid sequences is determined using the Needleman-Wunsch algorithm (Needleman and Wunsch,1970, J. Mol. Biol. 48: 443-453) as implemented in the Needle program of the EMBOSS package(EMBOSS: The European Molecular Biology Open Software Suite, Rice etal., 2000, Trends Genet.16: 276-277), preferably version 3.0.0 or later. The optional parameters used are gap open penaltyof 10, gap extension penalty of 0.5, and the EBLOSUM62 (EMBOSS version of BLOSUM62) 8 WO 2011/051489 PCT/EP2010/066572 substitution matrix. The output of Needle labelled “longest identity” (obtained using the -nobriefoption) is used as the percent identity and is calculated as follows: (Identical Residues x 100)/(Length of Alignment - Total Number of Gaps in Alignment)

For purposes of the present invention, the degree of sequence identity between twodeoxyribonucleotide sequences is determined using the Needleman-Wunsch algorithm(Needleman and Wunsch, 1970, supra) as implemented in the Needle program of the EMBOSSpackage (EMBOSS: The European Molecular Biology Open Software Suite, Rice et al., 2000,supra), preferably version 3.0.0 or later. The optional parameters used are gap open penalty of 10,gap extension penalty of 0.5, and the EDNAFULL (EMBOSS version of NCBI NUC4.4) substitutionmatrix. The output of Needle labeled “longest identity” (obtained using the -nobrief option) is usedas the percent identity and is calculated as follows: (Identical Deoxyribonucleotides x 100)/(Length of Alignment - Total Number of Gaps in Alignment)

Fragment: The term “fragment” means a polypeptide having one or more (several) aminoacids deleted from the amino and/or carboxyl terminus of an albumin and/or an internal region ofalbumin that has retained the ability to bind to FcRn. Fragments may consist of one uninterruptedsequence derived from HSA or it may comprise two or more sequences derived from HSA. Thefragments according to the invention have a size of more than approximately 20 amino acidresidues, preferably more than 30 amino acid residues, more preferred more than 40 amino acidresidues, more preferred more than 50 amino acid residues, more preferred more than 75 aminoacid residues, more preferred more than 100 amino acid residues, more preferred more than 200amino acid residues, more preferred more than 300 amino acid residues, even more preferredmore than 400 amino acid residues and most preferred more than 500 amino acid residues.

Allelic variant: The term “allelic variant” means any of two or more alternative forms of agene occupying the same chromosomal locus. Allelic variation arises naturally through mutation,and may result in polymorphism within populations. Gene mutations can be silent (no change in theencoded polypeptide) or may encode polypeptides having altered amino acid sequences. An allelicvariant of a polypeptide is a polypeptide encoded by an allelic variant of a gene.

Coding sequence: The term “coding sequence" means a polynucleotide, which directlyspecifies the amino acid sequence of its translated polypeptide product. The boundaries of thecoding sequence are generally determined by an open reading frame, which usually begins with theATG start codon or alternative start codons such as GTG and TTG and ends with a stop codonsuch as TAA, TAG, and TGA. The coding sequence may be a DNA, cDNA, synthetic, orrecombinant polynucleotide. cDNA: The term "cDNA" means a DNA molecule that can be prepared by reversetranscription from a mature, spliced, mRNA molecule obtained from a eukaryotic cell. cDNA lacks9 WO 2011/051489 PCT/EP2010/066572 intron sequences that may be present in the corresponding genomic DNA. The initial, primary RNAtranscript is a precursor to mRNA that is processed through a series of steps, including splicing,before appearing as mature spliced mRNA.

Nucleic acid construct: The term "nucleic acid construct" means a nucleic acid molecule,either single- or double-stranded, which is isolated from a naturally occurring gene or is modified tocontain segments of nucleic acids in a manner that would not otherwise exist in nature or which issynthetic. The term nucleic acid construct is synonymous with the term “expression cassette” whenthe nucleic acid construct contains the control sequences required for expression of a codingsequence of the present invention.

Control sequences: The term “control sequences” means all components necessary forthe expression of a polynucleotide encoding a variant of the present invention. Each controlsequence may be native or foreign to the polynucleotide encoding the variant or native or foreign toeach other. Such control sequences include, but are not limited to, a leader, polyadenylationsequence, propeptide sequence, promoter, signal peptide sequence, and transcription terminator.At a minimum, the control sequences include a promoter, and transcriptional and translational stopsignals. The control sequences may be provided with linkers for the purpose of introducing specificrestriction sites facilitating ligation of the control sequences within the coding region of thepolynucleotide encoding a variant.

Operably linked: The term “operably linked” means a configuration in which a controlsequence is placed at an appropriate position relative to the coding sequence of a polynucleotidesuch that the control sequence directs the expression of the coding sequence.

Expression: The term “expression” includes any step involved in the production of thevariant including, but not limited to, transcription, post-transcriptional modification, translation, post-translational modification, and secretion.

Expression vector: The term “expression vector” means a linear or circular DNA moleculethat comprises a polynucleotide encoding a variant and is operably linked to additional nucleotidesthat provide for its expression.

Host cell: The term "host cell" means any cell type that is susceptible to transformation,transfection, transduction, and the like with a nucleic acid construct or expression vector comprisinga polynucleotide of the present invention. The term “host cell” encompasses any progeny of aparent cell that is not identical to the parent cell due to mutations that occur during replication.

Plasma half-life: Plasma half-life is ideally determined in vivo in suitable individuals.However, since it is time consuming and expensive and there inevitable are ethical concernsconnected with doing experiments in animals or man it is desirable to use an in vitro assay fordetermining whether plasma half-life is extended or reduced. It is known that the binding of albumin10 WO 2011/051489 PCT/EP2010/066572 to its receptor FcRn is important for plasma half-life and the correlation between receptor bindingand plasma half-life is that a higher affinity of albumin to its receptor leads to longer plasma half-life.Thus for the present invention a higher affinity of albumin to FcRn is considered indicative of anincreased plasma half-life and a lower affinity of albumin to its receptor is considered indicative of areduced plasma half-life.

In this application and claims the binding of albumin to its receptor FcRn is described usingthe term affinity and the expressions “stronger” or “weaker”. Thus, it should be understood that amolecule having a higher affinity to FcRn than HSA is considered to bind stronger to FcRn thanHSA and a molecule having a lower affinity to FcRn than HSA is considered to bind weaker toFcRn than HSA.

The terms “longer plasma half-life” or “shorter plasma half-life” and similar expressions areunderstood to be in relationship to the corresponding parent albumin molecule. Thus, a longerplasma half-life with respect to a variant albumin of the invention means that the variant has longerplasma half-life than the corresponding albumin having the same sequences except for thealteration(s) in positions corresponding to 417, 440, 464, 490, 492, 493, 494, 495, 496, 499, 500,501, 503, 504, 505, 506, 510, 535, 536, 537, 538, 540, 541, 542, 550 , 573, 574, 575, 577, 578,579, 580, 581,582 and 584 in SEQ ID NO: 2.

Conventions for Designation of Variants

For purposes of the present invention, the mature polypeptide disclosed in SEQ ID NO: 2 isused to determine the corresponding amino acid residue in another albumin. The amino acidsequence of another albumin is aligned with the mature polypeptide disclosed in SEQ ID NO: 2,and based on the alignment, the amino acid position number corresponding to any amino acidresidue in the mature polypeptide disclosed in SEQ ID NO: 2 is determined using the Needleman-Wunsch algorithm (Needleman and Wunsch, 1970, J. Mol. Biol. 48: 443-453) as implemented inthe Needle program of the EMBOSS package (EMBOSS: The European Molecular Biology OpenSoftware Suite, Rice etal., 2000, Trends Genet. 16: 276-277), preferably version 3.0.0 or later.

Identification of the corresponding amino acid residue in another albumin can be confirmedby an alignment of multiple polypeptide sequences using “ClustalW” (Larkin et al., 2007,Bioinformatics 23: 2947-2948).

When the other polypeptide (or protein) has diverged from the mature polypeptide of SEQID NO: 2 such that traditional sequence-based comparison fails to detect their relationship (Lindahland Elofsson, 2000, J. Mol. Biol. 295: 613-615), other pairwise sequence comparison algorithmscan be used. Greater sensitivity in sequence-based searching can be attained using searchprograms that utilize probabilistic representations of polypeptide families (profiles) to search 11 WO 2011/051489 PCT/EP2010/066572 databases. For example, the PSI-BLAST program generates profiles through an iterative databasesearch process and is capable of detecting remote homologs (Atschul et al., 1997, Nucleic AcidsRes. 25: 3389-3402). Even greater sensitivity can be achieved if the family or superfamily for thepolypeptide has one or more representatives in the protein structure databases. Programs such asGenTHREADER (Jones, 1999, J. Mol. Biol. 287: 797-815; McGuffin and Jones, 2003,Bioinformatics 19: 874-881) utilize information from a variety of sources (PSI-BLAST, secondarystructure prediction, structural alignment profiles, and solvation potentials) as inputs to a neuralnetwork that predicts the structural fold for a query sequence. Similarly, the method of Gough etal.,2000, J. Mol. Biol. 313: 903-919, can be used to align a sequence of unknown structure within thesuperfamily models present in the SCOP database. These alignments can in turn be used togenerate homology models for the polypeptide, and such models can be assessed for accuracyusing a variety of tools developed for that purpose.

For proteins of known structure, several tools and resources are available for retrieving andgenerating structural alignments. For example the SCOP superfamilies of proteins have beenstructurally aligned, and those alignments are accessible and downloadable. Two or more proteinstructures can be aligned using a variety of algorithms such as the distance alignment matrix (Holmand Sander, 1998, Proteins 33: 88-96) or combinatorial extension (Shindyalov and Bourne, 1998,Protein Engineering 11: 739-747), and implementations of these algorithms can additionally beutilized to query structure databases with a structure of interest in order to discover possiblestructural homologs (e.g., Holm and Park, 2000, Bioinformatics 16: 566-567).

In describing the albumin variants of the present invention, the nomenclature describedbelow is adapted for ease of reference. The accepted IUPAC single letter or three letter amino acidabbreviation is employed.

Substitutions. For an amino acid substitution, the following nomenclature is used: Originalamino acid, position, substituted amino acid. Accordingly, for example the substitution of threoninewith alanine at position 226 is designated as “Thr226Ala” or “T226A”. Multiple mutations areseparated by addition marks (“+”), e.g., “Gly205Arg + Ser411Phe” or “G205R + S411F”,representing substitutions at positions 205 and 411 of glycine (G) with arginine (R) and serine (S)with phenylalanine (F), respectively. The Figures also use (“/”), e.g., “E492T/N503D” this should beviewed as interchangeable with (“+”).

Deletions. For an amino acid deletion, the following nomenclature is used: Original aminoacid, position*. Accordingly, the deletion of glycine at position 195 is designated as “Gly195*” or“G195*”. Multiple deletions are separated by addition marks (“+”), e.g., “Gly195* + Ser411*” or“G195* + S411*”. 12 WO 2011/051489 PCT/EP2010/066572

Insertions. For an amino acid insertion, the following nomenclature is used: Original aminoacid, position, original amino acid, inserted amino acid. Accordingly the insertion of lysine afterglycine at position 195 is designated “Gly195GlyLys” or “G195GK”. An insertion of multiple aminoacids is designated [Original amino acid, position, original amino acid, inserted amino acid #1,inserted amino acid #2; etc.]. For example, the insertion of lysine and alanine after glycine atposition 195 is indicated as “Gly195GlyLysAla” or “G195GKA”.

In such cases the inserted amino acid residue(s) are numbered by the addition of lowercase letters to the position number of the amino acid residue preceding the inserted amino acidresidue(s). In the above example, the sequence would thus be:

Parent: Variant: 195 195 195a 195b G G - K - A

Multiple alterations. Variants comprising multiple alterations are separated by additionmarks (“+”), e.g., “Arg170Tyr+Gly195Glu” or “R170Y+G195E” representing a substitution oftyrosine and glutamic acid for arginine and glycine at positions 170 and 195, respectively.

Different substitutions. Where different substitutions can be introduced at a position, thedifferent substitutions are separated by a comma, e.g., “Arg170Tyr,Glu” represents a substitution ofarginine with tyrosine or glutamic acid at position 170. Thus, “Tyr167Gly,Ala + Arg170Gly,Ala”designates the following variants: “Tyr1 67Gly + Arg1 70Gly”, “Tyr167Gly + Arg 170Ala”, “Tyr167Ala+Arg 1 70Gly”, and“Tyr167Ala+Arg170Ala”.

Parent albumin

Albumins are proteins and constitute the most abundant protein in plasma in mammals andalbumins from a long number of mammals have been characterized by biochemical methods and/orby sequence information. Several albumins, e.g., human serum albumin (HSA), have also beencharacterized crystallographically and the structure determined. HSA is a preferred albumin according to the invention and is a protein consisting of 585amino acid residues and has a molecular weight of 67 kDa. In its natural form it is not glycosylated.The amino acid sequence of HSA is shown in SEQ ID NO: 2. The skilled person will appreciate thatnatural alleles may exist having essentially the same properties as HSA but having one or moreamino acid changes compared to SEQ ID NO: 2, and the inventors also contemplate the use ofsuch natural alleles as parent albumin according to the invention. 13 WO 2011/051489 PCT/EP2010/066572

Albumins have generally a long plasma half-life of approximately 20 days or longer, e.g.,HSA has a plasma half-life of 19 days. It is known that the long plasma half-life of HSA is mediatedvia interaction with its receptor FcRn, however, an understanding or knowledge of the exactmechanism behind the long half-life of HSA is not essential for the present invention.

According to the invention the term “albumin” means a protein having the same, or verysimilar three dimensional structure as HSA and having a long plasma half-life. As examples ofalbumin proteins according to the invention can be mentioned human serum albumin, primateserum albumin, (such as chimpanzee serum albumin, gorilla serum albumin), rodent serum albumin(such as hamster serum albumin, guinea pig serum albumin, mouse albumin and rat serumalbumin), bovine serum albumin, equine serum albumin, donkey serum albumin, rabbit serumalbumin, goat serum albumin, sheep serum albumin, dog serum albumin, chicken serum albuminand pig serum albumin. HSA as disclosed in SEQ ID NO: 2 or any naturally occurring allele thereof,is the preferred albumin according to the invention.

The parent albumin, a fragment thereof, or albumin part of a fusion polypeptide comprisingalbumin or a fragment thereof according to the invention has generally a sequence identity to thesequence of HSA shown in SEQ ID NO: 2 of at least 60%, preferably at least 70%, preferably atieast 80%, preferably at ieast 85%, preferably at least 86%, preferably at least 87%, preferably atleast 88%, preferably at least 89%, preferably at least 90%, preferably at least 91%, preferably atleast 92%, preferably at least 93%, preferably at least 94%, preferably at least 95%, more preferredat least 96%, more preferred at least 97%, more preferred at least 98% and most preferred at least99%.

The parent preferably comprises or consists of the amino acid sequence of SEQ ID NO: 2.In another aspect, the parent comprises or consists of the mature polypeptide of SEQ ID NO: 2.

In another embodiment, the parent is an allelic variant of the mature polypeptide of SEQ IDNO: 2.

In a second aspect, the parent is encoded by a polynucleotide that hybridizes under verylow stringency conditions, low stringency conditions, medium stringency conditions, medium-highstringency conditions, high stringency conditions, or very high stringency conditions with (i) themature polypeptide coding sequence of SEQ ID NO: 1, (ii) the mature polypeptide coding sequenceof SEQ ID NO: 1, or (iii) the full-length complementary strand of (i) or (ii) (J. Sambrook, E.F. Fritsch,and T. Maniatis, 1989, Molecular Cloning, A Laboratory Manual, 2d edition, Cold Spring Harbor,New York).

The polynucleotide of SEQ ID NO: 1 or a subsequence thereof, as well as the amino acidsequence of SEQ ID NO: 2 or a fragment thereof, may be used to design nucleic acid probes toidentify and clone DNA encoding a parent from strains of different genera or species according to14 WO 2011/051489 PCT/EP2010/066572 methods well known in the art. In particular, such probes can be used for hybridization with thegenomic or cDNA of the genus or species of interest, following standard Southern blottingprocedures, in order to identify and isolate the corresponding gene therein. Such probes can beconsiderably shorter than the entire sequence, but should be at least 14, e.g., at least 25, at least35, or at least 70 nucleotides in length. Preferably, the nucleic acid probe is at least 100 nucleotidesin length, e.g., at least 200 nucleotides, at least 300 nucleotides, at least 400 nucleotides, at least500 nucleotides, at least 600 nucleotides, at least 700 nucleotides, at least 800 nucleotides, or atleast 900 nucleotides in length. Both DNA and RNA probes can be used. The probes are typicallylabelled for detecting the corresponding gene (for example, with 32P, 3H, 35S, biotin, or avidin). Suchprobes are encompassed by the present invention. A genomic DNA or cDNA library prepared from such other organisms may be screened forDNA that hybridizes with the probes described above and encodes a parent. Genomic or otherDNA from such other organisms may be separated by agarose or polyacrylamide gelelectrophoresis, or other separation techniques. DNA from the libraries or the separated DNA maybe transferred to and immobilized on nitrocellulose or other suitable carrier material. In order toidentify a clone or DNA that is homologous with SEQ ID NO: 1 or a subsequence thereof, thecarrier material is used in a Southern blot.

For purposes of the present invention, hybridization indicates that the polynucleotidehybridizes to a labelled nucleotide probe corresponding to the polynucleotide shown in SEQ IDNO: 1, its complementary strand, or a subsequence thereof, under low to very high stringencyconditions. Molecules to which the probe hybridizes can be detected using, for example, X-ray filmor any other detection means known in the art.

In one aspect, the nucleic acid probe is the mature polypeptide coding sequence of SEQ IDNO: 1. In another aspect, the nucleic acid probe is nucleotides 1 to 1785 of SEQ ID NO: 1. Inanother aspect, the nucleic acid probe is a polynucleotide that encodes the polypeptide of SEQ IDNO: 2 or a fragment thereof. In another aspect, the nucleic acid probe is SEQ ID NO: 1.

For long probes of at least 100 nucleotides in length, very low to very high stringencyconditions are defined as prehybridization and hybridization at 42°C in 5X SSPE, 0.3% SDS, 200micrograms/ml sheared and denatured salmon sperm DNA, and either 25% formamide for ver/ lowand low stringencies, 35% formamide for medium and medium-high stringencies, or 50%formamide for high and very high stringencies, following standard Southern blotting procedures for12 to 24 hours optimally. The carrier material is finally washed three times each for 15 minutesusing 2X SSC, 0.2% SDS at 45°C (very low stringency), 50°C (low stringency), 55°C (mediumstringency), 60°C (medium-high stringency), 65°C (high stringency), or70°C (very high stringency). 15 WO 2011/051489 PCT/EP2010/066572

For short probes that are about 15 nucleotides to about 70 nucleotides in length, stringencyconditions are defined as prehybridization and hybridization at about 5°C to about 10°C below thecalculated Tm using the calculation according to Bolton and McCarthy (1962, Proc. Natl. Acad. Sci.USA 48: 1390) in 0.9 M NaCI, 0.09 M Tris-HCI pH 7.6, 6 mM EDTA, 0.5% NP-40, 1X Denhardt'ssolution, 1 mM sodium pyrophosphate, 1 mM sodium monobasic phosphate, 0.1 mM ATP, and 0.2mg of yeast RNA per ml following standard Southern blotting procedures for 12 to 24 hoursoptimally. The carrier material is finally washed once in 6X SCC plus 0.1% SDS for 15 minutes andtwice each for 15 minutes using 6X SSC at 5°C to 10°C below the calculated Tm.

In a third aspect, the parent is encoded by a polynucleotide with a sequence identity to themature polypeptide coding sequence of SEQ ID NO: 1 of at least 60%, e.g., at least 65%, at least70%, at least 75%, at least, 80%, at least 85%, at least 90%, at least 95%, at least 96%, at least97%, at least 98%, at least 99%, or 100%, which encodes a polypeptide which is able to function asan albumin. In an embodiment, the parent is encoded by a polynucleotide comprising or consistingof SEQ ID NO: 1.

Preparation of Variants

In a further aspect the invention relates to a method for preparing a variant albumin,fragment thereof, or fusion polypeptide comprising variant albumin or a fragment thereof comprisingthe steps of: a. Identifying one or more amino acid residue positions being important for the binding ofalbumin to FcRn, in an albumin or a fragment thereof or the albumin part of a fusionpolypeptide comprising albumin or a fragment thereof; b. Providing a nucleic acid encoding said albumin, the fragment thereof or the albuminpart of a fusion polypeptide comprising albumin or the fragment thereof; c. Modifying the nucleic acid provided in b., so that the one or more (several) amino acidresidue located at the positions identified in a., are deleted or substituted or insertedwith a different amino acid; d. Expressing the modified nucleic acid in a suitable host cell; and e. Recovering the variant albumin, the fragment thereof or the fusion polypeptidecomprising variant albumin or the fragment thereof.

The identification of one or more amino acid residue positions being important for thebinding of albumin to FcRn, in albumin, fragment thereof or the albumin part of a fusion polypeptidecan be done in several ways including, but not limited to, random mutagenesis followed by analysisof the generated mutants and comparison with the non-mutated parent molecule, and identification 16 WO 2011/051489 PCT/EP2010/066572 based on structural considerations optionally followed by generation of variants having theidentified alterations and comparison with the non-mutated patent molecule. A preferred method for identification of one or more amino acid residue positions to bechanged to in order to prepare a variant HSA having an altered binding to FcRn compared withnatural HSA, comprises the following steps: i) Identifying a non-human albumin having a different binding property to FcRn; ii) Identifying the amino acid residues of the human serum albumin interacting withFcRn; iii) Comparing the primary and/or the tertiary structure of the identified non-humanalbumin and human serum albumin with respect to the amino acid residuesidentified in step ii) and identifying the amino acid residues that differ between saidnon-human albumin and human serum albumin as being responsible for theobserved binding difference; and iv) Optionally preparing variants of HSA at the positions identified in step iii) andconfirming that the prepared variants have altered binding to FcRn compared withHSA.

Step i) above may be done using the SPR assay described below. However, the skilledperson will appreciate that other methods may be used to identify non-human albumins havingdifferent binding properties to FcRn than HSA, and that the method is not dependent on how thenon- human albumin, having different binding properties to FcRn, has been identified.

In one preferred embodiment the identified non-human albumin has a stronger binding toFcRn than HSA. Examples of non-human albumins having stronger binding to FcRn than HSAinclude donkey serum albumin, rabbit serum albumin, dog serum albumin, hamster serum albumin,guinea pig serum albumin, mouse serum albumin and rat serum albumin. Step ii) may beaccomplished by considering the structure of FcRn, HSA and the binding complex of these two. Inthe absence of an available structure of the binding complex it is possible to use a model where theHSA structure is docked into the structure of the FcRn structure and thereby identify amino acidresidues of HSA interacting with FcRn.

In another preferred embodiment the identified non-human albumin has a weaker binding toFcRn than HSA. Examples of non-human albumins having weaker binding to FcRn than HSAinclude bovine serum albumin, goat serum albumin, sheep serum albumin and chicken serumalbumin. Step ii) may be accomplished by considering the structure of FcRn, HSA and the bindingcomplex of these two. In absence of an available structure of the binding complex it is possible touse a model where the HSA structure is docked into the structure of the FcRn structure and therebyidentify residues of HSA interacting with FcRn. 17 WO 2011/051489 PCT/EP2010/066572

In this invention and claims, an amino acid residues of HSA interacting with FcRn isconsidered any amino acid residues of HSA being located less than 10A from an amino acid in theFcRn or any amino acid residue that is involved in a hydrogen bond, a salt bridge or a polar ornonpolar interaction with an amino acid residue that is located less than 10Afrom an amino acid inthe FcRn. Preferably the amino acid in HSA residues are located less than 10Afrom amino acids inthe FcRn, more preferred less than 6A from amino acids in the FcRn and most preferred less than3A from amino acids in the FcRn.

Step iii) and iv) can be done using techniques well known to the skilled person.

The present invention also relates to methods for obtaining a variant albumin or fragmentsthereof, or fusion polypeptides comprising the variant albumin or fragments thereof, or associatesof variant albumin or fragment thereof comprising: (a) introducing into a parent albumin orfragments thereof, or fusion polypeptides comprising the parent albumin or fragments thereof analteration at one or more (several) positions corresponding to positions 417, 440, 464, 490, 492,493, 494, 495, 496, 499, 500, 501, 503, 504, 505, 506, 510, 535, 536, 537, 538, 540, 541, 542,550, 573, 574, 575, 577, 578, 579, 580, 581, 582 and 584 of the mature polypeptide of SEQ IDNO: 2; and (b) recovering the variant albumin or fragments thereof, or fusion polypeptidescomprising the variant albumin or fragments thereof.

The variants can be prepared by those skilled persons using any mutagenesis procedureknown in the art, such as site-directed mutagenesis, synthetic gene construction, semi-syntheticgene construction, random mutagenesis, shuffling, etc.

Site-directed mutagenesis is a technique in which one or more (several) mutations arecreated at one or more defined sites in a polynucleotide encoding the parent.

Site-directed mutagenesis can be accomplished in vitro by PCR involving the use ofoligonucleotide primers containing the desired mutation. Site-directed mutagenesis can also beperformed in vitro by cassette mutagenesis involving the cleavage by a restriction enzyme at a sitein the plasmid comprising a polynucleotide encoding the parent and subsequent ligation of anoligonucleotide containing the mutation in the polynucleotide. Usually the restriction enzyme thatdigests at the plasmid and the oligonucleotide is the same, permitting ligation of the plasmid andinsert to one another. See, e.g, Scherer and Davis, 1979, Proc. Natl. Acad. Sci. USA 76: 4949-4955; and Barton etal, 1990, Nucleic Acids Res. 18: 7349-4966.

Site-directed mutagenesis can also be accomplished in vivo by methods known in the art.See, e.g, U.S. Patent Application Publication No. 2004/0171154; Storici et al, 2001, NatureBiotechnol. 19: 773-776; Kren etal, 1998, Nat. Med. 4: 285-290; and Calissano and Macino, 1996,Fungal Genet. Newslett. 43: 15-16. 18 WO 2011/051489 PCT/EP2010/066572

Any site-directed mutagenesis procedure can be used in the present invention. There aremany commercial kits available that can be used to prepare variants.

Synthetic gene construction entails in vitro synthesis of a designed polynucleotide moleculeto encode a polypeptide of interest. Gene synthesis can be performed utilizing a number oftechniques, such as the multiplex microchip-based technology described by Tian et al. (2004,Nature 432: 1050-1054) and similar technologies wherein olgionucleotides are synthesized andassembled upon photo-programable microfluidic chips.

Single or multiple amino acid substitutions, deletions, and/or insertions can be made andtested using known methods of mutagenesis, recombination, and/or shuffling, followed by arelevant screening procedure, such as those disclosed by Reidhaar-Olson and Sauer, 1988,Science 241: 53-57; Bowie and Sauer, 1989, Proc. Natl. Acad. Sci. USA 86: 2152-2156;WO 95/17413; or WO 95/22625. Other methods that can be used include error-prone PCR, phagedisplay (e.g., Lowman et al., 1991, Biochemistry 30: 10832-10837; U.S. Patent No. 5,223,409;WO 92/06204) and region-directed mutagenesis (Derbyshire et al., 1986, Gene 46: 145; Ner et al.,1988, DNA 7: 127).

Mutagenesis/shuffling methods can be combined with high-throughput, automatedscreening methods to detect activity of cloned, mutagenized polypeptides expressed by host cells(Ness et al., 1999, Nature Biotechnology 17: 893-896). Mutagenized DNA molecules that encodeactive polypeptides can be recovered from the host cells and rapidly sequenced using standardmethods in the art. These methods allow the rapid determination of the importance of individualamino acid residues in a polypeptide.

Semi-synthetic gene construction is accomplished by combining aspects of synthetic geneconstruction, and/or site-directed mutagenesis, and/or random mutagenesis, and/or shuffling. Semi-synthetic constuction is typified by a process utilizing polynucleotide fragments that aresynthesized, in combination with PCR techniques. Defined regions of genes may thus besynthesized de novo, while other regions may be amplified using site-specific mutagenic primers,while yet other regions may be subjected to error-prone PCR or non-error prone PCR amplification.Polynucleotide sub sequences may then be shuffled.

Variants

The present invention also provides variant albumins or fragments thereof, or fusionpolypeptides comprising the variant albumin or fragments thereof, of a parent albumin, comprisingan alteration at one or more (several) positions corresponding to positions 417, 440, 464, 490, 492,493, 494, 495, 496, 499, 500, 501, 503, 504, 505, 506, 510, 535, 536, 537, 538, 540, 541, 542,550, 573, 574, 575, 577, 578, 579, 580, 581, 582 and 584 in SEQ ID NO: 2, wherein each 19 WO 2011/051489 PCT/EP2010/066572 alteration is independently a substitution, insertion or deletion with the provision that the and thevariant is not SEQ ID NO: 2 having the substitution D494N, E501K, K541E, D550GA K573E orK574N..

The variant albumin, a fragment thereof, or albumin part of a fusion polypeptide comprisingvariant albumin or a fragment thereof according to the invention has generally a sequence identitythe sequence of HSA shown in SEQ ID NO: 2 of at least 60%, preferably at least 70%, preferably atleast 80%, preferably at least 85%, preferably at least 90 %, more preferred at least 95%, morepreferred at least 96%, more preferred at least 97%, more preferred at least 98% and mostpreferred at least 99%.

In one aspect, the number of alterations in the variants of the present invention is 1-20, e.g.,1-10 and 1-5, such as 1,2, 3, 4, 5, 6, 7, 8, 9 or 10 alterations.

The variant albumin, a fragment thereof or fusion polypeptide comprising the variantalbumin or fragment thereof has altered plasma half-life compared with the corresponding parentalbumin, fragment thereof, or fusion polypeptide comprising the variant albumin or fragmentthereof.

In a particular preferred embodiment the parent albumin is HSA and the variant albumin, afragment thereof or fusion polypeptide comprising the variant albumin or fragment thereof hasaltered plasma half-life compared with the HSA, the corresponding fragment or fusion polypeptidecomprising HSA or fragment thereof.

The correlation between binding of albumin to its receptor and plasma half-life has beenrealized by the present inventors based on the natural occurring allele of HSA D494N. Theinventors have analyzed this allele and found that it has a lower affinity to its receptor FcRn.

Further, it has been disclosed that a transgenic mouse having the natural mouse FcRnreplaced with human FcRn has a higher serum albumin level than normal mouse; see (J Exp Med.(2003) 197(3):315-22). The inventors have discovered that human FcRn has a higher affinity tomouse serum albumin than mouse FcRn has to mouse serum albumin and, therefore, the observedincrease in serum albumin in the transgenic mice corresponds with a higher affinity between serumalbumin and its receptor, confirming the correlation between albumin binding to FcRn and plasmahalf-life. In addition, variants of albumin that have little or no binding to FcRn have been shown tohave reduced half-life in a mouse model, Kenanova et al (2009) J. Nucl. Med.; 50 (Supplement2):1582).

One way to determine whether the affinity of a variant albumin to FcRn is higher or lowerthan the parent albumin is to use the Surface Plasmon Resonance assay (SPR) as describedbelow. The skilled person will understand that other methods might be useful to determine whetherthe affinity of a variant albumin to FcRn is higher or lower than the affinity of the parent albumin to 20 WO 2011/051489 PCT/EP2010/066572

FcRn, e.g., determination and comparison of the binding constants KD. Thus, according to theinvention variant albumins having a KD that is lower than the KD for natural HSA is considered tohave a higher plasma half-life than HSA and variant albumins having a KD that is higher than theKD for natural HSA is considered to have a lower plasma half-life than HSA.

The variants of albumin or fragments thereof or fusion polypeptides comprising albumin orfragments thereof comprise one or more alterations, such as substitutions, deletions or insertions atone or more (several) positions corresponding to the positions in HSA selected from the groupconsisting of 417, 440, 464, 490, 492, 493, 494, 495, 496, 499, 500, 501, 503, 504, 505, 506, 510,535, 536, 537, 538, 540, 541, 542, 550, 573, 574, 575, 577, 578, 579, 580, 581, 582 and 584. Thesubstitution may be any substitution where the amino acid in the natural albumin sequence issubstituted with a different amino acid selected among the remaining 19 natural occurring aminoacids.

In one aspect, a variant comprises an alteration at one or more (several) positionscorresponding to positions 417, 440, 464, 490, 492, 493, 494, 495, 496, 499, 500, 501, 503, 504, 505, 506, 510, 535, 536, 537, 538, 540, 541, 542, 550, 573, 574, 575, 577, 578, 579, 580, 581,582and 584 in SEQ ID NO: 2. In another aspect, a variant comprises an alteration at two positionscorresponding to any of 417, 440, 464, 490, 492, 493, 494, 495, 496, 499, 500, 501, 503, 504, 505, 506, 510, 535, 536, 537, 538, 540, 541, 542, 550, 573, 574, 575, 577, 578, 579, 580, 581, 582 and584 in SEQ ID NO: 2. In another aspect, a variant comprises an alteration at three positionscorresponding to any of positions 417, 440, 464, 490, 492, 493, 494, 495, 496, 499, 500, 501, 503,504, 505, 506, 510, 535, 536 , 537, 538, 540, 541, 542, 550, 573, 574, 575, 577, 578, 579, 580,581, 582 and 584 in SEQ ID NO: 2. In another aspect, a variant comprises an alteration at eachposition corresponding to positions 417, 440, 464, 490, 492, 493, 494, 495, 496, 499, 500, 501,503 , 504, 505, 506, 510, 535, 536, 537, 538, 540, 541, 542, 550, 573, 574, 575, 577, 578, 579,580, 581,582 and 584 in SEQ ID NO: 2.

In another aspect, the variant comprises the substitution Q417A,H of the mature polypeptideof SEQ ID NO: 2. In another aspect, the variant comprises the substitution H440Q of the maturepolypeptide of SEQ ID NO: 2. In another aspect, the variant comprises the substitution H464Q ofthe mature polypeptide of SEQ ID NO: 2. In another aspect, the variant comprises the substitutionA490D of the mature polypeptide of SEQ ID NO: 2. In another aspect, the variant comprises thesubstitution E492G, Τ,Ρ,Η of the mature polypeptide of SEQ ID NO: 2. In another aspect, thevariant comprises the substitution V493P,L of the mature polypeptide of SEQ ID NO: 2. In anotheraspect, the variant comprises the substitution D494N,Q,A,E,P of the mature polypeptide of SEQ IDNO: 2. In another aspect, the variant comprises the substitution E495Q,A of the mature polypeptideof SEQ ID NO: 2. In another aspect, the variant comprises the substitution T496A of the mature 21 WO 2011/051489 PCT/EP2010/066572 polypeptide of SEQ ID NO: 2. In another aspect, the variant comprises the substitution P499A ofthe mature polypeptide of SEQ ID NO: 2. In another aspect, the variant comprises the substitutionK500E,G,D,A,S,C,P,H,F,N,W,T,M,Y,V,Q,L,l,R of the mature polypeptide of SEQ ID NO: 2. Inanother aspect, the variant comprises the substitution E501 A,P,Q of the mature polypeptide of SEQID NO: 2. In another aspect, the variant comprises the substitution N503K,D,H of the maturepolypeptide of SEQ ID NO: 2. In another aspect, the variant comprises the substitution A504E ofthe mature polypeptide of SEQ ID NO: 2. In another aspect, the variant comprises the substitutionE505K, D of the mature polypeptide of SEQ ID NO: 2. In another aspect, the variant comprises thesubstitution T506F, S of the mature polypeptide of SEQ ID NO: 2. In another aspect, the variantcomprises the substitution H510Q of the mature polypeptide of SEQ ID NO: 2. In another aspect,the variant comprises the substitution H535Q of the mature polypeptide of SEQ ID NO: 2. Inanother aspect, the variant comprises the substitution K536A of the mature polypeptide of SEQ IDNO: 2. In another aspect, the variant comprises the substitution P537A of the mature polypeptide ofSEQ ID NO: 2. In another aspect, the variant comprises the substitution K538A,H of the maturepolypeptide of SEQ ID NO: 2. In another aspect, the variant comprises the substitution T540S ofthe mature polypeptide of SEQ ID NO: 2. In another aspect, the variant comprises the substitutionK541A,D,G,N,E of the mature polypeptide of SEQ ID NO: 2. In another aspect, the variantcomprises the substitution E542P.D of the mature polypeptide of SEQ ID NO: 2. In another aspect,the variant comprises the substitution D550N of the mature polypeptide of SEQ ID NO: 2. Inanother aspect, the variant comprises the substitutionK573Y,W,P,H,F,V,I,T,N,S,G,M,C,A,E,Q,R,L,D of the mature polypeptide of SEQ ID NO: 2. Inanother aspect, the variant comprises the substitution K574N of the mature polypeptide of SEQ IDNO: 2. In another aspect, the variant comprises the substitution Q580K of the mature polypeptide ofSEQ ID NO: 2. In another aspect, the variant comprises the substitution L575F of the maturepolypeptide of SEQ ID NO: 2. In another aspect, the variant comprises the substitution A577T,E ofthe mature polypeptide of SEQ ID NO: 2. In another aspect, the variant comprises the substitutionA578R,S of the mature polypeptide of SEQ ID NO: 2. In another aspect, the variant comprises thesubstitution S579C,T of the mature polypeptide of SEQ ID NO: 2. In another aspect, the variantcomprises the substitution Q580K of the mature polypeptide of SEQ ID NO: 2. In another aspect,the variant comprises the substitution A581D of the mature polypeptide of SEQ ID NO: 2. Inanother aspect, the variant comprises the substitution A582T of the mature polypeptide of SEQ IDNO: 2. In another aspect, the variant comprises the substitution G584A of the mature polypeptide ofSEQ ID NO: 2.

In one aspect, the variant comprises an alteration at a position corresponding to position417. In another aspect, the amino acid at a position corresponding to position 417 is substituted22 WO 2011/051489 PCT/EP2010/066572 with Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, lie, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, or

Val, preferably with Ala or His. In another aspect, the variant comprises the substitution Q417A, H of the mature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position440. In another aspect, the amino acid at a position corresponding to position 440 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, He, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Ala. In another aspect, the variant comprises the substitution H440Q of themature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position464. In another aspect, the amino acid at a position corresponding to position 464 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, He, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Ala. In another aspect, the variant comprises the substitution H464Q of themature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position490 in another aspect, the amino acid at a position corresponding to position 490 is substituted withAla, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, He, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, or Val.In another aspect, the variant comprises the substitution A490G of the mature polypeptide of SEQID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 492. In another aspect, the amino acid at a position corresponding to position 492 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, He, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Gly. In another aspect, the variant comprises the substitution E492G of themature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 493. In another aspect, the amino acid at a position corresponding to position 493 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, He, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Pro. In another aspect, the variant comprises the substitution V493P of themature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 494. In another aspect, the amino acid at a position corresponding to position 494 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, He, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Asn, Gin or Ala. In another aspect, the variant comprises the substitutionD494N,Q, A of the mature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 495. In another aspect, the amino acid at a position corresponding to position 495 is substituted 23 WO 2011/051489 PCT/EP2010/066572 with Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, Ile, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, or Val, preferably with Gin or Ala. In another aspect, the variant comprises the substitution E495Q or A of the mature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 496. In another aspect, the amino acid at a position corresponding to position 496 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, Ile, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Ala. In another aspect, the variant comprises the substitution T496A of themature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 499. In another aspect, the amino acid at a position corresponding to position 499 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, Ile, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Ala. In another aspect, the variant comprises the substitution P499A of themature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 500. In another aspect, the amino acid at a position corresponding to position 500 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, Ile, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Ala. In another aspect, the variant comprises the substitutionK500E,G,D,A,S,C,P,H,F,N,W,T,M,Y,V,Q,L,l,R of the mature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 501. In another aspect, the amino acid at a position corresponding to position 501 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, lie, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Ala or Gin to reduce affinity and Pro to increase affinity. In another aspect, thevariant comprises the substitution E501A, Q, P of the mature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 503. In another aspect, the amino acid at a position corresponding to position 503 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, Ile, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Asp or Lys or His. In another aspect, the variant comprises the substitutionN503D, K, H of the mature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 504. In another aspect, the amino acid at a position corresponding to position 504 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, Ile, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal. In another aspect, the variant comprises the substitution A504 of the mature polypeptide ofSEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 505. In another aspect, the amino acid at a position corresponding to position 505 is substituted 24 WO 2011/051489 PCT/EP2010/066572 with Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, lie, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, or Val. In another aspect, the variant comprises the substitution E505D of the mature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 506. In another aspect, the amino acid at a position corresponding to position 506 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, lie, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal. In another aspect, the variant comprises the substitution T506S,F of the mature polypeptide ofSEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position510. In another aspect, the amino acid at a position corresponding to position 510 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, lie, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Gin. In another aspect, the variant comprises the substitution H510Q of themature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 535. In another aspect, the amino acid at a position corresponding to position 535 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, lie, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Gin. In another aspect, the variant comprises the substitution H535Q of themature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 536. In another aspect, the amino acid at a position corresponding to position 536 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, lie, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Ala. In another aspect, the variant comprises the substitution K536A of themature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 537. In another aspect, the amino acid at a position corresponding to position 537 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, He, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Ala. In another aspect, the variant comprises the substitution P537A of themature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 538. In another aspect, the amino acid at a position corresponding to position 538 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, lie, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Ala. In another aspect, the variant comprises the substitution K538H, A of themature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 540. In another aspect, the amino acid at a position corresponding to position 540 is substituted25 WO 2011/051489 PCT/EP2010/066572 with Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, lie, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal. In another aspect, the variant comprises the substitution T540S of the mature polypeptide ofSEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 541. In another aspect, the amino acid at a position corresponding to position 541 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, lie, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Gly, Asp or Ala. In another aspect, the variant comprises the substitutionK541G, D A, N of the mature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 542. In another aspect, the amino acid at a position corresponding to position 542 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, lie, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Asp or Pro. In another aspect, the variant comprises the substitution E542D, Pof the mature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position550. In another aspect, the amino acid at a position corresponding to position 550 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, lie, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Asn to reduce affinity, preferably with Glu to increase affinity.

In another aspect, the variant comprises an alteration at a position corresponding to position 573. In another aspect, the amino acid at a position corresponding to position 573 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, lie, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Tyr, Trp, Pro, His. Phe, Val, lie, Thr, Asn, Ser, Gly, Met, Cys, Ala, Glu, Gin, Arg,Leu, Asp. In another aspect, the variant comprises the substitutionK573Y,W,P,H,F,V,I,T,N,S,G,M,C,A,E,Q,R,L,D of the mature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 574. In another aspect, the amino acid at a position corresponding to position 574 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, He, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Asn. In another aspect, the variant comprises the substitution K574N of themature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 575. In another aspect, the amino acid at a position corresponding to position 575 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, He, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Phe. In another aspect, the variant comprises the substitution L575F of themature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 577. In another aspect, the amino acid at a position corresponding to position 577 is substituted26 WO 2011/051489 PCT/EP2010/066572 with Ala, Arg, Asn, Asp, Cys, Gln, Glu, Gly, His, lie, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Thr or Glu. In another aspect, the variant comprises the substitution A577TE ofthe mature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 578. In another aspect, the amino acid at a position corresponding to position 578 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gln, Glu, Gly, His, lie, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Arg or Ser. In another aspect, the variant comprises the substitution A578R,Sof the mature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 579. In another aspect, the amino acid at a position corresponding to position 579 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gln, Glu, Gly, His, lie, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Cys or Thr. In another aspect, the variant comprises the substitution S579C,Tof the mature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 580. In another aspect, the amino acid at a position corresponding to position 580 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gln, Glu, Gly, His, lie, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Lys. In another aspect, the variant comprises the substitution Q580K of themature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 581. In another aspect, the amino acid at a position corresponding to position 581 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gln, Glu, Gly, His, lie, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Asp. In another aspect, the variant comprises the substitution A581D of themature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 582. In another aspect, the amino acid at a position corresponding to position 582 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gln, Glu, Gly, His, lie, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Thr. In another aspect, the variant comprises the substitution A582T of themature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position584. In another aspect, the amino acid at a position corresponding to position 584 is substitutedwith Ala, Arg, Asn, Asp, Cys, Gln, Glu, Gly, His, He, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, orVal, preferably with Ala. In another aspect, the variant comprises the substitution G584A of themature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at positions corresponding topositions 494 and 496 in SEQ ID NO: 2, such as those described above. 27 WO 2011/051489 PCT/EP2010/066572

In another aspect, the variant comprises alterations at positions corresponding to positions492 and 493 in SEQ ID NO: 2, such as those described above.

In another aspect, the variant comprises alterations at positions corresponding to positions494 and 417 in SEQ ID NO: 2, such as those described above.

In another aspect, the variant comprises alterations at positions corresponding to positions492 and 503 in SEQ ID NO: 2, such as those described above.

In another aspect, the variant comprises alterations at positions corresponding to positions492 and 573 in SEQ ID NO: 2, such as those described above.

In another aspect, the variant comprises alterations at positions corresponding to positions492, 503, and 573 in SEQ ID NO: 2, such as those described above.

In one embodiment the variant albumin or fragments thereof, or fusion polypeptidescomprising the variant albumin or fragments thereof according to the invention contains onesubstitution at a position corresponding to a position in HSA selected from the group consisting of417, 440, 464, 490, 492, 493, 494, 495, 496, 499, 500, 501, 503, 504, 505, 506, 510, 535, 536,537, 538, 540, 541, 542, 550, 573, 574, 575, 577, 578, 579, 580, 581,582 and 584 in SEQ ID NO:2 provided that the variant albumin is not the variant consisting of SEQ ID NO: 2 with thesubstitution D494N, E501K, K541E, D550G,A, K573E or K574N. The variant albumin, fragmentthereof or fusion polypeptides comprising variant albumin or a fragment thereof according to theinvention may comprise additional substitutions, insertions or deletions at one or more (several)positions corresponding to other positions in HSA.

In another embodiment the variant albumin or fragments thereof, or fusion polypeptidescomprising variant albumin or fragments thereof according to the invention contains two, three,four, five , six, seven, eight, nine, ten, eleven, twelve , thirteen, fourteen fifteen, sixteen, seventeen,eighteen, nineteen twenty or even more substitutions at positions corresponding to positions inHSA selected from the group consisting of 417, 440, 464, 490, 492, 493, 494, 495, 496, 499, 500,501, 503, 504, 505, 506, 510, 535, 536, 537, 538, 540, 541, 542, 550, 573, 574, 575, 577, 578,579, 580, 581, 582 and 584 of SEQ ID NO: 2. The variant albumin or fragments thereof, or fusionpolypeptides comprising variant albumin or fragments thereof according to the invention maycomprise additional substitutions, insertions or deletions at positions corresponding to otherpositions in HSA.

In a further embodiment the variants of albumin or fragments thereof, or fusion polypeptidescomprising variant albumin or a fragment thereof according to the invention have a plasma half-lifethat is longer than the plasma half-life of the parent albumin fragment thereof or fusion polypeptidecomprising the parent albumin or a fragment thereof. Examples according to this embodimentinclude variants of albumin or fragments thereof, or fusion polypeptides comprising variant albumin28 WO 2011/051489 PCT/EP2010/066572 or a fragment thereof comprising a substitution in the position corresponding to 492, 503, 542, 550,573, 574, 580, 581, 582 or 584 in SEQ ID NO: 2. Preferred substitutions according to thisembodiment of the invention include the substitution of the amino acid residue in the positioncorresponding to 492 in SEQ ID NO: 2 with a G residue, substitution of the amino acid residue inthe position corresponding to 503 in SEQ ID NO: 2 with a H or a K residue, substitution of theamino acid residue in the position corresponding to 550 in SEQ ID NO: 2 with an E residue, thesubstitution of the amino acid residue in a position corresponding to 573 in SEQ ID NO: 2 with anΥ,νν,Ρ,Η,Ρ,ν,Ι,Τ,Ν,ε,Ο,Μ,Ο,Α,Ε,Ο,Ρ,Ι. or a D, the substitution of the amino acid residue in aposition corresponding to 574 in SEQ ID NO: 2 with an N residue, or the substitution of the aminoacid residue in the position corresponding to 580 in SEQ ID NO: 2 with an K residue. Otherpreferred variants have a substitution in the position corresponding to 492 in SEQ ID NO: 2 with aG residue and a substitution in the position corresponding to 573 in SEQ ID NO: 2 with an A or a Presidue. Other preferred variant has a number of substitutions corresponding to position 492 inSEQ ID NO: 2 with an H residue in position 503 in SEQ ID NO: 2.

Other preferred variants have a substitution in the position corresponding to 492 in SEQ IDNO: 2 with a G residue and a substitution in the position corresponding to position 503 in SEQ IDNO: 2 corresponding to a H or a K and a substitution in position 573 in SEQ ID NO: 2 with an A or aP residue.

In a further embodiment the variants of albumin or fragments thereof, or fusion polypeptidescomprising variant albumin or fragments thereof according to the invention have a plasma half-lifethat is shorter than the plasma half-life of the parent albumin fragment thereof or fusion polypeptidecomprising the parent albumin or a fragment thereof. Examples according to this embodimentinclude variants of albumin or fragments thereof, or fusion polypeptides comprising variant albuminor a fragment thereof comprising a substitution in the position corresponding to 417, 440, 494, 495,496, 499, 500, 501, 536, 537, 538, 541, 494+496 or 492+493 in SEQ ID NO: 2. Preferredsubstitutions include the substitutions corresponding to Q417A, H440Q, D494E+Q417H,D494N,Q,A, E495Q,A, T496A, D494N+T496A or, P499A, K500A, E501A , E501Q, K536A, P537A ,K538A, K541G, K541A K541D or D550N in SEQ ID NO: 2.

In another embodiment of the invention the variants of albumin or fragments thereof, orfusion polypeptides comprising variant albumin or a fragment thereof according to the inventionhave lost their ability to bind FcRn. In this connection variants of albumin or fragments thereof, orfusion polypeptides comprising variant albumin or fragments thereof is considered to have lost theability to bind FcRn if the measured resonance units for the variant in the SPR assay describedbelow is less than 10% of the measured resonance units for the corresponding parent albumin orfragment thereof. Examples according to this embodiment include variants of albumin or fragments 29 WO 2011/051489 PCT/EP2010/066572 thereof, or fusion polypeptides comprising variant albumin or fragments thereof comprising asubstitution at a position corresponding to 464, 500, 510 or 535 in SEQ ID NO: 2. Preferredsubstitutions include the substitutions corresponding to H464Q, K500A,P,C,S,A,D.G H510Q orH535Q in SEQ ID NO: 2.

In addition to the one or more substitutions at one or more positions corresponding topositions 417, 464, 490, 492, 493, 494, 495, 496, 499, 500, 501, 503, 504, 505, 506, 510, 535, 536,537, 538, 540, 541,542, 550, 573, 574, 580 581, 582 and 584 in SEQ ID NO: 2 the variant albuminor fragments thereof, or fusion polypeptides comprising variant albumin or fragments thereofaccording to the invention may contain additional substitutions, deletions or insertions in otherpositions of the molecules. Such additional substitutions, deletions or insertions maybe useful inorder to alter other properties of the molecules such as but not limited to altered glycosylation;introduction of reactive groups of the surface such a thiol groups, removing/generating acarbamoylation site; etc.

Residues that might be altered in order to provide reactive residues on the surface andwhich advantageously could be applied to the present invention has been disclosed in theunpublished patent application WO 2010/092135 (Included by reference). Particular preferredresidues include the positions corresponding to positions in SEQ ID NO: 2 .

As examples of alterations that can be made in SEQ ID NO: 2 or in corresponding positionsin other albumins in order to provide a reactive thiol group on the surface includes alterationscorresponding to following alterations in SEQ ID NO: 2: L585C, D1C, A2C, D562C, A364C, A504C,E505C, T79C, E86C, D129C, D549C, A581C, D121C, E82C, S270C, A578C, L595LC, D1DC,A2AC, D562DC, A364AC, A504AC, E505EC, T79TC, E86EC, D129DC, D549DC, A581AC,A581AC, D121DC, E82EC, S270SC, A579AC, C360*, C316*, C75*, C168*, C558*, C361*, C91*,C124*, C169* and C567*. Alternatively a cysteine residue may be added to the N or C terminal ofalbumin.

Polynucleotides

The present invention also relates to isolated polynucleotides that encode any of thevariants of the present invention.

Nucleic Acid Constructs

The present invention also relates to nucleic acid constructs comprising a polynucleotideencoding a variant of the present invention operably linked to one or more (several) controlsequences that direct the expression of the coding sequence in a suitable host cell underconditions compatible with the control sequences. 30 WO 2011/051489 PCT/EP2010/066572 A polynucleotide may be manipulated in a variety of ways to provide for expression of avariant. Manipulation of the polynucleotide prior to its insertion into a vector may be desirable ornecessary depending on the expression vector. The techniques for modifying polynucleotidesutilizing recombinant DNA methods are well known in the art.

The control sequence may be a promoter sequence, which is recognized by a host cell forexpression of the polynucleotide. The promoter sequence contains transcriptional controlsequences that mediate the expression of the variant. The promoter may be any nucleic acidsequence that shows transcriptional activity in the host cell including mutant, truncated, and hybridpromoters, and may be obtained from genes encoding extracellular or intracellular polypeptideseither homologous or heterologous to the host cell.

In a yeast host, useful promoters are obtained from the genes for Saccharomycescerevisiae enolase (ENO-1), Saccharomyces cerevisiae protease A (PRA1), Saccharomycescerevisiae protease B (PRB1), Saccharomyces cerevisiae translation elongation factor (TEF1),Saccharomyces cerevisiae translation elongation factor (TEF2), Saccharomyces cerevisiaegalactokinase (GAL1), Saccharomyces cerevisiae alcohol dehydrogenase/glyceraldehyde-3-phosphate dehydrogenase (ADH1, ADH2/GAP), Saccharomyces cerevisiae triose phosphateisomerase (TPI), Saccharomyces cerevisiae metallothionein (CUP1), and Saccharomycescerevisiae 3-phosphoglycerate kinase. Other useful promoters for yeast host cells are described byRomanos et al., 1992, Yeast 8: 423-488.

The control sequence may also be a suitable transcription terminator sequence, which isrecognized by a host cell to terminate transcription. The terminator sequence is operably linked tothe 3’-terminus of the polynucleotide encoding the variant. Any terminator that is functional in thehost cell may be used.

Preferred terminators for yeast host cells are obtained from the genes for Saccharomycescerevisiae enolase, Saccharomyces cerevisiae cytochrome C (CYC1), Saccharomyces cerevisiaealcohol dehydrogenase (ADH1) and Saccharomyces cerevisiae glyceraldehyde-3-phosphatedehydrogenase. Other useful terminators for yeast host cells are described by Romanos et al.,1992, supra.

The control sequence may also be a suitable leader sequence, a nontranslated region of anmRNA that is important for translation by the host cell. The leader sequence is operably linked tothe 5’-terminus of the polynucleotide encoding the variant. Any leader sequence that is functional inthe host cell may be used.

Suitable leaders for yeast host cells are obtained from the genes for Saccharomycescerevisiae enolase (ENO-1), Saccharomyces cerevisiae 3-phosphoglycerate kinase, 31 WO 2011/051489 PCT/EP2010/066572

Saccharomyces cerevisiae alpha-factor, and Saccharomyces cerevisiae alcoholdehydrogenase/glyceraldehyde-3-phosphate dehydrogenase (ADH2/GAP).

The control sequence may also be a polyadenylation sequence, a sequence operably linkedto the 3’-terminus of the variant-encoding sequence and, when transcribed, is recognized by thehost cell as a signal to add polyadenosine residues to transcribed mRNA. Any polyadenylationsequence that is functional in the host cell may be used.

Useful polyadenylation sequences for yeast host cells are described by Guo and Sherman,1995, Mol. Cellular Biol. 15: 5983-5990.

The control sequence may also be a signal peptide coding region that encodes a signalpeptide linked to the N-terminus of a variant and directs the variant into the cell’s secretorypathway. The 5’-end of the coding sequence of the polynucleotide may inherently contain a signalpeptide coding region naturally linked in translation reading frame with the segment of the codingregion that encodes the variant. Alternatively, the 5’-end of the coding sequence may contain asignal peptide coding region that is foreign to the coding sequence. The foreign signal peptidecoding region may be required where the coding sequence does not naturally contain a signalpeptide coding region. Alternatively, the foreign signal peptide coding region may simply replacethe natural signal peptide coding region in order to enhance secretion of the variant. However, anysignal peptide coding region that directs the expressed variant into the secretory pathway of a hostcell may be used.

Useful signal peptides for yeast host cells are obtained from the genes for Saccharomycescerevisiae alpha-factor and Saccharomyces cerevisiae invertase. Other useful signal peptidecoding sequences are described by Romanos eta!., 1992, supra.

Where both signal peptide and propeptide regions are present at the N-terminus of avariant, the propeptide region is positioned next to the N-terminus of the variant and the signalpeptide region is positioned next to the N-terminus of the propeptide region.

Methods of Production

The variants of the present invention can be prepared using techniques well known to theskilled person. One convenient way is by cloning nucleic acid encoding the parent albumin or afragment thereof or fusion polypeptide comprising albumin or a fragment thereof, modifying saidnucleic acid to introduce the desired substitution(s) at one or more (several) positionscorresponding to positions 417, 464, 490, 492, 493, 494, 495, 496, 499, 500, 501, 503, 504, 505,506, 510, 535, 536, 537, 538, 540, 541, 542, 550, 573, 574 and 580 in SEQ ID NO: 2, where thevariant is not the variant consisting of SEQ ID NO:2 with the substitution D494N, E501K, K541E,D550G,A, K573E or K574N., preparing a suitable genetic construct where the modified nucleic acid 32 WO 2011/051489 PCT/EP2010/066572 is placed in operative connection with suitable regulatory genetic elements, such as promoter,terminator, activation sites, ribosome binding sites etc., introducing the genetic construct into asuitable host organism, culturing the transformed host organism under conditions leading toexpression of the variant and recovering the variant. All these techniques are known in the art and itis within the skills of the average practitioner to design a suitable method for preparing a particularvariant according to the invention.

The variant polypeptide of the invention may also be connected to a signal sequence inorder to have the variant polypeptide secreted into the growth medium during culturing of thetransformed host organism. It is generally advantageous to have the variant polypeptide secretedinto the growth medium in order to ease recovery and purification.

Techniques for preparing variant polypeptides have also been disclosed in WO 2009019314(included by reference) and these techniques may also be applied to the present invention.

Albumins have been successfully expressed as recombinant proteins in a range of hostsincluding fungi (including but not limited to Aspergillus (WO06066595), Kluyveromyces (Fleer 1991,Bio/technology 9, 968-975), Pichia (Kobayashi 1998 Therapeutic Apheresis 2, 257-262) andSaccharomyces (Sleep 1990, Bio/technology 8, 42-46)), bacteria (Pandjaitab 2000, J. Allergy Clin.Immunol. 105, 279-285)), animals (Barash 1993, Transgenic Research 2, 266-276) and plants(including but not limited to potato and tobacco (Sijmons 1990, Bio/technology 8, 217 and Farran2002, Transgenic Research 11, 337-346). The variant polypeptide of the invention is preferablyproduced recombinantly in a suitable host cell. In principle any host cell capable of producing apolypeptide in suitable amounts may be used and it is within the skills of the average practitioner toselect a suitable host cell according to the invention. A preferred host organism is yeast, preferablyselected among Saccharomycacae, more preferred Saccharomyces cerevisiae.

The variant polypeptides of the invention may be recovered and purified from the growthmedium using a combination of known separation techniques such as filtration, centrifugation,chromatography, and affinity separation techniques etc. It is within the skills of the averagepractitioner to purify the variants of the invention using a particular combination of such knownseparation steps. As an example of purification techniques that may be applied to the variants ofthe present invention can be mentioned the teaching of W00044772.

The variant polypeptides of the invention may be used for delivering a therapeuticallybeneficial compound to an animal or a human individual in need thereof. Such therapeuticallybeneficial compounds include, but are not limited, to labels and readily detectable compounds foruse in diagnostics, such as various imaging techniques; pharmaceutical active compounds such asdrugs, or specifically binding moieties such as antibodies. The variants of the invention may evenbe connected to two or more different therapeutically beneficial compounds, e.g., an antibody and a 33 WO 2011/051489 PCT/EP2010/066572 drug, which gives the combined molecule the ability to bind specifically to a desired target andthereby provide a high concentration of the connected drug at that particular target.

Fusion polypeptides

The variants of albumin or fragments thereof according to the invention may also be fusedwith a non-albumin polypeptide fusion partner. The fusion partner may in principle be anypolypeptide but generally it is preferred that the fusion partner is a polypeptide having therapeuticor diagnostic properties. Fusion polypeptides comprising albumin or fragments thereof are known inthe art. It has been found that such fusion polypeptide comprising albumin or a fragment thereofand a fusion partner polypeptide have a longer plasma half-life compared to the unfused fusionpartner polypeptide. According to the invention it is possible to alter the plasma half-life of thefusion polypeptides according to the invention compared to the corresponding fusion polypeptidesof the prior art.

One or more therapeutic polypeptides may be fused to the N-terminus, the C-terminus ofalbumin, inserted into a loop in the albumin structure or any combination thereof. It may or it maynot comprise linker sequences separating the various components of the fusion polypeptide.

Teachings relating to fusions of albumin or a fragment thereof are known in the art and theskilled person will appreciate that such teachings can also be applied to the present invention. WO2001/79271 A and WO 2003/59934 A also contain examples of therapeutic polypeptides that maybe fused to albumin or fragments thereof, and these examples apply also to the present invention.

Conjugates

The variants of albumin or fragments thereof according to the invention may be conjugatedto a second molecule using techniques known within the art. Said second molecule may comprise adiagnostic moiety, and in this embodiment the conjugate may be useful as a diagnostic tool such asin imaging; or the second molecule may be a therapeutic compound and in this embodiment theconjugate may be used for therapeutic purposes where the conjugate will have the therapeuticproperties of the therapeutic compound as well as the long plasma half-life of the albumin.Conjugates of albumin and a therapeutic molecule are known in the art and it has been verified thatsuch conjugates have long plasma half-life compared with the non-conjugated, free therapeuticmolecule as such. The conjugates may conveniently be linked via a free thio group present on thesurface of HSA (amino acid residue 34 of mature HSA) using well known chemistry.

In one particular preferred aspect the variant albumin or fragment thereof is conjugated to abeneficial therapeutic compound and the conjugate is used for treatment of a condition in a patientin need thereof, which condition is responsive to the particular selected therapeutic compound.Techniques for conjugating such a therapeutically compound to the variant albumin or fragment 34 218858/2thereof are known in the art. WO 2009/019314 discloses examples of techniquessuitable for conjugating a therapeutically compound to a polypeptide which techniquescan also be applied to the present invention. Further WO 2009/019314 disclosesexamples of compounds and moieties that may be conjugated to substituted transferrin5 and these examples may also be applied to the present invention. The teaching of WO2009/019314 is included herein by reference. HSA contains in its natural form one free thiol group that conveniently may beused for conjugation. As a particular embodiment within this aspect the variant albuminor fragment thereof may comprise further modifications provided to generate additionalio free thiol groups on the surface. This has the benefit that the payload of the variantalbumin or fragment thereof is increased so that more than one molecule of thetherapeutic compound can be conjugated to each molecule of variant albumin orfragment thereof, or two or more different therapeutic compounds may be conjugatedto each molecule of variant albumin or fragment thereof, e.g., a compound having15 targeting properties such as an antibody specific for example a tumour; and a cytotoxicdrug conjugated to the variant albumin or fragment thereof thereby creating a highlyspecific drug against a tumour. Teaching of particular residues that may be modified toprovide for further free thiol groups on the surface can be found in copending patentapplication WO 2010/092135. 20

Associates

The variants of albumin or fragments thereof may further be used in form of“associates”. In this connection the term “associate” is intended to mean a compoundcomprising a variant of albumin or a fragment thereof and another compound bound or25 associated to the variant albumin or fragment thereof by non-covalent binding. As anexample of such an associate can be mentioned an associate consisting variantalbumin and a lipid associated to albumin by a hydrophobic interaction. Suchassociates are known in the art and they may be prepared using well knowntechniques. As an example of a preferred associate according to the invention can be30 mentioned an associate comprising variant albumin and paclitaxel.

Other uses

The variant albumin or fragments thereof or fusion polypeptides comprisingvariant albumin or fragments thereof according to the invention have the benefit that35 their plasma half-life is altered compared to the parent albumin or fragments thereof orfusion polypeptides comprising parent albumin or fragments thereof. This has theadvantage that the plasma half-life of conjugates comprising variant albumin or afragment thereof or fusion polypeptide comprising variant albumin or a fragmentthereof, or an associate comprising variant albumin or a fragment thereof according to40 the invention can be selected in accordance with the particular therapeutic purpose. 35 WO 2011/051489 PCT/EP2010/066572

For example for a conjugate, associate or fusion polypeptide used for imaging purposes inanimals or human beings, where the imaging moiety has an very short half-life and a conjugate or afusion polypeptide comprising HSA has a plasma half-life that is far longer than needed for theimaging purposes it would be advantageous to use a variant albumin or fragment thereof of theinvention having a shorter plasma half-life than the parent albumin or fragment thereof, to provideconjugates of fusion polypeptides having a plasma half-life that is sufficiently long for the imagingpurpose but sufficiently short to be cleared form the body of the particular patient on which it isapplied.

In another example for a conjugate, an associate or fusion polypeptide comprising atherapeutic compound effective to treat or alleviate a particular condition in a patient in need forsuch a treatment it would be advantageous to use the variant albumin or fragment thereof having alonger plasma half-life than the parent albumin or fragment thereof, to provide associates orconjugates or fusion polypeptides having longer plasma half-lives which would have the benefit thatthe administration of the associate or conjugate or fusion polypeptide of the invention would beneeded less frequently or reduced dose with less side affects compared to the situation where theparent albumin or associates thereof or fragment thereof was used.

In a further aspect the invention relates to compositions comprising the variant albumin,associates thereof or fragment thereof, variant albumin fragment or associates thereof or fusionpolypeptide comprising variant albumin or fragment thereof according to the invention. Thecompositions are preferably pharmaceutical compositions. The composition may be prepared usingtechniques known in the area such as disclosed in recognized handbooks within thepharmaceutical field.

In a particular embodiment the compositions comprise a variant albumin or a fragmentthereof according to the invention and a compound comprising a pharmaceutically beneficial moietyand an albumin binding domain (ABD). According to the invention ABD means a site, moiety ordomain capable of binding to circulating albumin in vivo and thereby conferring transport in thecirculation of the ABD and any compound or moiety bound to said ABD. ABD’s are known in theart and have been shown to bind very tight to albumin so a compound comprising an ABD bound toalbumin will to a certain extent behave as a single molecule. The inventors have realized by usingthe variant albumin or fragment thereof according to the invention together with a compoundcomprising a pharmaceutically beneficial moiety and an ABD makes it possible to alter the plasmahalf-life of the compound comprising a pharmaceutically beneficial moiety and an ABD compared tothe situation where said compound were injected as such in a patient having need thereof oradministered in a formulation comprising natural albumin or a fragment thereof. 36 218858/2

The variant albumin or fragments thereof, conjugates comprising variantalbumin or a fragment thereof or fusion polypeptide comprising variant albumin or afragment thereof, or an associate comprising variant albumin or a fragment thereofaccording to the invention may also be incorporated into nano- or microparticles usingtechniques well known within the art. A preferred method for preparing nano- ormicroparticles that may be applied to the variant albumins or fragments thereofaccording to the invention is disclosed in WO 2004/071536.

Compositions

The present invention is also directed to the use of a variant of albumin or afragment thereof or fusion polypeptides comprising variant albumin or fragmentsthereof, or a conjugate comprising a variant of albumin or a fragment thereof, or anassociate comprising a variant of albumin or a fragment thereof for the manufacture ofa pharmaceutical composition, where in the variant of albumin or a fragment thereof orfusion polypeptides comprising variant albumin or fragments thereof, or a conjugatecomprising a variant of albumin or a fragment thereof, or an associate comprising avariant of albumin or a fragment thereof has an altered plasma half-life compared withHSA or the corresponding fragment thereof or fusion polypeptide comprising HSA orfragment thereof or conjugate comprising HSA.

In this connection the corresponding fragment of HSA is intended to mean afragment of HSA that aligns with and has same number of amino acids as the fragmentof the variant albumin with which it is compared. Similarly the corresponding fusionpolypeptide comprising HSA or conjugate comprising HSA is intended to meanmolecules having same size and amino acid sequence as the fusion polypeptide ofconjugate comprising variant albumin, with which it is compared.

Preferably the variant of albumin or a fragment thereof or fusion polypeptidescomprising variant albumin or fragments thereof, fragment thereof, or a conjugatecomprising a variant of albumin or a fragment thereof has a plasma half-life that ishigher than the plasma half-life of HSA or the corresponding fragment thereof or fusionpolypeptide comprising HSA or fragment thereof.

Alternatively, this may be expressed as the variant of albumin or a fragmentthereof or fusion polypeptides comprising variant albumin or fragments thereof,fragment thereof, or a conjugate comprising a variant of albumin or a fragment thereofhas a KD to FcRn that is lower that the corresponding KD for HSA or the correspondingfragment thereof or fusion polypeptide comprising HSA or fragment thereof. Preferably,is KD for the variant of albumin or a fragment thereof or fusion polypeptides comprisingvariant albumin or fragments thereof, fragment thereof, or a conjugate comprising avariant of albumin or a fragment thereof less than 0.9X KD for HSA, more 37 WO 2011/051489 PCT/EP2010/066572 preferred less than 0.5X KD for HSA, more preferred less than 0.1 X KD for HSA, even morepreferred less than 0.05X KD for HSA, even more preferred less than 0.02X KD for HSA and mostpreferred less than 0.01 X KD for HSA.

The variant of albumin or a fragment thereof or fusion polypeptides comprising variantalbumin or fragments thereof, fragment thereof, or a conjugate comprising a variant of albumin or afragment thereof is preferably the variant of albumin or a fragment thereof or fusion polypeptidescomprising variant albumin or fragments thereof, fragment thereof, or a conjugate comprising avariant of albumin or a fragment thereof according to the invention.

The present invention is further described by the following examples that should not beconstrued as limiting the scope of the invention.

Examples

Materials and Methods ELISA:

Wells were coated with wild-type HSA or variants diluted in phosphate buffered saline (PBS)to stated concentrations, incubated overnight at 4 C and then blocked with 4% skimmed milk(Acumedia) for 1 hour at room temperature. The wells were then washed four times withPBS/0.005% TWEEN® 20 (PBS/T) pH 6.0 before glutathione-S-transferase (GST) -fused QshFcRn(0.5 pg/ml) as described in FEBS J. 2008 Aug;275(16):4097-110. pre-incubated with anhorseradish peroxidase (HRP)-conjugated polyclonal anti-GST from goat (1:5000; GE Healthcare),diluted in 4% skimmed milk PBS/0.005% TWEEN® 20 (PBS/T) pH 6.0 was added to each well andincubated for 1.5 h at room temperature followed by washing four times with PBS/T pH 6.0. Onehundred μΙ of the substrate tetramethylbenzidine (TMB) (Calbiochem) was added to each well andincubated for 45 min before 100 μΙ of 0.25 M HCI was added. The absorbance was measured at450 nm using a Sunrise TECAN spectrophotometer (TECAN, Maennedorf, Switzerland).

The same ELISA was repeated with PBS/T pH 7.4.

Surface Plasmon Resonance (SPR): SPR experiments were carried out using a Biacore 3000 instrument (GE Healthcare). Flowcells of CM5 sensor chips were coupled with shFcRn-GST (-1400-5000RU) using amine couplingchemistry as described in the protocol provided by the manufacturer. The coupling was performedby injecting 10pg/ml of the protein in 10mM sodium acetate pH 5.0 (GE healthcare). Phosphatebuffer (67mM phosphate buffer, 0.15M NaCI, 0.005% TWEEN® 20) at pH 6.0) was used as runningbuffer and dilution buffer. Regeneration of the surfaces were achieved using injections of HBS-EPbuffer (0.01 M HEPES, 0.15M NaCI, 3mM EDTA, 0.005% surfactant P20) at pH 7.4 (Biacore AB).For binding to immobilized shFcRn-GST, 1.0-0.5 μΜ of each HSA variant was injected over the 38 WO 2011/051489 PCT/EP2010/066572 surface at constant flow rate (40 μΙ/ml) at 25 C. In all experiments, data was zero adjusted and thereference cell subtracted. Data evaluation was performed using BIAevaluation 4.1 software(BIAcore AB).

The same SPR assay was repeated with HBS-EP buffer pH 7.4.

For the purposes of this patent unless otherwise stated HSA, WT HSA, rHA refer toRecombinant human serum albumin commercially available under the registered tradenameRECOMBUMIN (available from Novozymes Biopharma UK Ltd, Nottingham UK) was used for theexamples.

Serum albumin from other species: The albumins were recombinant wheres stated,produced using sequences provided from publicly available databases. Or purchased fromcommercial suppliers.

FcRn Expression and purification of soluble Human (shFcRn) and Mouse (smFcRn) FcRn :Methods for the generation of shFcRn and smFcRn expression plasmids, expression andpurification of each heterodimer can be found in Bemtzen et al. (2005) J. Immunol. Methods298:93-104).AIternatively shFcRn FcRn heterodimer was produced by GeneArt AG (Germany).Sequences for the two sub units of the heterodimer can be found in SEQ ID NO: 3 and SEQ ID NO:4. The soluble receptor was expressed in HEK293 cells and purified from culture supernatant usingNi-HiTrap chromatography columns.

Example 1. Preparation of variants

Preparation of specific HSA mutein expression plasmids

Methods for the expression of HSA mutant variants and HSA fusion variants were producedusing several techniques. Standard molecular biology techniques were employed throughout suchas described in Sambrook, J. and D. W. Russell, 2001. Molecular Cloning: a laboratory manual, 3rded. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.

Method 1. Amino acid substitutions in HSA detailed in Tablel

Synthetic DNA Ncol/Sacl fragments (859 bp) were generated by gene assembly (GeneArtAG, Germany) containing point mutations within the HSA-encoding gene (SEQ ID NO: 1) tointroduce the desired amino acid substitution in the translated protein. Table 2 details the codonsused to introduce the amino acid substitutions into the HSA-encoding gene. The nucleotidesequence of the synthetic fragment encoding unchanged amino acids (i.e. wild type) was identicalto that in pDB2243 (described in WO 00/44772). The synthetic nucleotide fragments were ligatedinto /Vcol/Sacl-digested pDB2243 to produce plasmids pDB3876 - pDB3886 (Table 1). For theproduction of expression plasmids, pDB3876 - pDB3886 (see Table 1) were each digested with 39 WO 2011/051489 PCT/EP2010/066572

Notl and PvlA, the DNA fragments were separated through a 0.7% (w/v) TAE gel, and 2992bpfragments (‘Not\ cassettes’ including PRB1 promoter, DNA encoding the fusion leader (FL)sequence (disclosed in WO 2010/092135), nucleotide sequence encoding HSA and ADH1terminator; see Figure 1) were purified from the agarose gel using a Qiagen Gel Extraction Kit5 following the manufacturer’s instructions. ‘Not\ cassettes’ were ligated into a· /Votl/Shrimp Alkaline

Phosphatase (Roche)- treated "disintegration" plasmid pSAC35, disclosed in EP-A-286 424 anddescribed by Sleep, D., et al. (1991) Bio/Technology 9, 183 - 187. Ligation mixtures were used totransform chemically-competent E. coli DH5a. Expression plasmids pDB3887 - pDB3897,pSAC35-derivatives containing the “Not\ cassettes”, were identified using standard techniques. 10 Disintegration plasmids pDB3887 - pDB3897 and pDB2244 (For the expression of wild type HSA,described in WO 00/44772) (Table 1) were used to transform S. cerevisiae BXPIOcir0 (aspreviously described WO/2001/079480 as described below.

Table 1: Plasmid, amino acid substitution introduced into HSA

Plasmid Construct PDB2244 HSA PDB3876 HSA D494N pDB3877 HSA D494A PDB3878 HSA E495Q PDB3879 HSA E495A PDB3880 HSA D494Q PDB3881 HSA D494N, T496A pDB3882 HSA T496A pDB3883 HSA E492G PDB3884 HSA E492G, V493P PDB3885 HSA E492P PDB3886 HSA E492H PDB3887 HSA D494N PDB3888 HSA D494A PDB3889 HSA E495Q PDB3890 HSA E495A PDB3891 HSA D494Q PDB3892 HSA D494N, T496A PDB3893 HSA T496A PDB3894 HSA E492G PDB3895 HSA E492G, V493P PDB3896 HSA E492P pDB3897 HSA E492H 40 WO 2011/051489 PCT/EP2010/066572 n/a = Not applicable. pDB3876-pDB3886 are sub-cloning plasmids.

Table 2: Codons used to introduce amino acid substitutions into HSA

Aminoacid Codon Gly GGT Glu GAA Asp GAT Val GTT Ala GCT Arg AGA Lys AAA Asn AAT Met ATG lie ATT Thr ACT Trp TGG Cys TGT Tyr TAT Leu TTG Phe TTT Ser TCT Gin CAA His CAT Pro CCA Stop TAA 10

Method 2. Production of HSA variants D494N+E495Q+T496A and E495Q+T496A A PCR-based method, using a QuickChange Lightening Kit (Statagene), was employed tointroduce point mutations into HSA. Oligonucleotide pairs xAP094 (SEQ ID NO: 5)/xAP095 (SEQID NO: 6) and xAP096 (SEQ ID NO: 7)/xAP097 (SEQ ID NO: 8) were used to generate two HSAvariants (D494N+E495Q+T496A and E495Q+T496A, respectively). Plasmid pDB3927(disclosed inWO 2010/092135) was used as template DNA and the methodology recommended by themanufacturer of the kit was followed. The resulting plasmids were named pDB3995 and pDB3996(contain HSA D494N+E495Q+T496A and E495Q+T496A expression cassettes, respectively).pDB3995 and pDB3996 were digested with BsfEII/BsrBI and the linearised DNA molecules werepurified using standard techniques. One hundred ng of each Bs/EII/BsrBI digested DNA, purifiedusing a Qiagen PCR-Purification kit following the manufacturer’s instructions, was mixedindividually with 100ng Acc65l/BamHI-digested pDB3936) (disclosed in WO 2010/092135) and 41 WO 2011/051489 PCT/EP2010/066572 used to directly transform S. cerevisiae BXP10cir° using the Sigma Yeast Transformation kit described below.

Method 3. Amino acid substitutions in HSA detailed in Table 3

Plasmid pDB3927 (disclosed in WO 2010/092135) (containing an identical nucleotidesequence encoding HSA as in pDB2243) was manipulated to amino acid substitutions within themature HSA protein. Synthetic DNA fragments were generated (GeneArt AG, Germany or DNA2.0Inc, USA) (Nco\/Bsu36\, Avri\ISph\ or Sac\ISph\ fragments), containing point mutations within theHSA-encoding gene to introduce the desired amino acid substitution(s) into the translated proteinsequence. Table 2 details the codons used to introduce the amino acid substitutions into the HSA-encoding gene. The nucleotide sequence of the synthetic fragment encoding unchanged aminoacids (i.e. wild type) was identical to those in pDB3927. Synthetic DNA fragments were sub-clonedinto Nco\/Bsu36\, AvrWSphl-, SacVSph -digested pDB3927 (described in PCT 11527.204-WO) togenerate pDB4006-pDB4010, pDB4083-pDB4101 and pDB4103-pDB4111 and pDB4194,pDB4200,pDB4202 (see Table 3).

Similarly, BamHI/Sa/l fragments containing point mutations in the nucleotide sequenceencoding HSA were generated by gene assembly (DNA2.0 Inc, USA) and ligated into BamHI/Sa/l-digested pDB3964 (described in WO 2010/092135) to produce plasmids pDB3986-pDB3989 (Table3).

The C-terminal string of amino acids from position 573-585 (KKLVAASQAALGL) (SEQ IDNO: 9) in HSA were mutated to those in macaque (PKFVAASQAALA) (SEQ ID NO: 10), mouse(PNLVTRCKDALA) (SEQ ID NO: 11), rabbit (PKLVESSKATLG) (SEQ ID NO: 12) and sheep(PKLVASTQAALA) (SEQ ID NO: 13) serum albumin. The codons used to introduce each aminoacid substitution are given in Table 2. Synthetic DNA fragments (Saci/Sphl) were generated(DNA2.0 Inc, USA) by gene assembly (the nucleotide sequence of the synthetic fragment encodingunchanged amino acids (i.e. wild type) was identical to that in pDB3927) and were sub-cloned intoSacl/Sphl-digested pDB3927 to produce plasmids pDB4114-4117 (Table 3).

Plasmids pDB3883 (Table 1), pDB4094 and pDB4095 (Table 3) were digested withNcol/Sacl and 857bp fragments from each digest were purified before being ligated into Ncol/Sacl-digested pDB4006 or pDB4110 (8.688kb) (Table 3) to produce pDB4156-pDB4161.

Expression plasmids were generated in vivo (i.e. via homologous recombination in S.cerevisiae', a technique referred to as gap repair or in vivo cloning- see Orr-Weaver & Szostak.1983. Proc. Natl. Acad. Sci. USA. 80:4417-4421). Modified plasmids listed in Table 3 were digestedwith BsfEII/BsrBI and the linearised DNA molecules were purified using standard techniques. Onehundred ng of each BsfEII/BsrBI digested DNA, purified using a Qiagen PCR-Purification kit42 WO 2011/051489 PCT/EP2010/066572 following the manufacturer’s instructions, was mixed individually with 100ng Acc65i/BamHl-digested pDB3936 (disclosed in WO 2010/092135) and used to directly transform S. cerevisiaeBXP10cir° using the Sigma Yeast Transformation kit described below. 5 Table 3.

Plasmid Amino acid substitution in HSA pDB3986 HSA H440Q pDB3987 HSA H464Q PDB3988 HSA H510Q PDB3989 HSA H535Q PDB4006 HSAK573A PDB4007 HSA E492T/N503K/K541A PDB4008 HSAK541G pDB4009 HSA K541D pDB4010 HSA D550N pDB4083 HSA D494E/Q417H pDB4084 HSA Q417A PDB4085 HSA P499A pDB4086 HSA K500A PDB4087 HSA K536A pDB4088 HSA P537A PDB4089 HSA K538A pDB4090 HSA E492G/V493P/K538H/K541N/E542D pDB4091 HSA E492P/N503K/K541G/E542P pDB4092 HSA N503K pDB4093 HSA N503H pDB4094 HSA E492G/N503K PDB4095 HSA E492G/N503H PDB4096 HSA E492T pDB4097 HSA N503D pDB4098 HSA E492T/N503D pDB4099 HSA K538H PDB4100 HSAK541A pDB4101 HSA K541N pDB4103 HSA E542D PDB4104 HSA E542P pDB4105 HSA D550E pDB4106 HSA E492H/E501P/N503H/E505D/T506S/T540S/K541E pDB4107 HSA A490D/E492T/V493L/E501P/N503D/A504E/E505K/T506F/K541D pDB4108 HSA E501A pDB4109 HSA E501Q PDB4110 HSA K573P 43 WO 2011/051489 PCT/EP2010/066572

PDB4111 HSA E492G/K538H/K541N/E542D pDB4114 HSA K573P/L575F/G584A PDB4115 HSA K573P/K574N/A577T/A578R/S579C/Q580K/A581D/G584A PDB4116 HSA K573P/A577E/A578S/Q580K/A582T pDB4117 HSA K573P/A578S/S579T/G584A pDB4156 HSA E492G K573A pDB4157 HSA E492G N503K K573A pDB4158 HSA E492G N503H K573A pDB4159 HSA E492G K573P PDB4160 HSA E492G N503K K573P pDB4161 HSA E492G N503H K573P PDB4194 HSA D550E pDB4200 HSA K574N pDB4202 HSA Q580K

Table 4: K500 primers and plasmids

Original primers CODONSUSED XAP216 CTTTGGAAGTCGACGAAACTTACGTTCCAGGTGAATTCAACGCTG Gly GGT (SEQ ID NO: 14) XAP217 CTTTGGAAGTCGACGAAACTTACGTTCCAGAAGAATTGAACGCTG(SEQ ID NO: 15) Glu GAA XAP218 CTTTGGAAGTCGACGAAACTTACGTTCCAGACGAATTCAACGCTG (SEQ ID NO: 16) Asp GAC XAP219 CTTTGGAAGTCGACGAAACTTACGTTCCAGTTGAATTCAACGCTG(SEQ ID NO: 17) Val GTT XAP220 CTTTGGAAGTCGACGAAACTTACGTTCCAAGAGAATTCAACGCTG (SEQ ID NO: 18) Arg AGA XAP221 CTTTGGAAGTCGACGAAACTTACGTTCCAAACGAATTCAACGCTG (SEQ ID NO: 19) Asn AAC XAP222 CTTTGGAAGTCGACGAAACTTACGTTCCAATGGAATTCAACGCTG(SEQ ID NO: 20) Met ATG XAP223 CTTTGGAAGTCGACGAAACTTACGTTCCAATTGAATTCAACGCTG(SEQ ID NO: 21) He ATT XAP224 CTTTGGAAGTCGACGAAACTTACGTTCCAACOGAATTCAACGCTG(SEQ ID NO: 22) Thr ACC XAP225 CTTTGGAAGTCGACGAAACTTACGTTCCATGGGAATTCAACGCTG (SEQ ID NO: 23) Trp TGG XAP226 CTTTGGAAGTCGACGAAACTTACGTTCCATGTGAATTCAACGCTG(SEQ ID NO: 24) Cys TGT XAP227 CTTTGGAAGTCGACGAAACTTACGTTCCATAOGAATTCAACGCTG (SEQ ID NO: 25) Tyr TAC XAP228 CTTTGGAAGTCGACGAAACTTACGTTCCATTGGAATTCAACGCTG(SEQ ID NO: 26) Leu TTG XAP229 CTTTGGAAGTCGACGAAACTTACGTTCCATTCGAATTCAACGCTG Phe TTC 44 •I IDz^_vv WO. 2011/051489 PCT/EP2010/066572 (SEQ ID NO: 27) XAP230 CTTTGGAAGTCGACGAAACTTACGTTCCATCTGAATTCAACGCTG (SEQ ID NO: 28) Ser TCT XAP231 CTTTGGAAGTCGACGAAACTTACGTTCCACAAGAATTCAACGCTG (SEQ ID NO: 29) Gin CAA XAP232 CTTTGGAAGTCGACGAAACTTACGTTCCACACGAATTCAACGCTG (SEQ ID NO: 30) His CAC xAP233 CTTTGGAAGTCGACGAAACTTACGTTCCACCAGAATTCAACGCTG (SEQ ID NO: 31) Pro CCA XAP234 CTTTGGAAGTCGACGAAACTTACGTTCCATAAGAATTCAACGCTG (SEQ ID NO; 32) STOP taa XAP235 GAATTAAGC7TATTACAAACCCAAAGCAGCTTGGGAAGC (SEQ ID NO: 33)

Table 5: K573 primers and plasmids

Original primers CODONSUSED XAP187 ATAAGCCTAAGGCAGCTTGACTTGCAGCAACAAGTTTACCACCCTCCTCG(SEQ ID NO: 34) Gly GGT XAP188 ATAAGCCTAAGGCAGCTTGACTTGCAGCAACAAGTTTTTCACCCTCCTCG(SEQ ID NO: 35) Glu GAA XAP189 ATAAGCCTAAGGCAGCTTGACTTGCAGCAACAAGTTTATCACCCTCCTCG (SEQ ID NO: 36) Asp GAT XAP190 ATAAGCCTAAGGCAGCTTGACTTGCAGCAACAAGTTTAACACCCTCCTCG(SEQ ID NO: 37) Val GTT XAP191 ATAAGCCTAAGGCAGCTTGACTTGCAGCAACAAGTTTTCTACCCTCCTCG(SEQ ID NO: 38) Arg AGA XAP192 ATAAGCCTAAGGCAGCTTGACTTGCAGCAACAAGTTTATTACCCTCCTCG(SEQ ID NO: 39) Asn AAT XAP193 ATAAGCCTAAGGCAGCTTGACTTGCAGCAACAAGTTTCATACCCTCCTCG(SEQ ID NO: 40) Met ATG XAP194 ATAAGCCTAAGGCAGCTTGACTTGCAGCAACAAGTTTAATACCCTCCTCG(SEQ ID NO: 41) He ATT XAP195 ATAAGCCTAAGGCAGCTTGACTTGCAGCAACAAGTTTAGTACCCTCCTCG(SEQ ID NO: 42) Thr ACT XAP196 ATAAGCCTAAGGCAGCTTGACTTGCAGCAACAAGTTTCCAACCCTCCTCG (SEQ ID NO: 43) Trp TGG XAP197 ATAAGCCTAAGGCAGCTTGACTTGCAGCAACAAGTTTACAACCCTCCTCG (SEQ ID NO: 44) Cys TGT XAP198 ATAAGCCTAAGGCAGCTTGACTTGCAGCAACAAGTTTATAACCCTCCTCG(SEQ ID NO: 45) Tyr TAT XAP199 ATAAGCCTAAGGCAGCTTGACTTGCAGCAACAAGTTTQAAACCCTCCTCG (SEQ ID NO: 46) Leu TTG XAP200 ATAAGCCTAAGGCAGCTTGACTTGCAGCAACAAGTTTAAAACCCTCCTCG (SEQ ID NO: 47) Phe TTT XAP201 ATAAGCCTAAGGCAGCTTGACTTGCAGCAACAAGTTT.AGAACCCTCCTCG (SEQ ID NO: 48) Ser TCT 45 ( lowy WO 2011/051489 PCT/EP2010/066572 XAP202 ATAAGCCTAAGGCAGCTTGACTTGCAGCAACAAGTTTTTGACCCTCCTCG (SEQ ID NO: 49) Gin CAA XAP203 ATAAGCCTAAGGCAGCTTGACTTGCAGCAACAAGTTTATGACCCTCCTCG (SEQ ID NO: 50) His CAT XAP204 ATAAGCCTAAGGCAGCTTGACTTGCAGCAACAAGTTTTTAACCCTCCTCG(SEQ ID NO: 51) STOP taa XAP205 AATGCTGCCATGGAGATCTGCTTGAATGTGCTGATG (SEQ ID NO: 52)

Method 4. HSA K500 and K573 permutation library PCR was used to produce two permutation libraries in which the codons encoding aminoacid 500 or 573 of mature HSA were changed (mutated) to alternative non-wild type amino acidsand a termination codons (K5XXSTOP). Mutagenic oligonucletides (Table 4 and Table 5), weredesigned to amplify HSA- encoding DNA and incorporate the desired changes. That is, for thechanges at position 500, pDB4082 (Figure 1) was used as a template DNA. pDB4082 is aderivative of pDB2305 (disclosed in EP 1788084) and was produced as follows. pDB2305 (Figure 2)was digested with NsiVSpe\ and the yielded 8.779kb Nsi\ fragment was self-ligated to producepDB4005 (Figure 3). A synthetic DNA fragment (Bsal/Sphl) was generated by gene assembly(DNA2.0 Inc, USA) (SEQ ID NO: 1) (containing 3’ region of the PRB1 promoter, modified fusionleader sequence, nucleotide sequence encoding HSA and 5’ region of the modified ADH1terminator), and ligated into H/nc/lll/Spril-digested pDB4005 (Figure 3) to produce pDB4082. Note.The H/nc/lll site in PRB1 promoter site has been removed and a Sacll site within the nucleotidesequence encoding HSA has been introduced.

For the permutation library for position 500 of HSA, the nucleotide sequence encoding HSAcorresponding to that between the Sa/I / H/ndlll sites (see plasmid map pDB4082, Figure 1) wasgenerated using the New England Biolabs Phusion kit (Table 6) and oligonucleotides listed in Table4. Table 7 describes the PCR method employed.

The permutation library at amino acid position 573 in HSA was generated using pDB3927as template DNA and involved amplifying the albumin-encoding DNA corresponding to thatbetween the Nco\ and Bsu36l sites using oligonucleotides detailed in Table 5.

Table 6:. PCR ingredients 500 library 573 library 20DI Buffer HF(5X) 2DI dNTP mix(10mM) 2Dl oligonucleotide(10DM) XAP235 XAP205 2 DI oligonucleotide (10OM) XAP216- XAP234 XAP187 XAP204 46 WO 2011/051489 PCT/EP2010/066572

1 DI Phusion polymerase i 1 □ I Template DNA(~5ng) pDB4082 PDB3927 72OI dH2O

Table 7: PCR conditions: 98°C for 2 min 1 cycle 98°Cfor 10sec 35 cycles 57°C for 30sec 72°Cfor20sec 72°C for 5 min 1 cycle

For the albumin variants based at positions 500 and 573, each PCR-product was purifiedusing a Qiagen PCR-clean up kit (according to the manufactures instructions), digested withSa/l/H/ridlll (position 500 library) or Ncol/Bsu36\ (position 573 library). The digested DNAs werethen purified using a Qiagen PCR-clean up kit and ligated into Sall/ Hindlll - or Ncol/Bsu3Ql -digested pDB4082 or pDB3927, respectively, replacing the equivalent native sequence. Ligationswere transformed into E. coli DH5a, subsequent plasmids isolated from transformants using aQiagen miniprep kit (according to the manufacturer’s instructions) and the correct constructsidentified by restriction analysis. This produced a collection of plasmids, pDB4204 - pDB4222(position 500 library) pDB4173 to pDB4190 (position 573 library), containing albumin genes whichdiffered only in their sequence corresponding to the codon for the amino acid at position 500 or 573Table 4 and 5, respectively). The specific changes in each plasmid were confirmed by sequencing.

The resultants plasmids were used to generate expression plasmids and albumin fusionproducing yeast by in vivo cloning as described above. That is, S. cerevisiae was transformedusing the Sigma Yeast Transformation kit (described below), using a mixture of a 100 ngBsfEII/BsrBI-digeste HSA variant containing plasmid and 100 ng Acc65l/BamHI digested pDB3936.

Transformation of S. cerevisiae S. cerevisiae BXP10 cir° (as previously described WO/2001/079480) or Strain A cir°(described in WO/2005/061718) was streaked on to YEPD plates (1% (w/v) yeast extract, 2%(w/v) Bactopeptone, 2% (w/v) glucose), 1.5% agar) and allowed to grow for 4 days at 30°C prior totransformation. One pg of whole plasmid (i.e. circular plasmids) or, for gap repair, 100 ngBstEll/BsrBl- or Λ/s/l/Pvul-digested HSA variant or HSA variant fusion containing plasmid and 100ng Acc65l/BamHI digested pDB3936 were used to transform S. cerevisiae using a Sigma YeastTransformation kit using a modified lithium acetate method (Sigma yeast transformation kit,YEAST-1, protocol 2; Ito et al. (1983) J. Bacteriol., 153, 16; Elble, (1992) Biotechniques, 13, 18).The protocol was amended slightly by incubating the transformation at room temperature for 4 h 47 WO 2011/051489 PCT/EP2010/066572 prior to heat shock. Following heat shock, the cells were briefly centrifuged before being re-suspended in 200μΙ 1M sorbitol then spread over BMMD agar plates, the composition of BMMD isdescribed by Sleep et al., (2001), Yeast, 18, 403. Plates were incubated at 30°C for 4 days beforeindividual colonies were patched on to fresh BMMD plates. Yeast strain numbers are detailed inTable 1.

Stocks were prepared for each yeast strain as follows: BMMD broth was inoculated with aheavy loop of each yeast patch and grown for 24h at 30°C with orbital shaking at 200rpm. Cellswere harvested by centrifugation at 1900 χ g for 5 min in a Sorval RT600 centrifuge, 15mLsupernatant was removed and replaced by trehalose 40% (w/v). The cells were resuspended andtransferred to cyrovials (1mL) for storage at -80°C.

Shake flask growth of S. cerevisiae BMMD (recipe 0.17% (w/v) yeast nitrogen base without amino acid and ammonium sulphate(Difco), 37.8mM ammonium sulphate, 29mM citric acid, 142mM disodium hydrogen orthophosphatedehydrate pH6.5, 2% (w/v) glucose) media (10mL) was inoculated with each yeast strain andgrown for 12h at 30°C with orbital shaking at 200rpm. An aliquot of each starter culture (4mL) wasused to inoculate 2 χ 200mL BMMD media and grown for 36h at 30°C with orbital shaking at200rpm. Cells were harvested by filtration through 0.2pm vacuum filter membranes (Stericup,Millipore) including a GF-D prefilter (Whatman) and the supernatant retained for purification.

Primary concentration

Retained culture supernatant was concentrated using Tangential Flow Filtration using a PallFiltron LV system fitted with a Omega 10KD (0.093sq.m2) filter (LV CentramateTM cassette, PallFiltron) with a transmembrane pressure of 20psi and a recirculation rate of 180mL.min*1.

Fermentation

Fed-batch fermentations were carried out in a 10 L Sartorius Biostat C fermenter at 30°C;pH was monitored and adjusted by the addition of ammonia or sulphuric acid as appropriate. Theammonia also provided the nitrogen source for the cultures. The level of dissolved oxygen wasmonitored and linked to the stirrer speed, to maintain the level at >20% of saturation. Inoculawere grown in shake flasks in buffered minimal media (recipe). For the batch-phase the cultureswas inoculated into fermenter media (approximately 50% of the fermenter volume) containing 2%(w/v) sucrose. The feed stage was automatically triggered by a sharp rise in the level of dissolvedoxygen. Sucrose was kept at growth-limiting concentrations by controlling the rate of feed to a setnominal growth rate. The feed consisted of fermentation media containing 50% (w/v) sucrose, allessentially as described by Collins. (Collins, S.H., (1990) Production of secreted proteins in yeast,in: T.J.R. Harris (Ed.) Protein production by biotechnology, Elsevier, London, pp. 61-77). GP-HPLC quantitation 48 WO 2011/051489 PCT/EP2010/066572

Purified albumin variants, fusions and conjugates were analysed by GP-HPLC andquantification as follows. Injections of 25pL were made onto a 7.8mm id x 300mm length TSKG3000SWXL column (Tosoh Bioscience), with a 6.0mm id x 40mm length TSK SW guard column(Tosoh Bioscience). Samples were chromatographed in 25mM sodium phosphate, 100mM sodiumsulphate, 0.05% (w/v) sodium azide, pH 7.0 at 1 mL/min, Samples were quantified by UV detectionat 280nm, by peak area, relative to a recombinant human albumin standard of known concentration(10mg/mL) and corrected for their relative extinction coefficients.

Purification of albumin variants from shake flask

Albumin variants were purified from shake flask (either culture supernatant or concentratedculture supernatant) using a single chromatographic step using an albumin affinity matrix(AlbuPure™ - ProMetic BioSciences, Inc.). Chromatography was performed at a constant linearvelocity of 240cm/h throughout. Culture supernatant was applied to a 6cm bed height, 2.0mLpacked bed pre-equilibrated with 50mM sodium acetate pH5.3. Following load the column waswashed with 10 column volume (CV) of equilibration buffer, then 50mM ammonium acetate pH8.0(10CV). Product was eluted with either 50mM ammonium acetate 10mM octanoate pH8.0, 50mMAmmonium Acetate 30mM Sodium Octanoate 200mM Sodium Chloride pH7.0 or 200mMPotassium thiocyanate. The column was cleaned with 0.5M NaOH (3cv) and 20mM NaOH (3.5cv).Eluate fraction from each albumin variant were concentrated and diafiltered against 10 volumes of50mM sodium chloride (Vivaspin20 10,000 MWCO PES with optional diafiltration cups, Sartorius).Purified albumin variants were quantified by GP-HPLC as described above.

Purification of albumin-fusion variants from shake flask

Albumin-fusion variants were purified from shake flask culture supernatant using a singlechromatographic step using an albumin affinity matrix (AlbuPure™ - ProMetic BioSciences, Inc.).Chromatography was performed at a constant linear velocity of 240cm/h throughout. Culturesupernatant or concentrated culture supernatant was applied to a 6cm bed height, 2.0mL packedbed pre-equilibrated with 50mM sodium acetate pH5.3. Following load the column was washedwith 10 column volume (cv) equilibration buffer then 50mM ammonium acetate pH8.0 (10cv).Product was eluted with either 50mM ammonium acetate 10mM octanoate pH8.0, 50mMAmmonium Acetate 30mM Sodium Octanoate 200mM Sodium Chloride pH7.0, 50mM AmmoniumAcetate 100mM Sodium Octanoate pH9.0 or 200mM Potassium thiocyanate. The column wascleaned with 0.5M NaOH (3cv) and 20mM NaOH (3.5cv). Eluate fraction from each albuminvariant-fusion were concentrated and diafiltered against 10 volumes of 25 mM Tris, 150 mM NaCI,2 mM KCI, pH 7.4 (Vivaspin20 10,000 MWCO PES with optional diafiltration cups, Sartorius).Purified albumin-fusion variants were quantified by GP-HPLC as described above.

Purification of albumin variants from fermentation 49 WO 2011/051489 PCT/EP2010/066572

Albumin variants were purified from high cell density fed batch fermentation supernatantsafter separation by centrifugation, using a Sorvall RC 3C centrifuge (DuPont). Culture supernatantwas chromatographed through an 11cm bed height column 8.6mL packed bed packed with acustom synthesised albumin affinity matrix (AlbuPure™ - ProMetic BioSciences, Inc.) as describedabove. Product was eluted using elution buffers describe above at a flow rate of 120cm/h. Theeluate fraction(s) was analysed by GP-HPLC. (above) .and reducing SDS-PAGE for purity and ifrequired concentrated (Vivaspin20 10,000 MWCO PES) and applied to a 2.4x96cm column packedwith Superdex 75 run at a flow rate of 39cm/h in 25 mM Tris, 150 mM NaCI, 2 mM KCI, pH 7.4 .The peak was fractionated, assayed by GP-HPLC and pooled in order to generate the monomericprotein of interest. Pooled fractions were concentrated (Vivaspin20 10,000 MWCO PES,Sartorius).

All proteins to be assayed for receptor (FcRn) binding properties and or other analysis werequantified by GP-HPLC as described above corrected fortheir relative extinction coefficients.

Example 2. Determination of receptor (shFcRn) binding properties of blood derived HSAand recombinant human albumin

Essentially fatty acid-free HSA (Sigma-Aldrich) was further purified by size exclusionchromatography as described in Andersen et a/(2010). J.Biol.Chem. 285, (7),4826-4836. Ten μΜof monomeric HSA and rHA were analysed using SPR as described above and the data presentedin Figure 4.

Direct comparison of HSA (blood derived) with recombinant human albumin (Recombumin)at the same concentration (10μΜ) (Figure 4A and 4B) shows for both samples binding toimmobilized shFcRn (pH6.0, pH7.4 respectively) was reversible and pH dependent. In addition,comparison of HSA vs recombinant human albumin by Bosse et al (2005).. J. Clin. Pharmacol. 45;57-67, demonstrated equivalent half life in vivo human study

Example 3. Determination of receptor (shFcRn) binding properties of albumin variants

Two established FcRn binding assays were used, ELISA and SPR. There are majordifferences between the assays: In the ELISA system HSA is coated directly in wells and shFcRn-GST is added in solution whereas in the SPR assay shFcRn-GST is immobilized to a CM5 chip andHSA injected in solution. The pH can be varied in both systems.

The variants were analysed using ELISA at pH 6.0 and pH 7.4. Results are disclosed inFigure 5. The ELISA values represent the mean of duplicates. 50 WO 2011/051489 PCT/EP2010/066572

The variants were analysed using SPR analysis at pH 6.0 and pH 7.4. Results are disclosedfor a representative number of variants in Figure 6 using a concentration of the variants of 0.2 μΜand in Figure 7 using a concentration of the variants of 1 μΜ.

The SPR data disclosed in Figures 6 and 7 were normalized and the relative binding ofvariants at each concentration is shown in Figure 8 A and B respectively.

The conclusions of the analysis are that all tested variants have the characteristic binding tothe receptor at pH 6.0 but no binding at pH 7.4. The variants D494N,Q,A, E495Q,A, T496A, andD494N+T496A show reduced binding to the receptor compared to HSA.

Example 4. Determination of receptor (shFcRn/smFcRn) binding properties of albuminvariants

Using the SPR analysis method below the association constant Ka , the dissociationconstant Kd and the binding constant KD calculated for HSA and mouse serum albumin (MSA)binding to human and mouse FcRn (Table 8). SPR analyses- SPR analyses were performed on a BIAcore 3000 instrument (GEHealthcare) using CM5 chips and immobilization of smFcRn-GST and shFcRn-GST variants orsmFcRn was performed using the amine coupling kit (GE Healthcare). Protein samples (10 pg/ml)were injected in 10 mM sodium acetate at pH 4.5 (GE Healthcare), all as described by themanufacturer. Unreacted moieties on the surface were blocked with 1 M ethanolamine. For allexperiments, phosphate buffer (67 mM phosphate buffer, 0.15 M NaCI, 0.005% TWEEN® 20) at pH6.0 or pH 7.4, or HBS-P buffer (0.01 M HEPES, 0.15 M NaCI, 0.005% surfactant P20) at pH 7.4were used as running buffer or dilution buffer. Kinetic measurements were performed using a lowdensity immobilized surface (100-200 resonance units (RU)). Serial dilutions of hlgG1 (2000.0-31.2nM), mlgG1 (1000.0-15.6 nM), MSA (20.0-0.3 μΜ) and HSA (200.0-3.1 μΜ) were injected at pH 6.0or pH 7.4, at a flow rate 50 μΙ/minute at 25°C. Additive binding was recorded by injecting HSA (10μΜ), MSA (5 μΜ), hlgG1 (100 nM) or mlgG1 (100 nM) alone or two at a time at 25°C at 20μΙ/minute at pH 6.0 over immobilized shFcRn (-600 RU) or smFcRn (-600 RU). Competitivebinding was measured by injecting shFcRn (50 nM) or smFcRn (100 nM) alone or together withdifferent amounts of HSA or MSA (10.0-0.05 μΜ) over immobilized HSA (-2600 RU)or MSA(-2000 RU). In all cases, to correct for nonspecific binding and bulk buffer effects, responsesobtained from the control surfaces and blank injections were subtracted from each interactioncurve. Kinetic rate values were calculated using predefined models (Langmuir 1:1 ligand model,heterogeneous ligand model and steady state affinity model) provided by the BIAevaluation 4.1software. The closeness of the fit, described by the statistical value χ2 that represents the meansquare, was lower than 2.0 in all affinity estimations. 51 WO 2011/051489 PCT/EP2010/066572

Table 8: Binding constants of HSA and MSA shFcRn and smFcRn.

AlbuminSpecies FcRnSpecies Ka (103/Ms) Kd(103/s) KD(μΜ) KDReq.(μΜ) MSA Mouse 4.2±0.5 39.4±3.1 9.3±0.4 MSA Human 3.8±0.0 3.1±0.1 0.8+0.2 ND HSA Mouse NA NA NA 86.2±4.1 HSA Human 2.7±1.3 12.2±5.9 4.5+0.1 4.6 ±0.5

The KD’s were generated using the BIAevaluation 4.1 software) A Langmuir 1:1 ligand model was used throughout. The kinetic values represent the average of triplicates. ND means:Not determined. NA means: Not acquired

Example 5. Binding of albumins from other species to human FcRn

Commercially available animal albumin (either Sigma-Aldrich or Calbiochem) were furtherpurified as described in Andersen et al (2010). J.Biol.Chem. 285, (7),4826-4836. The binding ofdonkey serum albumin, bovine serum albumin, goat serum albumin, sheep serum albumin, rabbitserum albumin, dog serum albumin, hamster serum albumin, guinea pig albumin, rat serumalbumin and chicken serum albumin to shFcRn was determined using the techniques described inMaterials and Methods.. The ELISA results are disclosed in Figure 9 A-D and the relative bindingssummarized in Figure 9 E.

The SPR results are shown in Figure 10, where the binding at pH 6.0 and pH 7.4 for eachalbumin species are shown. Table 10 shows an overview of the relative binding responsesmeasured using ELISA and SPR:

Table 10: Cross-species albumin-FcRn binding

Albuminspecie shFcRn ELISA SPR pH6.0 pH7.4 pH6.0 pH7.4 Human ++(+) - ++(+) - Donkey +++ - ++ 25 Cow ++ - ++ - Sheep +/- - - - Goat +/- - - - Rabbit ++++ - +++ - Dog NDa ND +++ ~—3O- G. pig ++++ + ++++ + Hamster +++ - +++ - Rat +++ - +++ - Mouse +++ - +++ □ ς Chicken - - - JJ 52 WO 2011/051489 PCT/EP2010/066572

Relative binding responses are categorized from strongest (++++) to weakest (+) and no binding (-)a: Not determined (ND). A hierarchy ranging from strongest to weakest binding is as follows; guinea pig =/> rabbit >hamster/dog > rat/mouse > donkey > human > bovine > goat/sheep > chicken. This data showsthat animal albumins have different affinities for shFcRn.

Example 6. Kinetics of the HSA variant for shFcRn

The binding constants for variants according to the invention were determined according tothe methods described in Materials and Methods.

Table 11: Binding constants of HSA variants for shFcRn

t Albumin i Ka kd / KD KD Req i (i Variant (103/Ms) j (10'3/s) (μΜ) (μΜ) WT 3 2±0.2 : 15.5±2.5 4.8 5.4 D494N 17+0 0 i 18.6±0 0 10.9 11.8 ; D494A 2 3±0 1 : 53.4±0.3 23 2 170 = D494Q 2 1 ±0.0 58.2±3.8 27.7 ND = E495Q 2 5±0.0 24.1 ±0.2 9.6 10.9 E495A 2 1±0-° 14.0±0.0 7.0 8.6 D494N+T496A 2.5±0.0 11.0±0.0 4.4 55 \\ T496A 2.3±0 0 11.7±0.5 5 1 7.1 [ E492G Ϊ 4.1±0.0 11.0±0.0 2.7 ND

The KD’s were generated using the BIAevaluation 4.1 software) A Langmuir 1:1 ligandmodel was used throughout. The kinetic values represent the average of triplicates. ND means:Not determined.

The results correspond with the conclusions made in Example 3 based on SPR and ELISAdata but in addition shows that E492G has increased affinity to its receptor,

Example 7. Competitive analysis of the HSA variants

Competitive analysis of the HSA variants prepared in example 1 and WT HSA wasperformed using the methods described in example 4. Results are shown in Figure 15.

The results show that the variant E492G, unlike E492H E492P and E492G+V493P, has strongerbinding to shFcRn than HSA..

Example 8. Analysis of Q417 substitutions

Using the method of Example 1 variants of HSA having the substitutions Q417A andD494E+Q417H were constructed. The kinetic properties of these variants were tested using themethods in Materials and Methods and are shown in Table 12. 53 WO 2011/051489 PCT/EP2010/066572

Table 12: Binding constants of HSA variants for shFcRn

Albumin variant3 ka : kd : KDb KD Reqc h (103/Ms) (10‘3/s) : (μΜ) (μΜ) WT J3.2+0 2 15 5+2.5 4 8 54 Q417A 3.2±0 1 26.0±0 0 8.1 . ND D494E+Q417H ; 3.1±0.1 20.5±0 5 . 6.6 ND a: Dilutions of HSA variants were injected over immobilized shFcRn (-1500 RU). b: The kinetic rate constants were obtained using a simple first-order (1:1) bimolecular interactionmodel. c: The steady state affinity constant was obtained using an equilibrium (Req) binding modelsupplied by the BIAevaluation 4.1 software. The kinetic values represent the average of triplicates,d: Not determined (ND).

The data show that variants Q417A and D494E+Q417H bind weaker to the receptor thanthe wild-type HSA.

Example 9. Analysis of HSA variants in position 499, 500, 536, 537, 538 and 573

Using the method of Example 1 variants of HSA having the substitutions P499A, K500A, K536A, P537A, K538A and K573A were constructed. The receptor binding properties of thesevariants were tested as described in Materials and Methods. Results are shown in Figure 11.

The data demonstrated that variants P499A, K536A, P537A and K538A had a reducedbinding affinity to shFcRn relative to HSA. Variant K500A had almost completely lost its ability tobind to shFcRn and K573A had an increased binding affinity to shFcRn both relative to HSA.

Example 10. Analysis of variants in position 501 of HSA

Using the method of Example 1 variants of HSA having the substitutions E501A and E501Qwere constructed. The kinetic properties of these variants were tested as described in Materialsand Methods.

Table 13 Binding constants of HSA variants for shFcR

Albumin variant3 ka (103/Ms) kd (10‘3/s) KDb(μΜ) KD Reqc i! (μΜ) WT 3.2±0 2 15.5±2.5 4.8 : i 5.4 E501A 3.3±0.0 26.0+0 0 7.8 i : ND E501Q 2.7 ±0.1 15.5+0.5 5.7 ί ND i a: Dilutions of HSA variants were injected over immobilized shFcRn (-1500 RU). b: The kinetic rate constants were obtained using a simple first-order (1:1) bimolecular interactionmodel. c: The steady state affinity constant was obtained using an equilibrium (Req) binding model supplied by the BIAevaluation 4.1 software. The kinetic values represent the average of triplicates d: Not determined (ND). 54 WO 2011/051489 PCT/EP2010/066572

The data shows that variants E501A and E501Q have a slightly decreased binding affinity toshFcRn relative to HSA.

Example 11. Analysis of HSA variants in position 573 5 Using the method of Example 1 variants of HSA having a substitution at position 573 were constructed. All variants at position 573 were generated and the receptor binding properties ofthese variants were tested as described in Materials and Methods but with SPR analysis performedat pH5.5. Results are shown in the table 14 below and Figure 12 and 13. 10 Table 14 /.Kinetics of HSA K573.single point mutants.

Albumin variant3 ka (103/Ms) ii kd • (10’3/s) / kbbh (nM) WT 9 0±0.0 6.910.1 i 766 K573A 7 4+0.0 | 2.210.0 h 297 Ϊ K573C 4.210.0 p 1.110.2 / 262 L K573D 7 9+0.2 4 1+0 3 /518 1 K573E 9 010.0_______ 2.910.0 j 322 ii K573F 7.810.1 |i 0.510.1 / 74 ii K573G 8.510.0 P 1.810.1 H 212 ii K573H 12.010.2 0.810.0 i 68 / K573I 8 6+0.0 0.810.2 99 K573L / 5.1+0.2 ... 2.310.1 :. 451 K573M 8.6+0.0 1.910.0 / 221 K573N 7.3+0.2 1.110.3 151 K573P 9.810.0 0.610.1 61 K573Q 7.7±0.2 2.610.0 338 K573R 8.510.0 3.010.2 353 K573S 7.910.2 1.210.2 = 152 K573T 8 7+0.2 1.110.1 126 K573V 8.1+0 0 0 6+0.2 i 80 K573W 15.0+0.2 I: 0.410.3 : 29 K573Y 22.010.1 0.510.1 /23 ί K573STOP ND ND 141000 a: Dilutions of HSA variants were injected over immobilized shFcRn (-1500 RU). b: The kinetic rate constants were obtained using a simple first-order (1:1) bimolecular interactionmodel. 15 c: The steady state affinity constant was obtained using an equilibrium (Req) binding model supplied by the BIAevaluation 4.1 software. The kinetic values represent the average of duplicates, d: Not determined (ND). 55 WO 2011/051489 PCT/EP2010/066572

The results show that all variants having substitution in position 573 have improved bindingto shFcRn compared with WT HSA. In particular the variants K573F, K573H, K573P, K573W andK573Y have more than 10 fold lower KD to shFcRn than the parent HSA. The variant K573STOPis a truncated albumin having a stop codon in position 573. The sensorgram for the K573STOP5 variant show significantly reduced binding compare to the WT HSA and generated a high KD. Theincreased affinity that we have shown for the variant K573E, a natural variant characterized byOtagiri (2009). Biol.Pharm. Bull. 32(4) 527-534, is predicted to have increased half-life in vivo.

Example 12. Analysis of further HSA variants 10 Using the method of Example 1 variants of HSA having the substitutions E492G, E492G+N503H, N503H, D550E, E492G+N503K, E542P, H440Q, K541G, K541D, D550NE492G+K538H+K541N+E542D, E492T+N503K+K541A, E492P+N503K+K541G+E542P,

E492H+E501P+N503H+E505D+T506S+T540S+K541E, A490D+E492T+V493L+E501P +N503D+A504E+E505K+T506F+K541D, E492G+V493P+K538H+K541N+E542D were 15 constructed. The receptor binding properties of these variants were tested as described inMaterials and Methods, and the results are shown in Table 15 and Figure 14.

Table 15: Binding constants of HSA variants for shFcR

Albuminvariant3 i Ka (103/Ms) kd (10‘3/s) KDb(μΜ) ί KD Reqc I (μΜ) ϊ WT 3 2+0.2 15.5+2 5 4.8 5.4 E492G 4 1+0.0 11.010.0 2.7 ND E492G/N503H 6 9±0:1 14.5+0.5 2.1 ND N503H 5 4+0.0 24.0+0 1 4.4 ND D550E 3 2±0'4 11.8+0 0 3.6 : ND E492G/N503K : 5.910.1 16.0+0.0 2.7 ND . E542P :. 3.410.0 15.7+0.2 4.7 ND H440Q 3.2+0.1 ..20.8+0.0 6.5 ND K541G 3.210.0 23.0+0.0 7.1 ND ? K541D 2.610.0 ; 24.010.0 : 9.2 ND I D550N 2.510.0 i 30.0+0.0 i 12.0 ND 20 a: Dilutions of HSA variants were injected over immobilized shFcRn (-1500 RU). b: The kinetic rate constants were obtained using a simple first-order (1:1) bimolecular interactionmodel. c: The steady state affinity constant was obtained using an equilibrium (Req) binding model supplied by the BIAevaluation 4.1 software. The kinetic values represent the average of triplicates. 25 d: Not determined (ND). 56 WO 2011/051489 PCT/EP2010/066572

The results show that for position 550, a substitution to E results in an increased affinitywhilst a substitution to N resulted in reduced affinity for shFcRn at pH6.0. When this analysis wasrepeated for the D550E substitution at pH5.5 however no observable increase in affinity was seen.The substituted for an acid amino acid (E) maintains and improves the binding. However thesubstitution for an uncharged amide amino acid reduces binding at pH6.0. Based on thisobservation, we would predict for this position that substitutions to basic amino acids (Η, K and R)would result in further reductions in binding.

Example 13. Mutations in His residues

The following variants were generated using the methods described in Example 1: H440Q,H464Q, H510Q and H535Q.. Figure 15 shows SPR sensorgrams of these variants interacting withshFcRn as described in Materials and Methods.

It was found that the variant H440Q bound with comparable affinity as HSA. In contrastH464Q, H510Q and H535Q had significantly reduced affinity to shFcRn. This supports thepreviously published observations that mutagenesis of these Histidine residues significantlyreduced HSA binding to shFcRn (Wu et al (2010). PEDS,23(10)789-798). Wu et al show areduced half-life for a diabody fusion proteins (scFv-DIII)2 in mice with an order of removal fromslowest to fastest: Db-DIII WT>H535A>H510A>H464A>Db. Based on affinity to shFcRn and whencompared to smFcRn (example 5) we would predict the clearance order in humans to be (forglutamine (Q) substitutions) WT> H440Q>H510Q>H464Q>H535Q.

Example 14. Further variants

The following variants were generated using the methods described in Example 1: K574Nand Q580K in HSA. Binding of the variants to FcRn was tested using the SPR assay as describedin Materials and Methods and the results are shown in Table 16.

The results show that variants K574N and Q580K bound stronger to shFcRn.

Table16: Following kinetic data was found for these variants:

Albumin variant [ ka I kd H KD (103/Ms) i' (w3/s) (μΜ) WT 9.7+0.0 30 0±0.1 3 1 K574N i: 4.9+01 8 4±0.1 H 1-7 Q580K 6.0±0.0 9 3±0.0 L 1.5

Example 15. Analysis of HSA variants in position 500 57 WO 2011/051489 PCT/EP2010/066572

Using the method of Example 1 variants of HSA having a substitution at position 500 wereconstructed. All variants at position 500 were generated and the receptor binding properties ofthese variants were tested. Biacore X, Biacore X100 and Sensor Chip CM5 were used for allanalyses, both supplied by G E Healthcare. shFcRn produced by GeneArt AG (Germany) (diluted5 to 10pg/mL in 10mM sodium acetate pH5.0 (G E Healthcare)) was immobilised on flow cell 2 (FC2)to levels between 1600 - 2200 response units (RU) via standard amine coupling as permanufacturers instructions (G E Healthcare). A blank immobilisation was performed on flow cell 1(FC1) for it to serve as a reference cell. To stabilise the assay, 3-5 start up cycles were run first,with running buffer (67mM phosphate buffer, 0.15M NaCI, 0.005% Tween 20 at pH5.75 ±0.25) only,10 followed by regeneration. WT rHA and K500 library variants were injected at various concentrations(1μΜ - 150μΜ) for 90s at a constant flow rate of (30pl/min) at 25 °C followed by regeneration ofthe surface using HBS-EP buffer pH7.4 (G E Healthcare) until approximate initial baseline RU wasrestored (usually 12s pulse would suffice).

Results are shown in the Table 17 and Figure 16 15 Table 17: Kinetics of HSA K500 single point mutants. : Albumin = ka kd b KDb KD Reqc ί variant (103/Ms) (10_3/s) (μΜ) j: (μΜ) \ K500R 4 42 7.21 1 63 K500I 5 18 10.9 2.1 WT 4.24 9.2 2 2a K500L 3.73 : 11.9 I: 3.2 K500Q : 1.07 : 3.4 3.2 ; K500V 3.29 : 11.0 3.3 j K500Y 3.97 : 14.6 «·· 3.7 K500M 2.48 : 21.5 87 K500T 1 2 i 13.4 11.2 j K500W 0 5 5.4 11.7 K500N 1 3 ......... .J :18.2 14 j K500F 5 17 73.7 143 I K500H 1' 4 63.8 r 16 | K500P = . ND ND i; ND 51’ j K500C 2.38 .:: 124 ; 52 I K500S ND ND- ND 70.2* ; K500A 2 61 208 79.9 i K500D ND ND ND 83.3’ | K500G ND ND ND 95.4’ j K500E = KD not calculable see Figure 16 ; K500STOP Null binder 58 WO 2011/051489 PCT/EP2010/066572 a: Mean of 4 values. b: The kinetic rate constants were obtained using a simple first-order (1:1) bimolecular interactionmodel. c: The steady state affinity constant was obtained using an equilibrium (Req) binding modelsupplied by the BIAevaluation 4.1 software.

The results show for variants K500R and K500I have increased and comparable affinity forshFcRn compared to WT HSA respectively. Variant K500E bound tightly to immobilised shFcRnbut still demonstrated the characteristic pH-dependency of the FcRn interaction. This complex wasvery stable, such that kinetic analysis was not possible (Figure 16). All other variants have reducedbinding to shFcRn than wt rHA.

All variants bound to shFcRn (to some extent) at pH5.5. No binding of K500 library variantsto shFcRn was detectable at pH7.4.

Example 16. Fusion polypeptides

The generation of albumin fusions containing albumin muteins

Plasmids containing expression cassettes for the production of scFv (vHvL) genetically-fused to HSA, at either the N- or C-terminus or both, (described in, Evans et al., 2010. ProteinExpression and Purification. 73,113-124) were modified to allow the production of albumin fusionsusing in vivo cloning (describe above). That is, pDB3017 (Figure 17), pDB3021 (Figure 18),pDB3056 (Figure 19) were digested with Nsi\/Spe\ and /Vs/I fragments corresponding 9.511 kb,9.569kb and 8.795kb, respectively, were purified using standard techniques. Purified Λ/s/lfragments were self-ligated and used to transform chemically competent E. coli DH5a to producepDB4168, pDB4169 and pDB4170, respectively (Table 18).

Similarly, pDB3165 (containing the bivalent fusion) (Figure 20) was digested with Nott andthe expression cassette (4.506kb fragment) was purified before being ligated into Wl-digestedpDB3927 to produce pDB4172 (Figure 21, Table 18).

Synthetic Sal\IBsu36l DNA fragments (269bp), which contain point mutations within thealbumin encoding nucleotide sequence to introduce amino acid substitutions corresponding toK500A, or D550N or K573P into the translated albumin protein sequence, were generated by geneassembly (GeneArt AG, Germany). The SalUBsu36l fragments were individually ligated intoSa/l/Bsu36l-digested pDB4168-pDB4170 and pDB4172 and used to transform chemicallycompetent E. coli DH5a using standard techniques to generate plasmids pDB4265 - pDB4276(Table 18).

Table 18: Albumin variant fusions

Plasmid Construct 59 WO 2011/051489 PCT/EP2010/066572 pDB3017 scFv (anti-FITC) - HSA - FLAG pDB3021 HSA - GS linker - scFv (anti-FITC ) - FLAG pDB3056 HSA-FLAG pDB3165 scFv (anti-FITC) - HSA - GS linker - scFv (anti-FITC) - FLAG pDB4168 scFv (anti-FITC) - HSA - FLAG PDB4169 HSA - GS linker - scFv (anti-FITC) - FLAG PDB4170 HSA-FLAG PDB4172 scFv (anti-FITC) - HSA - GS linker - scFv (anti-FITC ) - FLAG PDB4265 scFv (anti-FITC ) - HSA K500A - FLAG pDB4266 scFv (anti-FITC ) - HSA D550N - FLAG pDB4267 scFv (anti-FITC ) - HSA K573P - FLAG pDB4268 HSA K500A - GS linker - scFv (anti-FITC) - FLAG pDB4269 HSA D550N - GS linker - scFv (anti-FITC) - FLAG pDB4270 HSA K573P - GS linker - scFv (anti-FITC) - FLAG pDB4271 HSA K500A - FLAG pDB4272 HSA D550N - FLAG PDB4273 HSA K573P - FLAG pDB4274 scFv (anti-FITC) - HSA K500A - GS linker - scFv (anti-FITC) - FLAG pDB4275 scFv (anti-FITC) - HSA D550N - GS linker - scFv (anti-FITC) - FLAG pDB4276 scFv (anti-FITC) - HSA K573P - GS linker - scFv (anti-FITC) - FLAG pDB4277 scFv (anti-FITC) - HSA K573A - FLAG pDB4278 HSA K573A - GS linker - scFv (anti-FITC) - FLAG pDB4279 HSA K573A - FLAG pDB4280 scFv (anti-FITC) - HSA K573A - GS linker - scFv (anti-FITC) - FLAG pDB4281 HSA K500A - GS linker - scFv (anti-FITC) pDB4282 HSA D550N - GS linker - scFv (anti-FITC) PDB4283 HSA K573P - GS linker - scFv (anti-FITC) pDB4284 HSA - GS linker - scFv (anti-FITC) pDB2613 HSA- GS linker -IL1RA (N84Q) pDB4285 HSA K573A- GS linker -IL1 RA (N84Q) pDB4286 HSA D550N- GS linker -IL1 RA (N84Q) pDB4287 HSA K500A- GS linker -IL1 RA (N84Q) pDB4288 HSA K573P- GS linker -IL1 RA (N84Q)

Similarly, a DNA fragment was generated by PCR (using standard techniques), to introducea K573A substitution in the translated albumin protein sequence. PCR was performed using theNew England Biolabs Phusion kit using pDB4267 (Figure 22) as template DNA and5 oligonucleotides xAP238 (SEQ ID NO: 53) and xAP239 (SEQ ID NO: 54):

Table 19 describes PCR cycling.

Table 19: PCR cycling 60 WO 2011/051489 PCT/EP2010/066572 98°C for 2 min 1 cycle 98°C for 10sec 35 cycles 57°C for 30sec 72°Cfor 10sec 72°C for 5 min 1 cycle

The PCR-product was purified, digested with Sal\/Bsu36\, and the fragment (269bp) isolatedwas ligated into Sa/l/Bsu36l-digested pDB4168-pDB4170 and pDB4172 and used to transformchemically competent E. coli DH5a. Resulting plasmids (pDB4277 - pDB4280) are listed in Table 18.

The nucleotide sequence encoding the FLAG tag was removed from plasmids pDB4168and pDB4268-4270 (plasmids for the expression of scFv N-terminally fused to HSA and HSAmuteins K500A, D550N and K573P, respectively. pDB4168 and pDB4268-4270 (Table 18) weredigested with Bsu36\ISph\ to remove a 231 bp product comprising 3’ region of HSA-encoding gene,nucleotide sequence encoding FLAG tag and 5’ region of ADH1 terminator. A Bsu36\/Sph\fragment (207bp), comprising 3’ region of HSA-encoding gene and 5’ region of mADH1 terminator(SEQ ID1) from pDB4181 was ligated into Bsu36l/Spftl-digested pDB4168 and pDB4268-pDB4270using standard techniques. Ligation mixtures were used to transform chemically competent E. coliDH5a using standard techniques to generate plasmids pDB4281-pDB4284 (Table 18) pDB4265-pDB4284 were digested with BstEII/BsrBI and the linearised DNA molecules werepurified using standard techniques. One hundred ng BstEII/BsrBI DNA samples were mixed with100ng Acc65l/BamHI-digested pDB3936 and used to transform S. cerevisiae BXP10cir° using theSigma Yeast Transformation kit described below. In each case the expression plasmid wasgenerated in the yeast by homologous recombination {in vivo cloning) between the albumin-fusioncontaining plasmid (pDB4265-pDB4280) (Table 18) and pDB3936.

Plasmids pDB3017, pDB3021, pDB3056 and pDB3165 (wild type HSA fusions, describedby Evans et al., 2010. Protein Expression and Purification. 73,113-124) were used to transform S.cerevisiae Strain Acir0 (described in WO/2005/061718) using the Sigma Yeast Transformation kitdescribed below.

The nucleotide sequence encoding human IL-1RA (interleukin-1 receptor antagonist)(accession number: CAA59087) could be synthetically generated by gene assembly. Thenucleotide sequence of the 708bp synthetic fragment (Bsu36l/Sphl fragment) is given in SEQ IDNO: 55 and includes the 3’region of the gene encoding HSA, the nucleotide sequence encoding aGS linker, the nucleotide sequence encoding human IL-1RA (N84Q to abolish the N-linkedglycosylation motif) and the 5’ region of the ADH1 terminator. The synthetic DNA fragment couldbe ligated into Bsu36l/Sphl-digested pDB3927 to produce pDB2588. 61 WO 2011/051489 PCT/EP2010/066572

Plasmids containing the expression cassettes for the production of IL-1RA genetically fusedto the C-terminus of HSA and the HSA variants K500A, D550N, K573A and K573P were preparedas follows. pDB2588 was digested with Bsu36l/Sphl and a 705bp fragment containing the '3 regionof the HSA encoding gene, nucleotide sequence encoding a GS linker, nucleotide sequenceencoding human IL1-RA (N84Q) and the 5’ region of a modified S. cerevisiae ADH1 terminator(SEQ ID3) was purified using standard techniques then ligated into Bsu36l/Sphl-digested pDB4006(containing HSA K573A expression cassette), pDB4010 (containing HSA D550N expressioncassette), pDB4086 (containing HSA K500A expression cassette), pDB4110 (containing HSAK573P expression cassette) to generate pDB4287, pDB4286, pDB4285 and pDB4288, respectively(for an example, see Figure 23). pDB4285-pDB4288 were digested with Nsil/PvuI and thelinearised DNA molecules were purified using standard techniques. One hundred ng Nsil/Pvul-digested DNA samples were mixed with 100ng Acc65l/BamHI-digested pDB3936 (9721 bp) (i.e. invivo cloning) and used to transform S. cerevisiae (i.e. by in vivo cloning) using the Sigma YeastTransformation kit described below.

Preparation of an S. cerevisiae strain expressing wild type HSA genetically fused to a GSlinker and IL1-RA (N84Q) (see Table 18) could also be generated following the methods describedabove.

The fusion polypeptides were analysed for their binding to FcRn using the SPR methoddescribed above and following results were obtained:

Table 20: Kinetics of HSA fusion variants.

: Albumin variant3 i ka (103/Ms) / kd (10'3/s) l· KDb (μΜ) HSAWT 9.710.0 30.010.1 3.1 K574N : 4.9101 8.410.1 : 1.7 Q580K 6.010.0 i; 9.310.0 : 1.5 K573P 2.810.0 0.410.0 i 0.1 / HSA-WT-FLAG 8.210.2 24.0+0.2 H 2.9 L HSA-D550N-FLAG 5.910.0 49.010.1 83 HSA-K500A-FLAG NDC ND ND HSA-K573A-FLAG 6.110.1 7.110.1 / 1.1 HSA-K573P-FLAG 6.210.1 1.210.1 02 HSA-WT-IL1RA 6.210.0 25.010.2 ’ 4.0 HSA-K500A-IL1RA ND ND ND HSA-D550N-IL1RA 7 310 2 38.010.0 52 HSA-K573A-IL1RA 6 110.0 7.110.1 : 1.1 l· HSA-K573P-IL1RA 6.210.1 1.310.1 E 0.2 [ scFv-HSA-K500A-Fl_AG ; ND ND l· ND 62 WO 2011/051489 PCT/EP2010/066572 scFv-HSA-D550N-FLAG 6.2±0.0 18.0±0.0 2.9 scFv-HSA-K573A-FLAG 6 4±0 1 < 5.7±0.2 0 9 scFv-HSA-K573P-FLAG 5 8±0 0 1.1±0.1 0.2 scFv-HSA-WT-scFv- FLAG .. 7.5±0.0 15.0±0.2 .: 2.0 scFv-HSA-K500A-scFv- FLAG ND ND ND scFv-HSA-D550N-scFv- FLAG 4.1±0.1 27.0±0.2 6.6 scFv-HSA-K573P-scFv- FLAG 6.0±0.2 0.7+0.1 0.1 HSA-K500A-scFv-FI_AG h ND v ND ND HSA-D550N-scFv-FLAG 7.3±0.1 42.0±0.3 :: 5.8 HSA-K573A-scFv-FLAG H 6.4±0.1 * 5.7±0.1 : 09 HSA-K573P-scFv-FLAG 4.7±0.1 0.7±0.1 0.1 scFv-HSA-K500A ND ND ND scFv-HSA-D550N 7.5±0.1 19.0±0.2 2.5 scFv-HSA-K573P 7.4±0.1 0.8+0.1 0.1 a: Dilutions of HSA variants were injected over immobilized shFcRn (-1500 RU). b: The kinetic rate constants were obtained using a simple first-order (1:1) bimolecular interactionmodel. The kinetic values represent the average of duplicates. c: Not determined due to weak binding (ND).

In example 8 it was shown that the K500A variant did not significantly bind shFcRn, inExample 10 it was shown that the K573P and K573A variants bind shFcRn stronger than HSA andin Example 11 it was shown that the D550N variant binds FcRn weaker than HSA.

In the present example it is shown that these observed difference in binding properties alsoare reflected in fusion polypeptides in different configurations: C-terminal fusions with a smallmoiety (HSA-FLAG), C-terminal fusions with a larger polypeptide (HSA-IL1RA); N-terminal fusionswith polypeptide (scFv-HSA); N- and C-terminal fusions (scFv-HSA-FLAG and scFv-HSA-scFv-FLAG).

Example 17. Conjugation of Horseradish peroxidase protein to Albumin and the K573Pvariant.

For conjugation analysis, commercially available recombinant albumin (Recombumin™)was used as a control molecule. For this example, a final 200mg/mL albumin K573P variant of theinvention was purified from a fed batch fermentation by means described in Material and Methods.A two step purification was carried out; 63 WO 2011/051489 PCT/EP2010/066572

The first step used a column (bed volume approximately 400mL, bed height 11cm) packedwith AlbuPureTM matrix (ProMetic). This was equilibrated with 50mM sodium acetate, pH 5.3 andloaded with neat culture supernatant, at approximately pH 5.5-6.5, to approximately 20 mg/mLmatrix. The column was then washed with approximately 5 column volumes each of 50mM sodiumacetate, pH 5.3, 50mM sodium phosphate, pH 6.0, 50mM sodium phosphate, pH 7.0 and 50mMammonium acetate, pH 8.0, respectively. Bound protein was eluted using approximately twocolumn volumes of 50mM ammonium acetate, 10mM octanoate, pH 7.0. The flow rate for the entirepurification was 154mL/min.

For the second step, the eluate from the first step was diluted approximately twofold withwater to give a conductivity of 2.5±0.5 mS/cm after adjustment to pH 5.5±0.3 with acetic acid. Thiswas loaded onto a DEAE-Sepharose Fast Flow (GE Healthcare) column (bed volumeapproximately 400mL, bed height 11cm), equilibrated with 80mM sodium acetate, 5mM octanoate,pH 5.5. Loading was approximately 30mg protein/mL matrix. The column was washed withapproximately 5 column volumes of 80mM sodium acetate, 5mM octanoate, pH 5.5. Followed byapproximately 10 column volumes of 15.7mM potassium tetraborate, pH 9.2. The bound proteinwas eluted using two column volumes of 110mM potassium tetraborate, 200mM sodium chloride,approximately pH 9.0. The flow rate was 183mL/min during the load and wash steps, and169mL/min during the elution step.

The eluate was concentrated and diafiltered against 145mM NaCI, using a Pall CentramateOmega 10,000 Nominal MWCO membrane, to give a final protein concentration of approximately200mg/mL.

Both 200mg/mL stock solutions of the rHA and K573P variant albumin were diluted down to5mg/mL, using phosphate buffer saline (PBS), pH adjusted to pH 6.5-6.7. This ensured afavourable pH environment for the maleimide reactive group of the EZ-Link® Maleimide ActivatedHorseradish Peroxidase (Thermo Scientific) to react with the free sulphydryl, to form a stablethioester bond. 2mg of the EZ-Link® Maleimide Activated Horseradish Peroxidase (HRP) wasmixed with either 1mL of the 5mg/ML rHA or K573P variant albumin. This mixture ensured anapproximate 2 fold molar excess of the albumin, or K573P variant albumin. This mixture wasminimally incubated at 4 °C, for 24 hours. The reaction mixtures were then checked forconjugation, using GP-HPLC.

To separate unconjugated species (rHA, or Albumin variant K573P and unreacted HRP)from the corresponding conjugated species the samples were first concentrated (Vivaspin20,10,000 MWCO PES, Sartorius), and then individually applied to a Tricorn Superdex™ 200, 10/300GL column (GE Healthcare), run at a flow rate of 45cm/hr in PBS. The elution peak wasfractionated and GP-HPLC analysed. Fractions containing the conjugated species were pooled,64 WO 2011/051489 PCT/EP2010/066572 concentrated and diafiltered against 50mM NaCI and analysed by GP-HPLC to demonstrate(Figure 24)

These samples were then assayed using the Biacore method described herein (Table 21).This example demonstrates that the K573P maintains its increased affinity for shFcRn comparedthe the WT HSA.

Example 18. Conjugation of Fluorescein to Albumin and the K573P variant.

The two same albumin samples used in Example 17, were also the start materials for thisexample. I.e. Approximately 200mg/mL rHA or the K573P albumin variant.

Fluorescein-5-Maleimide, Thermo Scientific (F5M) was dissolved in dimethylformamide, togive a final concentration of 25mg/mL. This was then further diluted into 18mls of PBS, pHadjusted to approximately pH 6.5. To this solution either 1ml of 200mg/mL rHA or 1mL of200mg/mL K573P variant was added. This gave an approximate 20 fold final molar excess of F5M.These samples were incubated and allowed to conjugate overnight at 4°C, in the dark, to allow themaleimide groups on the F5M to react with predominantly the free sulfhydryl, present in bothalbumin species.

Following overnight incubation aliquots of the reaction mixtures were extensively diafilteredagainst 50mM NaCI to remove unconjugated F5M, (Vivaspin20, 10,000 MW CO PES, Sartorius).Conjugation was confirmed by ultraviolet visualization of conjugated Fluorescein::AlbuminsFollowing standard SDS-PAGE (Figure 25).

These diafiltered samples were then assayed using the Biacore method described herein(Table 21). This example demonstrates that the conjugation of a small molecule to either rHA or avariant, e.g. K573P does not affect the trend in binding affinities to shFcRn.

Table 21: Representative Biacore assay KD values of conjugated rHA or a variant (K573P) whenbinding to immobilized shFcRn.

Analyte KD(pM) rHA::HRP 3 6 K573P::HRP 0.02 rHA::F5M 73 K573P::F5M 2.5 65 WO 2011/051489 PCT/EP2010/066572

Example 19. Further albumin variants.

The following variants were generated using the methods described in Example 1 E492T,N503D, E492T+N503D, K538H, E542D, D494N+E495Q+T496A, E495Q+T496A, N403K, K541Aand K541N. SPR analysis was carried out as described in Example 15 and the results presented inFigure 26 and figure 27.

Figure 30A and 30B shows the effect on shFcRn binding for the albumin variants.

Substitutions N503D, D494N+E495Q+T496A E492T+N503D, E495Q+T496A within HSA had anegative inpact on binding to shFcRn at pH5.5.

Example 20. Variants of albumin at the C-termini.

The following variants were generated using the methods described in Example 1. Bindingto the shFcRn was determined as described in Materials and Methods and the results arepresented in Table 22.

Table 22: Kinetics of the HSA C-terminal swapped variant interactions with shFcRn.

Albumin ka ; kd s KDb variant3 (103/Ms) (10¼) (μΜ) HSA 4.4+0.0 24.0+0.1 54 MacSA 3.1+0.1 8.6+0.1 2.7 HSA-MacC 4.1+0.1 5.6+0.0 1 3 MouseSAc 3.8+0.0 = 3.1+0.1 0.8 HSA-MouseC i: 3.7±0.1 : 1.3+0.0 0.3 RabbitSAd 1 9±0·.3... 1.7+0.1 0.9 HSA-RabC 3 5±0.0 1.6+0.0 0.4 SheepSA ND ND ND HSA-SheepC 3 3+0.0 i 2.1+0 0 0.6 a: Dilutions of HSA variants were injected over immobilized shFcRn (-1500 RU). b: The kinetic rate constants were obtained using a simple first-order (1:1) bimolecular interactionmodel. c: Data from Table 2 d: Data from Table 3

Not determined due to weak binding (ND) 66 WO 2011/051489 PCT/EP2010/066572

This example demonstrates that for all C-terminal swaps to human albumin tested anincrease in binding over the donor albumin was observed. All donor sequences contain the K573Psubstitution shown to significantly increase binding but less that the K573P alone (Table 20).

Example 21. Competitive binding analysis of variant albumin fusions

Competitive binding studies, using variant albumin fusions and a selection of variantalbumins prepared as described in Example 1, were performed as described in Example 4. Resultsare presented in Figures 28-31.

The competitive binding hierarchy was identical for the variants fusions of HSA-FLAG and ,N+C-terminal scFv HSA-FLAG to the hierarchy of the individual HSA variants (unfused and fused)affinity data. For the IL1 Ra variants K573P, K573A, and the K500A were as predicted, however theD550N appears to inhibit more efficiently than the WT fusion.

Example 22. Further HSA variants

The following variants were generated using methods described in Example 1: HSAE492G+K573A, HSA E492G+ N503K+ K573A, HSA E492G+ N503H + K573A, HSA E492G +K573P, HSA E492G + N503K + K573P, HSA E492G + N503H + K573P. SPR analysis wasperformed as described in Materials and Methods. Results (Figure 32) showed that all HSAvariants bound more strongly to shFcRn compared to wild type HSA at pH 5.5. No binding wasobserved at pH 7.4. HSA E492G+K573A, HSA E492G+ N503K+ K573A, unlike HSA E492G+ N503H + K573A,had marginally improved binding beyond that of HSA K573A. The combination variants containingK573P did not show improved binding over the K573P single variant. 67