AU2015205821B2 - Albumin variants - Google Patents

Abstract

The present invention relates to variants of a parent albumin having altered plasma half-life compared with the parent albumin. The present invention also relates to fusion polypeptides and conjugates comprising said variant albumin.

Description

ALBUMIN VARIANTS

The present application is a divisional application of Australian Application No. 2010311332, which is incorporated in its entirety herein by reference.

Reference to a Sequence Listing

This application contains a Sequence Listing in computer readable form, which is incorporated herein by reference.

Background of the Invention Field of the Invention

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

Any discussion of the prior art throughout the specification should in no way be considered as an admission that such prior art is widely known or forms part of common general knowledge in the field.

Description of the Related Art

Albumin is a protein naturally found in the blood plasma of mammals where it is the most abundant protein. It has important roles in maintaining the desired osmotic pressure 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 structural homology.

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. FcRn is a membrane bound protein, expressed in many cell and tissue types. FcRn has been found to salvage albumin from intracellular degradation (Roopenian D. C. and Akilesh, S. (2007), Nat. Rev. ImmunoU, 715-725.). FcRn is a bifunctional molecule that contributes to maintaining a high level of IgGs and albumin in serum in mammals such as human beings.

Whilst the FcRn-immunoglobulin (IgG) interaction has been characterized in the prior art, the FcRn-albumin interaction is less well characterized. The major FcRn binding site is localized within Dill (381-585). Andersen et a/(2010). Clinical Biochemistry 43,367-372. Data indicates that IgG and albumin bind non-cooperatively to 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 human FcRn appears to be more discriminating (Ober etal. (2001) Int. Immunol A3, 1551-1559). Andersen etal. (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 at physiological pH to either receptor. At acidic pH, a 100-fold difference In binding affinity was observed, in aii cases, binding of aibumin and igG from either species to both receptors were additive.

Human serum albumin (HSA) has been well characterized as a polypeptide of 585 amino acids, the segaence of which can be found in Peters, 7\, Jr. pBB$) Ataboui Aibumin; Bioe&emMry, Oett&tite'md Medical, Appilcaiiom ppiU Academic Press, !ne.s Orlando (ISBN 0-12-662110-3), it has a characteristic binding to its receptor FcRsv where it btndeiat 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 variant having lower plasma half-life has::'been identified (Peach, R; J, and Brennan, S. 0.,(1991) Biocbim Biopbys Aela,109T;49~54) having the substitution 0494N. This substitution generated art, N~ giycosylation site In this variant, which Is not present In the wild-type albumin, it Is not known whether the gfyeosyiafion or the amino acid change is responsible for the change in plasma half-life.

Albumin has a tong plasma half-iife and because of this property it has been suggested for use in drug delivery. Albumin has been conjugated to pharmaceutically beneficial compounds (WO 2000/699Q2A), and It was found that the conjugate - maintained tie tong plasma half-life of aibumin. The resulting pissma hatMife of the conjugate was generally considerably longer than the plasma half-fife of the beneficial therapeutic compound atone.

Further, albumin has been fused to iherapeuticaiiy beneficial peptides (W© 2001/79271 A and WO 2003/59934 A) with the typical result that the fusion has the activity of the therapeuicaily beneficial peptide and a considerably longer plasma haif-iife than the plasma haif-iife of the therapeutically beneficial peptides alone.

Otagirl ef a/|2069), Biol Hharm, Sufi. 32(4), 527-534, discloses that 77 aibumin variant are know, of these 25 are found in domain 111. A natural variant lacking the: last 175 amine adds at the oartroxy termini has been shown to have reduced- half-life (Andersen ei af (2016), Clinical Biochemistry 43, 367-372). iweo ef a(.(2Q07) studied: the half-life of naturally securing human aibumin variants using a mouse model, and found that K541E and K500E had reduced half-life, £501K and E570K had increased half-life arid K573E had almost no effect on half-life; (IwaOi et ai. (2007) B.B.A. proteins and Proteomies 1774, :1582-1560),

Galliano et U (1993} Slochim, 8iophys.Acta 1225. 27-32 discloses a natural variant E60SK, Minchrotfi ef a/.(1990) discloses a natural variant K536E, Minchiotti ef a/ (1987) Biddhim, Biophys. Acta 916, 411-418 discieses a natural variant K574N. Takahashl ei a/(1687} Proc. Mali. Acad. Sci, USA 84, 4413-4417, discloses a natural variant D550G. Carlson ef a/ (1962),

PmcWai.Acad,ScivUSA 89,8225- 8229, discloses a natural variant D550A.

Albumin haS the ability to bind a number of ligands and these become associated (associates) with albumih. This property has been utilized to extend the plasma half-life of drugs having the ability to noncovalently bind to albumin. This can also be achieved by binding a pharmaceutical beneficial compound, which has little or no albumin binding properties, to a moiety 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 which such a preparation maybe for example, but not limited to, a nano particle or micro particle of albumin. In these examples the delivery of a pharmaceutically beneficial compound or mixture of compounds may benefit from alteration in the albumins affinity to its receptor where the beneficial compound has been shown to associate with albumin for the means of delivery.

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

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

It may even be desirable to have variants of albumin that have little or no binding to FcRn in order to provide shorter half-lives or controlled serum pharmacokinetics as described by Kenanova etal(2009) J. Nucl. Med.·, 50 (Supplement 2):1582).

Summary of the Invention

According to a first aspect, the present invention provides a method for preparing a variant of albumin, a fragment thereof or a fusion polypeptide comprising said variant albumin or fragment thereof, comprising following steps: a. providing a nucleic acid encoding a parent albumin having at least 90% sequence identity to SEQ ID NO: 2; b. modifying the sequence of step a., to encode said variant of albumin, fragment thereof or fusion polypeptide comprising said variant albumin or fragment thereof having a substitution corresponding to a substitution in SEQ ID NO: 2 selected among: K500A,C,D,E,F,G,H,L,M,N, Q,S,T,V,W,Y; c. introducing the modified sequence of step b., in a suitable host cell; d. growing the cells in a suitable growth medium under conditions leading to expression of said variant of albumin, fragment thereof or fusion polypeptide comprising said variant albumin or fragment thereof; and e. recovering said variant of albumin, fragment thereof or fusion polypeptide comprising said variant albumin or fragment thereof from the growth medium; wherein said variant of albumin, fragment thereof or fusion polypeptide comprising said variant albumin or fragment thereof, has a shorter plasma half-life compared with the parent albumin, fragment thereof or fusion polypeptide comprising said variant albumin or fragment thereof.

According to a second aspect, the present invention provides an isolated variant of albumin, a fragment thereof or a fusion polypeptide comprising said variant albumin or a fragment thereof having a shorter plasma half-life compared with the parent albumin, fragment thereof or fusion polypeptide comprising said parent albumin or fragment thereof, comprising at least 90% identity to SEQ ID NO: 2 and a substitution at a position corresponding to position 500 in SEQ ID NO:2, where the variant is not the variant consisting of SEQ ID NO: 2 with the substitution K500I or K500R.

According to a third aspect, the present invention provides an isolated nucleic acid encoding the variant of albumin, fragment thereof or fusion polypeptide comprising said variant albumin or fragment thereof according to the second aspect.

According to a fourth aspect, the present invention provides a conjugate comprising: an isolated variant of albumin or fragment thereof having a shorter plasma half-life compared with the parent albumin, fragment thereof or fusion polypeptide comprising said parent albumin or fragment thereof, comprising a substitution at a position corresponding to position 500 in SEQ ID NO:2; and a beneficial therapeutic moiety; where the isolated variant of albumin is not the variant consisting of SEQ ID NO: 2 with the substitution K500I or K500R.

According to a fifth aspect, the present invention provides an associate comprising: an isolated variant of albumin or fragment thereof having a shorter plasma half-life compared with the parent albumin, fragment thereof or fusion polypeptide comprising said parent albumin or fragment thereof, comprising a substitution at a position corresponding to position 500 in SEQ ID NO:2; and a beneficial therapeutic moiety; where the isolated variant of albumin is not the variant consisting of SEQ ID NO: 2 with the substitution K500I or K500R.

According to a sixth aspect, the present invention provides a fusion polypeptide comprising: an isolated variant of albumin or fragment thereof having a shorter plasma half-life compared with the parent albumin, fragment thereof or fusion polypeptide comprising said parent albumin or fragment thereof, comprising a substitution at a position corresponding to position 500 in SEQ ID NO:2; and a fusion partner polypeptide; where the isolated variant of albumin is not the variant consisting of SEQ ID NO: 2 with the substitution K500I or K500R.

According to a seventh aspect, the present invention provides a composition comprising: an isolated variant of albumin or fragment thereof having a shorter plasma half-life compared with the parent albumin, fragment thereof or fusion polypeptide comprising said parent albumin or fragment thereof, comprising a substitution at a position corresponding to position 500 in SEQ ID NO:2; or a conjugate comprising said variant albumin, wherein the binding to the neonatal Fc receptor (FcRn) is weaker than for the corresponding albumin or fragment thereof or fusion polypeptide comprising said albumin or fragment thereof or a conjugate comprising said albumin; where the isolated variant of albumin is not the variant consisting of SEQ ID NO: 2 with the substitution K500I or K500R.

According to an eighth aspect, the present invention provides an isolated variant of albumin, a fragment thereof or a fusion polypeptide comprising said variant of albumin or said fragment thereof when prepared by the method according to the first aspect.

According to a ninth aspect, the present invention provides use of an isolated variant of albumin or fragment thereof having a shorter plasma half-life compared with the parent albumin, fragment thereof or fusion polypeptide comprising said parent albumin or fragment thereof, comprising a substitution at a position corresponding to position 500 in SEQ ID NO:2, to increase the half-life of a beneficial therapeutic moiety; where the isolated variant of albumin is not the variant consisting of SEQ ID NO: 2 with the substitution K500I or K500R.

Unless the context clearly requires otherwise, throughout the description and the claims, the words “comprise”, “comprising”, and the like are to be construed in an inclusive sense as opposed to an exclusive or exhaustive sense; that is to say, in the sense of “including, but not limited to”.

The present invention provides variants of a parent albumin with improved properties compared to its parent. In particular the invention provides variants of a parent albumin 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 parent albumin, comprising an 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 of the mature polypeptide of SEQ ID --y NO: 2, wherein the variant is not the variant consisting of SECt ID NO: 2 with the substltotion 04841% E501K, K54.1E, 05500,4: K573E or 14574N,

The alteration at one or mom 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; nucleic acid constructs, vectors, and host cells comprising the polynucleotides; and methods of producing the variants,

The present invention also relates to conjugates or associates comprising the variant albumin or fragment thereof according to the invention and al beneficial therapeutic moiety or to a fusion polypeptide comprising a variant albumin or fragment thereof of the: inven tion and a fusion partner polypeptide.

The invention further relates: to compositions comprising the variant albumin, fragment thereof, fusion polypeptide comprising variant albumin or fragment thereof or conjugates Comprising the variant albumin or fragment thereof, according to the Invention or associates comprising the variant albumin or fragment thereof, according to the invention. The compositions a re preferably pn amiaceulioai com positions.

The invention further relates to a pharmaceutical composition comprising a variant albumin, fragment thereof, fusion polypeptide comprising variant albumin or fragment thereof or conjugates comprising toe variant albumin or fragment thereof; or associates comprising the variant albumin or fragment thereof, wherein said variant albumin, fragment thereof, fusion polypeptide comprising variant album In or fragment thereof or conjugates Comprising the variant albumin or fragment or associates of variant albumin or fragment thereof has altered plasma half-life compared to the corresponding plasma half-life of the HS.4 or fragment thereof, fusion polypeptide comprising HSA or fragment thereof or conjugates or associates of HSA or, fragment thereof, comprising HSA or fragment 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 pD82305

Figure 3 shows a restriction map of the expression plasmid pDB4005

Figure 4 shows SPR sensorgrams 10 μΜ albumin injected over shFcRn HSA (JTA) ~ fatty add free HSA obtained from Sigma-Aidrich £43782)* HSA (Novozymes) ~ Oommepai Recombinant human serum albumin (RECOblBUMiN).

Figure 5 shows ELISA bidding of shFcRn-DST to human serum albumin pSA) variants (100-0.04S ng/mi). Binding of WT, B494N, D494Q and D494A pH 6,0 arid pH 7.4, Binding of WT, 0494N, 0494N/T498A and T498A at pH 6.0 and pH 7.4. Binding of WT5 E495Q and E495A at pH 6.0 and pH 7.4.

Figure € shows representative sen so rg rams of binding of 0.2 μΜ of HSA variants !o immobilized ShFcRn (-4699 RU), WT, D494N, 04940, D494A, D494H/T496A and T496A.

Figure 7 shows representative sensorgrams of binding of 1 uM of HSA variants to imnibfeiiized shFcRn (-1406 RU), WT, D494N, D494Q, ©494A, D494N/T498A and T498A.

Figure 8 shews relative binding ot the HSA variants compared to WT based on two independent SPR experiments as shown (A) Figure 6 and (8) Figure 7.

Figure 9 shows ELISA: (A) binding of shFcRn to albumins' from human, donkey, bovine, sheep, goat and rabbit at pH 8.0. (B) binding of shFcRn to albumin from guinea pig, hamster, rat and chicken at pH 8.0. (G| binding of shFcRn to albumin from human, donkey, bovine, sheep, goat and rabbit at pH 7.4, (D) binding of shFcRn to albumin from guinea pig, hamster, rat and chicken at pH 7.4. (E) relative binding of the different albumins. Relative binding of human albumin to shFcRn is defined as 1.0. The ELISA vaiues represent the mean of duplicates.

Figure 18 shows SPR: Binding ot shFcRn-GST to ai bum in from several species at pH 8.0 and pH 7,4. Representative sensorgrams showing binding of 5.0 μΜ of albumin from different species; (A) human, (B) donkey, {€} bovine, (D) goat, {£) sheep, (F) rabbit, p) dog, (H) guinea pig, (I) : hamster, (J) rat, fK) mouse and (L). chicken. The albumin variants were injected over immobilized GST-tagged shFcRn (-2100 RU), infections were performed at 26*0 at a rate of 49 μΙ/min.

Figure 11 shows SPR sensorgrams of selected HSA mutants compared with wild-type HSA. 20 μΜ of (A) WT and P499A (B) WT and K500A, (C) WT and K536A, (0) WT and P537A and |E) WT and K838A and (F).WT and KS37Awere injected over immobilized ShFcRn at pH 6.9 (-1600 RU)

Figure 12 shows SPR sensorgrams of HSA mutants compared with WT HSA, 10 pM of (A) WT and KS73A (8) WT and K573G, (C) WT and K573F, (D) WT and R6736 and (E) WT and K573L and (F) WT and R873M, (G) WT and K673G, (H) WT and &S73R and (!) WT and K573T and (j) WT and KB73V injected over immobilized ShFcRn at pH 5.5 and pH7.4. Injections were performed at 25*C at a flow rate of 89 pl/min.

Figure 13 shows SPR sensorgrams of HSA mutants compared with wild-type HSA. 10 pM Of |A) WT and K673D (8) WT and K673E, (C) WT and KS73H, (DJ WT and K573! and (E) WT and R573lsi and (F) WT and KS73P, (G) WT and K573S, |H| WT and KS73* and (1) WT and K573W and (J) WT and K573Y injected over immobilized shFcRn at pH 5.5 and pH7.4. injections were performed at 2S*Cat a flow rate of 80 μί/min.

Figure 14 shows SPR sensorgrams of HSA mutants compared with wild-type HSA. 20 plVI of (A) WT and E4B2G+K:638H^R541N^E642D (8) WT and E482T+N503K+K641A! (C) WT and E4 92P + M503k+K541G + E542P, (D) WT and E492H*£501P+N503Η + £505D +TS06S+T64OS+K541E and {£) WT and A490D+E492T+V493L+E501P+N503D+A504E *E50SK+T6Q8F+K541D and (F) WT and I4i2G+\M93P+K538H+K541N+E542D inJeci6d over immobilized shFcRn at pH 5.0. injections war# performed at 28"C at a flow rate of 80 μΙ/tdin»

Figure 15 shows SPR sensorgrams of HSA mutants compared With wild-type·HSA. Twenty plVf of f A) WT, (6) H440Q. (C) H464Q and (0) H535Q injected over immobilized shFcRn at pH 8.0. Injections ware performed at 2S°C at a flow rate of 80 μί/rhin.

Figure 16 shows SPR sensorgrams of HSA mutant K50OE compared with wild-type HSA. Ten μΜ of HSA mutant,.ΚδΟΟΕ injected over immobilized ShFcRn at pH 5,75. injections were performed at 25¾ at a flow rate of 30 pl/min.

Figure 17' shows a restriction map of the expression plasmid pDB3017 Figure 18 shows atrostrictlon map of the expression plasmid pDB302i Figure 18 shows a restriction map of the expression plasmid pD83056 Figure 20 shews a restriction map of the expression plasmid pDB3165 Figure 21 shows a restricted map of the expression plasmid pSB4172 Figure 22 shows a restriction map of the expression plasmid pDB4267 Figure 23 shows a restriction map of the expression piasmid pDB42S®

Figure 24 shows a GP-NPLC chromatogram of WT HSA and mutant i<573P HRP conjugates for shFcRn analysis. injections of 25pL were made onto a TSK G3000SWXL column (Tosoh Bioscience) as described in materials and methods.

Figure 25 shows SOS PAGE separation followed by both visual (A) and ultraviolet (B) detection of the RuoFesesio conjugated albumin. HSAcFoM (Lane 1), K573P::F5IV1 (Lane 2) and rHA standard (Lane 3).

Figure 26 shows shFcRn binding properties of HSA variants. IQpIVi of VVT rHA and E482T(A}, WT rH A and D494N/E48SQ/T4i6A(B)! WT rHA and N503D{C), WT rHA and N503K(D), WT rHA and E49.2T/N503D(E), WT rHA and E495Q/T496A(F), WT rHA and K538HCG), VVT rHA and E492D{H) injected over immobiiised shFcRn at pHS.S

Figure 27 shows shFoRh binding properties of HSA variants. 10μΜ of VVT rHA and i<541 A(i) and WT rHA and K54TNfJ) were injected over immobiiised shFcRn at pH5.5.

Figure 28 shows competitive binding of K573A and K573P measured by injecting shFcRn (100 niVi} alone or pre-incubated with different amounts of HSA K573A and K873F over immobilized HSA (42500 RU) at pH6.8

Figure 29 shows competitive binding of HSA-FLAG variants measured by injecting shftoRn (:100 hM) alone or together with different amounts of HSA-FLA6 variants over immpbiifzed HSA (-2500 RD) at pH6.0.

Figure 30 shews competitive binding of HSA-illRa variants measured by injecting shFcRn (100 nM) alone or together with different amounts of HSA-1L1 Ra variants over immobilized HSA (-2500 RU) at pH6.0

Figure 31 shows competitive binding of scFv~fused RSA variants measured by injecting shFcRn (100 nM) alone or together with different amounts: of (A) scFv-BSA-FLAG variants: or (B) RBA-soFv-FLAG variants over immobilized HSA (-2500 RU) at pH8.0.,

Figure 32 shows binding of HSA, single, double and triple mutant variants to shFoRm Samples bf 10 μΜ of each HSA variant were injected oyer immobilized shFcRn at pH 5.5 or pH 2,4.

Detailed Description of the Invention

The: present invention: relates to isolated variants of albumin or fragments thereof, or fusion polypeptides comprising variant: albumlb or fragments thereof, of a parent albumin, comprising ah alteration at. one Or more {several} positions eorrespohdlng to positions 417, 440, 464» 400:.: 482, 483, 494, 495, 496,: 499, 500, 501, 59¾ 504, 505, 506, 519, 53¾ 536, 537, 538, 540, 541, 542, 550, 573, 574, 575, 577, 578, 579, 580, 581,582 and 584 of the mature polypeptide of SEO ID HO; .2, wherein the variant is not toe variant consisting of SBQ iD NO: 2 with the substitution D494N:, E501K, KS4.1E, D530G A; #73E or R574M.

The alteration at one or mom 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 or more aiteraiion{s}s Te,, a substitution, insertion, and/or deietion, at one or more (several) positions, A substitution means a: replacement of an amino acid occupying a position with a different amino acid; a deletion means removal: of an amino acid occupying a position; and an Insertion means adding 1 or more, preferably 14$ amiho acids immedsateiy adjacent to an amino add occupying a position.

Mutant: The term “mutant" means a polynucleotide encoding a variant,

Wiid'Type Albumin: The term “wiidAype” (WT) albumin means albumin haying the same amino add 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 to which an alteration is made by the hand of man to produce the albumin variants of the present invention. The parent may be a naturally occurring (wlid-type) polypeptide or an aiieie thereof, or even a variant thereof,

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

Isolated variant: The term “isolated varianf means a vananithat is niodifted by the hand of man and separated completely" or partiaiiy from at least one component with which it naturally occurs, in one aspect, the variant is at least 1% pure, e.g,, at least 5% pure, at least 10% pure, at least 20% pure, at least 40% pure, at least 60% pure, at least 80% pure, and at least 90% pure, as determined by SDS-P&amp;SE or GP-HPLG

Substantially pure variant: The term “substantially pure variant" means a preparation that contains at most 10%, at most 8%, at most 0%, at most 8%, atrnost 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 or reoombinantly associated. Preferably, the varianf is at ieast 92% pure, e g,, at least 94% pure, at Ieasi 95% pure, at least 98% pure, at ieast 97% pure, at least 08% pure, at Ieasi 09%, at least 99,5% pure, and 100% pure by weight of the iota! polypeptide material present in the preparation. The variants of the present invention are preferably ima substantiaiiy pure form, This can be accomplished, for example, by preparing the variant by well known recombinant methods and by purification:^ methods,

Mature polypeptide: The term “mature poly peptide’' means a polypeptide in its final form following translation and any post-translational modifications, such as N-tormina! processing, G-lerminai truncation, giycosylatlon, phosphorylation, etc. In one aspect, the mature polypeptide is amino acids i to 585 of SEQ ID NO: 2, with the inclusion of any posf-transiationai modifications,

Mature polypeptide coding sequence: The term “mature polypeptide coding sequence” means a polynucleotide that encodes a mature albumin polypeptide. In one aspect, the mature poiypoptide coding: sequence is nucleotides 1 to 1:758 ofSEQ iD NO: 1,:

Sequence Identity: The relaledness between two amino acid sequences or between two nucleotide sequences is described by the parameter isseqyence ideniiiyT

For purposes of the present invention, the degree of sequence identity between two amino acid sequences is determined using the Needleman-Wunsoh algorithm {Needlernan and Wunsch, 1970, J. M&amp;L Biol. 48: 443-453) as Implemented in the Needle program of the EMBGSS package (EMBOSS: The European Molecular Biology Open Software Suite, Rise ef a/., 2000, Trends Gengt. 16:278-277¾ preferably version 3.0.0 or later, The optional parameters used are gap open penalty of IP, gap extension penalty of 0,5, and the EBLOSUMBS (EMBOSS version of BLOSUM62) subsidy lion matrix. The output erf Needle labelled longest identity" (obtained using the Hiobrief option) is used as the peroent Identity and is calculated as follows: (Identical Residues x 100}/(Lehgth of Alignment - Total Number of Gaps ini Alignment)

For purposes of the present invention, the degree of sequence identity between two deoxyribonucleotide sequences is determined using the Needieman-Wunseh algorithm (Needieman and VVunsch, 1870, sopra) as Implemented In the Needle program of the EMBGSS package: (EMBOSS: The European Molecular Biology Open Software Suite. Bice et a/., 2000, sopne)*: 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 NUQ4,4) substitution matrix. The output of Needle labeled longest identity" (obtained using the Hiobrief option) is used as the peroent identity end is calculated as'follows: (Idenilcd! BeoxyribonucleQtides .x 100)/(Length of Alignment - Total Number of in Alignment)

Fragment: The term: '‘fragment” means a polypeptide having one or more (several) amino aeids deleted from the amino a n d lo rear bo xy i te rmi n u s of an aibomln and/or an internal region of aibumin that has retained the ability to'bind to FoRn. Fragments: may consist of one uninterrupted sequence .derived from H8A or it may comprise two or more sequences derived from HSA. The fragments: according to the invention have a size: Of mere than approximately 20 amino acid residues, preferably more than 30 amine acid residues, more preferred more than 40 amino acid residues, more preferred more than 50 amino acid residues, more preferred more than 76 amino acid residues, more preferred more than 100 amino acid residues, more preferred more than 200 amino acid residues, more preferred more than 300 amino acid residues, even more preferred more than 400 amino add residues and most preferred more than 500 amino acid residues.

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

Coding sequencer The term "coding sequence" means:a polynucleotide, which direedy specifies the amino acid sequence of its translated polypeptide product. The boundaries of the coding sequence are generally determined by an open reading frame, which usually begins with the ATS start codon or alternative start codons such as GT© and TTG and ends with a stop codon such as TAA, TAG, and TGA. The coding sequence may be a DNA, cONA, synthetle, or recombinant polynucleotide, cDNA: The term foONA* means a DMA molecule that can be prepared by reverse tranaaription from a mature, spliced, mRNA molecule obtained from a eukaryotic ceil, oDNA lacks ibtron sequences that may be present in the corresponding genomic DMA. The initial, primary RNA transcript is a precursor to mRNA that is processed threcg h a senes of steps, including splicing, before appearing: as mature spliced mRNA,

Nucleic acid construct; The term ^nucleic acid construe?' means a nucleic acid molecule, either single- or : double-stranded. which is Isolated from a naturally occurring gene or is modified to contain segments of nucleic acids in a manner that would not otherwise exist in nature or which is synthehc. The term nucleic acid construct is synonymous with the term "expression cassette" when the nucleic acid construct contains the control sequences required for expression of a coding sequence erf the present invention,

Control sequences; The term "control sequences" means all components necessary for the expression of a polynucleotide encoding a variant of the present invention. Each control sequence may be native or foreign to the polynucleotide encoding the variant or native or foreign to each other. Such control sequences include, but are not limited to, a leader, pdiyadenyiation sequence, propeptide sequence, promoter, signal peptide sequence, and transcription terminator. At a minimum, the control sequences include a promoter, and transcriptional and translational stop signals. The control sequences may be provided with linkers for the purpose of introducing specific restriction sites facilitating ligation of the control sequences within the coding region of the polynucleotide encoding a variant.

Operabty Jinked: The term "operably finked" means a configuration in which a control sequence is placed at-an appropriate position relative to the coding sequence of a polynucleotide such that the control sequence directs the expression of the coding sequence.

Expression; The term "expression" includes any step involved in the production of the variant including,: but not iimited to, transcription, post-transcriptional modification, translation, post-translational modlficatfo^ secretion.

Expression vector; The term "expression vector' means a linear or circular DNA molecule that comprises a polynucleotide encoding a variant and is operably linked to additional nucleotides that provide for its: expression.

Host cell:: The term "host cell* moans any cell typo that is susceptible to transformation, ftansfiscfioh, transduction, end the like with a nucleic acid construct or expression vector comprising a; polynucleotide Of the present invention. The term: “host cell" encompasses: any progeny of a parent ceil that is: notidentical to the parent cell due to. mutations that Occur during replication.

Plasma halTfife: PSasma half-life is ideally determined Ip wvo In suitable individuals. However, since | is time consuming and expensive and there inevttahie are ethical concerns connected with doing experiments in animals dr man it is desirable to use an in vitro assay for determining whether plasma half-life is extended dr reduced* It Is known that the binding of albumin to its receptor FcRn Is important for plasma half-life and the correlation between receptor binding and 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 an Increased plasma half-life and a lower affinity of albumin to its receptor is considered Indicative of a reduced plasma half-life.

In this application and claims the binding of albumin to its receptor FcRn Is described using the term affinity and the expressions “stronger'* or “weaker'. Thus, it should be understood that a molecule having: a higher affinity to FcRn than HSA Is considered to blnd stronger to f pRn than HSA and a molecule having a lower affinity to FcRn than HSA is considered to bind weaker to FcRn than HSA,

The terms longer plasma half-life’* or “shorter plasma lialf-liW': and; similar expressions are Understood to be in relationship to the corresponding pafenfialbumln molecule..Thus, alonger plasma halhlifeyvitn respect to a variant albumin of the invention means:;that the variant has longer plasma balf-lifethan the; corresponding aibomin having the; same sequences except for the a iteration (s) in pos;hon$ corresponding :fe 417} 440, 464, 490, 49:2, 49¾ 494,496, 496, 499. S00, 591, 503, 504, §05, 506, 510, 533, 539, 537, 533, 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, tire mature polypeptide disclosed in SEQ !0 NO: 2 is used to determine the corresponding amino acid residue in another aibumin. The amino add sequence of another albumin is aligned with the mature polypeptide disciosed in SEQ iD NO: 2, and based on the alignment, the amino acid position number corresponding to any amino acid residue in the mature polypeptide disclosed in SEQ ID NO: 2 is determined using the Needieman-Wunseh algorithm (Needfeman and Wunseh, 1978, J. Mol Bioi 48: 443-453) as implemented in the Needie program of the EMBOSS package (EMBOSS: The European Molecular Biology Open Software Suite, Rice ef ah, 2000, Trends Genet 1S: 270-277), preferably version 3,0.0 or later, identification of the corresponding amino add residue In another albumin can be confirmed by an alignment of multiple polypeptide sequences using XlustalW" (Larkin el at, 2007, BioinformatiGS 23:2947-2948).

When the other polypeptide (pr protein):has diverged from the mature polypeptide of SEQ ID NO: 2 such that traditional sequence-based comparison fails to detect their relationship (Lindahi arid Elofsson, 2009, l itt Slot 295: 613-515), other pairwise sequence comparison algorithms can be used. Greater sensitivity in sequence-based searching ban he attained using search programs that utilize probabilistic representations of polypeptide families (profiles) to search databases, For example, the; PSPBLAST program generates profiles through an iieraflve database search process: and is capable of defecting remote homologs (AtSchul ei &amp;LS 1997, Nuclei® Acids Res, 25: 3389--94(52), Even greater sensitivity can be achieved If the: family or superfamily for the polypeptide has one or more representatives in the protein structure databases. Programs such as GenTHREADER (Jones, 199¾ J, Mo!. Βίοι 287: 797-815; MeGuffin and Jones,; 2003, Bieinfomistics It; 874-881:) utilize information from a variety of sources {PSi-BLAST, secondary structure prediction, structural alignment profiles, and solvation potentials) as inputs to a neural network that: predicts the structural fold for a query sequence. Similarly the method of Gough etat, 2000:, J. Mel. Βίοι 318: 908-819, can be used to align a sequence of unknown structure within the supeffamsly models present in the SCOP database. These alignments can in turn be used to generate homology modelslfor the polypeptide, and such models can be assessed for accuracy Using a variety of tdois doveipped for that purpose.

For proteins of khown structure, several tools and resources are available for retrieving and generating structural alignments. For example the SCOP superfamillee of proteins have been structurally aligned, and those alignments are accessible and downloadable. Two or more protein structures dan be aligned using a variety of algorithms such as the distance alignment matrix (Heim and Sander, 1998, Proteins 33; 88-98} or combinatorial extension (Shlndyaiov and Bourne, 1898, PiWein Engineering 11; 739-747), and Implementations of these algorithms can addifionaiiy be utilized to guery structure databases with a structure of interest in order to discover possible structural hemofegs Heim and Park, 2000, Bfomfermadcs 18; 566-587).

In describing the albumin variants of the present invention, the nomenclature described beiow is adapted for ease of reference. The accepted 1UPAC single letter or three letter amino add abbreviation is employed.

Substitutions. For an amino acid substitution, the following nomenclature is used: Original amino acid, position. substituted amino acid. Accordingly, for example the substitution of threonine with alanine at position 226 is designated as iThr226Ala'! or “T228AT Multiple mutations are separated by addition marks (“**), e.g., :iGly205Arg + Ser41 IPhe" 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., KEA92T/HS03DBths$ shoyid be viewed as interchangeable with (***),

Deletions. For an amino add deletion, the foliowing nomenclature is used; Original amino acid, position*. Accordingly, the deletion of glycine at position 195 is designated as “GiylSS*" or “0195*". Multiple deletions are separated by addition marks (“A!); “Gly10§* * Ser411*" or “0195* + $411**,

insertions, For an amino add insertion, the following nomenclature is used: Original amino add, position:, original amino add, inserted amino acid. Accordingly the insertion of lysine after glycine at position 1 §5 is designated 'OlyiSSGiyLys" or τβΙδδΟΚ”. An insertion eft multiple amino acids is designated[Original amino acid, position, original. amino acid, inserted amino acid #1: inserted amino acid #2; etc.]. For example, Itie insertion of lysine and alanine after glycine at position 1951s indicated as KG1y195GlyLpAta” er*<G195GHaA*K in such cases the inserted amino, acid residue(s) are numbered by the addition of lower case letters to the position number of the amino acid residue preceding the inserted amino acid residue(s). in the above example, the sequence would thus be:

Multiple alterations. Variants comprising multiple alterations are separated by addition marks C*”)> 0.0., Arg170Tyrr-Gly 195ΘΙιΓ or ‘R170Y*G195E:: representing a substitution of tyrosine and glutamic acid for arginine and giycine at positions 170 and 196, respectively.

Different substitutions:. Where different substitutions can he introduced at a position, the different substitutions at® separated By a comma,-&amp;>&amp;., ifAtg170TyrsGlLf! represents a substitution of arginine with tyrosine or glutamic acid at position 170. Thus, “Tyrl 6'7SlysAia + Arg170Gly,Aia” designates ibeftbllowing variants! <!Tyr 187Gly-rArg 170βiy , “Tyrl67Gly+Argi7ΟΑ1;00: “Tyrl67AiafoArg1?0Gly\ and “Tyrl67Aia+Arg17QAIa",

Parent albumin

Albumins are proteins and constitute the most abundant protein in plasma in mammals and aibuminsfrom a long number of mammals have: been charaptenzed by biochemical methods and/or by sequence information. Several albumins, e,g„ human serum: albumin; fHSA), have also been characterized erystallagraphleally and the structure determined. HSA is a preferred albumin according to the invention and Is a protein consisting of 885 amino 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 ND: 2, The skilled person will appreciate that natural aiielos may exist having essentially the same properties as HSA but having one or more amino acid changes compared to SEQ ID MO: 2, and tee inventors also contemplate the: use of such natural alleles as parent albumin according to the invention.

Albumins have generally a Jong plasma half-iife of approximately 20 days or longer, e.g., HSA has a plasma half-life of 19. days. if Is known that the long plasma haf-iffe of HSA is mediated via interaction with its receptor FcRn, however, an understanding or knowledge of the exact mechanism behind the tang half-life of HSA Is net essential for the present invention.

According to the invention the term *albumirf means a protein having the same, or very simiiar three dimensional structure as HSA and having a long plasma half-life. As examples of albumin proteins according to the invention can be mentioned human serum -albumin, primate serum albumin, (such as ehimpansee serum albumin, gorilla serum albumin), rodent serum albumin (such as hamster serum albumin, guinea pig serum albumin, mouse albumin and rat serum albumin),, beylhe serum: albumin, equine serum albumin, donkey serum albumin, rabbit: serum albumin, goatee rum albumin, sheep serum-albumin, dog serum albumin, chicken: serum albumin and.pig' serum albumin. HSA as disclosed in 8EQ ID NO-t 2 or any naturally Occurring altalelbereof, is the preferred albumin according to the invention.

The parent albumin, a fragment thereof, or albumin part of a fusion polypeptide comprising albumin or a fragment thereof according to the invention has generally a sequence identity to the sequence of HSA shown In SEO ID NO; 2 of at least 60%, preferably a* least.70%, preferably at feast @0%, preferably at least 85%, preferably at least 86%, preferably at least 87%, preferably at feast 88%, preferably at least 89%, preferably at least 90%, preferabiy at least 91%, preferably at least 92%, -preferably.at least 93%, preferably at least 94%, preferably at feast 95%, more preferred at least 96%, more preferred at least 97%, more preferred at least 98% and most preferred at feast 99%.:

The parent preferably comprises or consists of the amino acid sequence of iiQ 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 I'D

NOtZ in a second aspect, the parent is encoded: by a polynucleotide that hybridizes under very low stringency conditions, low stringency conditions, medium stringency conditions,/medium-high stringency conditions, high stringency conditions, or very' high stringency conditions with (!) the mature polypeptide coding sequence of SEQ ID NO; 1, (»} the mature polypeptide coding sequence of SEQ 10 NO; 1, or p) the full-length complementary strand of (i) or (IS) (J. Sambrook, E.F. Fritsch, and T. fdaniafis, 198¾ Molecular Glorilhg, 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 acid sequence of $EQ ID HQ; 2 or e fragment thereof, may be used to design nucleic acid probes to identify and clone DN.A encoding a parent from strains of different genera or species according to methods well known in the art, in particular, such probes can be used for hybridization with the genomic or cDNA of the genus or species of interest, following standard Southern blotting procedures, in order to identify and isolate the corresponding gene therein. Such probes can be considerably shorter than the entire sequence, but should he at least 14, e,g., at least 25» at least 35, or at least 70 nucleotides in length. Rreferahiy, the nucleic acid probe is at least 100 nucleotides in length, e:g.:, at least 200 nucleotides, at least 300 nucleotides, at least 400 nucleotides, at least 500 nucleotides, at least 500 nucleotides, at least 700 nucleotides, at feast 800 nucleotides, or at least 000 nucleotides in length. Both DMA and RNA probes can be used. The probes are typically labelled for detecting the corresponding gene (for example, with ^-/¾. ^S, biotin, otavidfn). Such probes are encompassed by the presebt invention. A genorhtc DNA oreDNA library prepared from such other organisms may be screened for DNA that hybridizes with the probes described above and encodes a parent. Genomic or Other DNA from such other orgahisms may be separated by agarose or pofyaefyiamsde gei electrophoresis, or other separation techniques. DNA from the libraries or the separated DNA may be transferred to and immobilized on nitrocellulose or other suitable carrier material, in order to identify a clone or DMA that is homologous with SEQ ID N©: 1 or a subsequence thereof, the carrier material is used in a Southern blot.

For purposes of the present invention, hybridization indicates that the polynucleotide hybridizes to a labelled nucleotide probe corresponding to the polynecieoide shown In SEQ ID NO; 1, its complementary strand, or a subsequence thereof, under low to Very high stdngeney conditions, Molecules to which the probe hybridizes can be detected using, for example, X-ray film or any ether detection means known in the art.

In one aspect, the nucleic acid probe is the mature polypeptide coding sequence of SEQ ID NO: 1, in another aspect, the nucleic acid probe is nucleotides 1 to 1785 of SEQ ID NO; 1. in another aspect, the nucleic acid probe is a poiynucieotide that encodes the polypeptide of SEQ ID NO: 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 stringency conditions are defined as prehybridization and hybridization at 42¾ in 5X 8SPE, 0-,3% SDS, 200 mlcrograms/ml sheered and denatured salmon sperm DNA, and either 25% formamide for very Sow and low stringencies* 35% formamide for medlMm and mediumphigh stringencies, or 50% formamide for high and very high stringencies, foilotying Standard Southern blotting procedures for 12 to 24 hours optimally, The carrier materiel is finally washed three times each for 15 minutes using 2X SSC, 0.2% SDS at 45CC (very low stringency), 50:'C (low stringency), S5°C (medium stringency), 60*C (medtMm-hiib stringency), 65°0 (high stringency), or70°C (very high stringency).

For short probes that are about 16 nucleotides to about 70 nucleotides in length, stringency conditions are defined as prehybridization and hybridization at about SrQ to about 19*C helow the calculated T(!! using thecalculation according to Bolton and fv1cC!arthy:(10§2, Pros, Natl, Amd.Sci. USA 48: 1300} in 0.9 M MaCi, 0.09; fvf Tris^HC! pH 7.6, 8 mi ΕΟΤΑ» 0.6% NfMG, 1X Denhardt’s solution, 1 mfvl sodium pyrophosphate, 1 mM sodium monobasic phosphate, 0.1 mM ATP, and 0.2 mg of yeast RNA per ml following standard Southern blotting procedures for 12 to 24 hours optimally. The carrier material Is finally washed:once in 8X SCG plus 0.1% SOS for 16 minutes and twice each for 15 minutes using 6X SSC at S-G to 10'C below the calculated T,„, in a third aspect, the parent is encoded by a poiynucieoilde with a sequence identity to the mature polypeptide coding sequence of SEQ ID NO: 1 of a! least 60%, e g., at feast 85%. at least 70%, at least 76%, at least 80%, at least 85%, at least 90%, at least 95%, at least 98%., at least 97%, at least 98%, at least 99%, or 100%, which encodes a polypeptide which is abieto fehetidh as an albumin, in an embodiment, the parent is encoded by a pblynucieotide comprising or consisting of 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 comprising the steps of: a. Identifying one or more amino acid residue positions being important for the binding of albumin to FcRn, in an albumin or a fragment thereof or the albumin part of a fusion polypeptide comprising albumin or a fragment thereof; b. Providing a nucleic acid encoding said albumin, the fragment thereof or the albumin part of a fusion polypeptide comprising albumin or the fragment thereof; o, Modifying the nucleic acid provided In b„ so that the one or more {several} amino acid residue located at the positions identified in a., are deleted or substituted or inserted with a different amino acid; d. Expressing the modified nucleic acid In a suitable host ceil: and e. Recovering the variant albumin, the fragment thereof or the fusion polypeptide comprising variant albumin or the fragment thereof.

The identification of one or more amino acid residue positions being important for the binding of albumin to FcRn, in albumin, fragment thereof or the albumin part of a fusion polypeptide can be done In several ways including, but not limited to, random mutagenesis followed by analysis of the generated mutants and comparison with the non-mutated parent molecule, and identification based on structural considerations optionally followed by generation of variants having the identified alterations and comparison with the nob-mutated patent molecule. A preferred method for identification of one or more amino add residue positions to be changed to in order to prepare a variant HSA having an altered binding to FcRn compared with natural HSA, comprises the following steps: i) identifying a non-human albumin having a different binding property to FcRn' si) identifying tne amino acid residues of the human serum albumin interacting with

FcRn; iii) Comparing the primary and/or the tertiary structure of the identified non-human albumin and human serum albumin with respect to ihe amino acid residues identified in step ii) and ideutify;ng the: amino acid residues that differ between said non-human albumin and human serum albumin as being responsible for the observed binding difference; and lv) Optionally preparing variants of HSA at the positions identified in step iii) and confirming that the prepared variants have altered binding to FcRn compared with HSA,

Stop I) above may be done using the SPR assay described below. However, the skilled person will appreciate that other methods may be used to identify non-human albumins having different binding properties to FcRn than HSA, and that the method is not dependent on how the non- human albumin, having different binding properties to FcRn, has been identified. in one preferred embodiment the identified non-human aibumin has a stronger binding to FcRn than HSA. Examples of non-human albumins having stronger binding to FcRn than HSA include donkey serum albumin, rabbit serum aibumin., dog serum aibumin, banister serum albumin, guinea pig serum albumin, mouse serum aibumin and rat serum aibumin. Step if) may be accomplished by considering the structure of FcRn, HSA and the binding complex of these two, in the absence of an available structure of the binding complex it is possible to use a mode! where the HSA structure is docked into the structure of the FcRn structure and thereby Identify amino acid residues of HSA Interacting with FcRb. in aNother preferred embodimahi the identiflad non-humail aibumin has a weaker binding to FcRn than HSA, Examples of hon-humah albumins having weaker binding to FcRn than HSA include bovine serum albumin,, goat serum albumin, sheep serum aibumin and chicken serum albumin, Step 1!) may be accomplished by considering the stryctyfe of FqRnyBSAand ibe binding complex of these two, In absence of an available structure of the binding complex ft is possible to use d model where the HSA structure is docked: into trie structure of the FcRn structure and thereby identify residues: Of HSA interacting with FcRn.

In this invention and claims, an amino acid residues of HSA Interacting with FcRn is considered any amino acid residues of HSA being located less than 10A from an amino acid in the FcRn or any amino acid residue that is involved in a hydrogen bond, a salt bridge or a'.polar or nonpolar interaction with an amino acid residue that is located less than fOA from an amino add in the FcRn.Preferably the amine acid in HSA residues are located less than 16A from amino acids in the FcRn, more preferred less than SAf'rom amino acids in thelFoRn and most preferred less than 3&amp; from amino: acids in the FcRn,

Step ill) and ivj can be done using techniques well known to the skilled person.

The present invention also relates to: methods for obtaining a variant albumin or fragments thereof, or fusion polypeptides comprising the variant albumin or fragments thereof, or associates of variant albumin or fragment thereof comprising: (a); Introducing into a parent albumin or fragments thereof, of fusion polypeptides comprising the parent albumin or fragments thereof an alteration at one; or more (several) positions corresponding to positions 417, 440, 464, 490, 402, 493, 494, 495, 406, 499, 500, 561* 503,, 504, 505, 505, :510¾ 535, 536, 537, 535, 540, 541, 542, 550, 573, 574, 575,, 577, :578, 579, 580¾ 581, 582 arid 584 of the mature polypeptide of SEQ IP NO: 2; and (b) recovering the variant albumin or fragments thereof, or fusion polypeptides comprising the variant albumin or fragments: thereof.

The variants can he prepared by those skilled persons using any mutagenesis procedure known In the art, such as site-directed mutagenesis, synthetic gene construction, semi-synthetic gene construction, random mutagenesis, shuffling, etc.

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

Site-directed mutagenesis can be aceempiished m vitro by FCR involving the use of oligonucleotide primers containing the desired mutation. Site-directed mutagenesis can aiso be performed in vitro by cassette mutagenesis involving the cleavage by a restriction enzyme at a site in the plasmid comprising a polynucleotide encoding the parent and subsequent ligation of an oligonucleotide containlngibe mutation in the polynucleotide, Usually the restriction enzyme that digests at the: plasmid and, the oligonucleotide is the same, permitting ligation of the plasmid and insert to one another, See, e.g,s Scharpr and Davis, 1979, fifbc. NM. Acad. ScL USA 76: 4849-4955; and Barton etal, 1999, Nucleic Acids Res. 15:7349-4866.

Site-directed mutagenesis can also be accompilshed #? vivo by methods known in the art. See, e,g,, U.S.. Patent Application Publication No. 2804/81:71154; Storici at ak ::2001, Nature Bbtechnoi, 19:773-776; Kren at pi, 1998, Nat.. Mad. 4: 235-280; and Oafissanb and tvladno, 1996, Bengal Genet. Newsiett 43:15-16.

Any site-directed mutagenesis procedure can be used in the present Invention. There are many commercial kits available that can be used to prepare:variants.

Synthetic gene construction entails in vitro synthesis of a designed polynucleotide molecule to encode a! polypeptide of interest. Gene synthesis can be performed utilizing a number of techniques, such as the multiplex microchip-based technoiogy described by Tlan ei al. (2Q04, Natum 432: 1050-1054} and similar teehRoiogies wherein olgionucleotldes ane synthesized and assembled upon phcto-programahle microfluidic chips.

Single or multiple amino acid substitutions^ deletions, and/or insertions can foe made and tested using known methods of mutagenesis, recombination, and/or shuffling, followed by a relevant screening procedure, such as those disclosed by Seidhaar-Oison and Sauer, 1888, Science 241: 53-57; Bowie and Sauer, 1989, Proo. Nati. Acad Sod USA 86: 2152-2158; WO 95/17413: or WO 95/22625. Other methods that can be Used include error-prone PCR, phage display (e.g., Lowman ef a/., 1991, Bio&amp;mmistiy 30; 10832-10837; U.S. Patent No, 5,223,49% WO 92/06204) and region-directed mutagenesis (Derbyshire at ai*< 1986, Gene 46: 145; Nor at ai < 1988, DNA 7:127). yulagenesis/shuffling methods can foe combined with high-throughput, automated screening methods to detect activity of cloned, mutageoized polypeptides expressed by host ceils (Ness et a/., 1999, Natum Biotechnology 17: 893-896). Mutagenized DNA molecules that encode active polypeptides can be recovered from the host cells and rapidly sequenced using standard methods in the art. These methods allow the rapid determination of the importance of Individual amino acid: residues In a polypeptide.

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

Warrants

The present invention also provides variant albumins or fmgments thereof, or fusion polypeptides comprising the variant albumin or fragments thereof, of a parent albumin, comprising an alteration at one dr more fseverai) positions corresponding to positions 417* 440; 464,490, 482, 493, 494, 495, 496, 499, 599, 591, 503.504, 505, 506, 51:0, 535, 538, 537, 538, 540, 541, 542, 550, 573, 574, 575, 577, 578, 579, 589, 581, 582 and 584 in SEC} iD NO: 2, wherein each alteration is independently a substitution, insertion or deletion with the provision that the and the variant is not SEQ ID NO: 2 having the substitution D494N, ES01K, K541E, Q550GA, K573B or K574N..

The variant albumin, a fragment thereof, or albumin part of a fusion polypeptide comprising variant albumin or a fragment thereof according to the invention has generally a sequence Identity the sequence of HSA shown in SEQ ID NO:: 2 of at least 00%, preferably at least 70%, preferably at least 80%, preferably at least 85%, preferably at.bast 90 more preferred at least 95%., more preferred at least 90%, more preferred at least 97%, more preferred at least 98% and most preferred at least 99%. in one aspect, the number of alterations in the variants of the present invehtibn Is 1-20, e,g,. 1-10 and 1~S, such as 1, 2, 3,4,5. 8,X 8.9 or 10 alterations.

The variant albumin, a fragment thereof or fusion polypeptide comprising the variant aibumin or fragment thereof has altered plasma half-life compared with the corresponding parent albumin, fragment thereof, or fusion polypeptide comprising the variant albumin or fragment thereof. in a particular preferred embodiment the parent albumin is HSA and the variant albumin, a fragment thereof or fusion polypeptide comprising the variant albumin or fragment thereof has altered plasma half-life compared wife the HSA, the corresponding fragment or fusion polypeptide comprising HSA:or fragment thereof.

The correlation between binding of albumin to its receptor and plasma half-life has been realized by the present id venters based: on the natural occurring allele of HSA 0404N, The inventors have analyzed this allele and found that it has a lower affinity to its receptor FcRn.

Further, St has been disclosed that a transgenic mouse having the natural mouse FcRn replaced with human FcRn has a higher serum: albumin level than normal mouse; see {3 Exp Med. (2003} 197(8:):315-22), The inventors: have discovered that human FcRn has a higher affinity to mouse serum: albumin than mouse FcRn has to mouse serum albumin and, therefore, the observed increase: in serum albumin in: the transgenic mice corresponds with a higher affinity between serum albumin and its receptor, confirming the correlation between albumin binding to FcRn and plasma half-life:, in addition, variants of albumin that have little or no binding to FcRn have been shown to hive reduced haibfife in a mouse: model, Kenanbva of a/ {2009} J. MM Med.: B0 {Supplement 2)::1582),

One: way to determine whether the affinity of a variant albumin to FcRn is' higher or lower than the parent albumin is to use the Surface PSasmon Resonance assay (8PR) as described beiow. The skilled person will understand that other methods might he useful to determine whether the affinity of a Variant albumin to FcRn is higher or lower than the affinity of the patent albumin to

FcRn, e.g., determination and comparison of the binding constants KD. Thus, according to the invention variant aibumins having a KD that is Sower than the KB for natural HSk is considered to have a higher piasma hatfdife than HSA and variant albumins having a KD that is higher than the KD for natural HSA is considered to have a lower plasma half life than HSA.

The variants of aibumin or fragments thereof or fusion polypeptides comprising albumin or fragments thereof comprise one or more alterations, such as substitutions, deiefions or insertions at one or more (several) positions corresponding to the positions in HSA 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, S3B, 540, 541, 542, 550, 573, 574, 575, 577,578, 579, 580, 581, 582 and 584. The substitution may be any substitution where the amino add in the natural albumin sequence is substituted with a different amino acid selected among the remaining 19 natural occurring amino acids.

In one aspect;, a variant comprises art alteration at one or more {several} positions corresponding to' positions 417, 44¾ 464, 490. 482, 483, 494, 495, 496, 499. 500, 501, 503,504, 505, ; 806, 510,635,:538, 537,538, 540, 541, 54¾ 66¾ 573,674, 575,; 677, 578, 679, 580, 881,582 and 884 in SEQ ID MO: 2. in another aspect, a variant comprises ah alteration at two positions corresponding to any Of 447, 440, 464, 490, 492.493, 494, 495, 496,499, 500. 501,503, 504, 505, 506, 510, 535, 536, 837, 833, 540, 541,542, 550, 573, 574,575,677, 578, 579, 580, 581, 582 and 584 in SEO ID NO: 2. In another aspect, a variant comprises an alteration at three positions corresponding to any of positions 417,440, 464,490. 492,493,484, 496, 486, 499,590, SOI, 503, 504, 805, 506, 510, 635, 536, 837, 538, 540, 541, 542, 550, 573, 574, 875, 577, 878, 679, 530, 561, 582 and 584 in SEQ ID NO: 2. in another aspect, a variant comprises an alteration at each position corresponding to positions 417, 440, 484, 490, 492, 493, 494, 495, 496, 499, 500. 501, 503, 504, 505, 506, 510, 535, 538, 537, 538, 540, 541, 542, 850, 573, 574, 575, 577, 578, 579, 580, 881,582 and 884 in SEQ ID NO: £ in another aspect, the variant; comprises the substitution Q417A,H of the mature polypeptide of SEQ ID NO: 2, In another aspect, the variant composes the substitution H44DQ of the mature polypeptide of SEQ ID NO: 2, in another aspect, the variant comprises the substitution H464Q of the mature polypeptide of SEQ ID NO: 2, In another aspect, the variant comprises the substitution A490D of the mature polypeptide of SEQ ID NO: 2, in another aspect, the variant comprises the substitution E4I2Q, Τ,Ρ,Μ of the mature polypeptide of SEQ IQ NO:: 2. in another aspect, the variant oj^pris®$;'the;sub^p|»i' V493f?,L of the mature polypeptide of SEQ ID NO: 2, In another aspect, the variant comprises the substitution D494N,Q,A,EfP of the mature polypeptide of SEQ ID NO: 2. in another aspect, the variant comprises the substitution E498Q.A of the mature polypeptide of SEQ ID NO: 2. in another aspect, the variant comprises the substitution T496A of the mature polypeptide of SEQ ID ND; 2, in another -ihe: variant :ic©mpflses;^^Substitution P489A of the mature polypeptide Of SEQ ID NO: 2, In another aspect, the vanant comprises the substitution ^ύΟΕ,β,Ο,^βΑΕ,Η,Ε,Μ^^ of the mature polypeptide of SEQ ID NQ:2. in another aspect, the variant comprises the substitution ES()1A,P?Q of the mature poiypeptide of SEQ 10 NO: 2. in another aspect, the variant comprises the substitution N5(33K\D,H of the mature polypeptide of SEQ ID NO: 2, in another aspect, the variant comprises the substitution A504E of the mature polypeptide of SEQ ID NO: 2. In another aspect, the variant comprises the substitution E505K, D of the mature poiypeptide of SEQ iD NO: 2, in another aspect, the variant comprises the substitution T506F, S of the mature poiypeptide of SEQ iD NO: 2, In another aspect, the variant comprises the substitution H510Q of the mature polypeptide of SEQ ID NO: 2. in another aspect, the vanant comprises the substitution HS3SQ of the mature poiypepfide of SEQ ID NO: 2. in another aspect, the variant comprises the substitution KS36A efthe maturepolypeptide of SEQ !D NO: 2, in another aspect, the variant comprises the substitution P537A of the mature poiypeptide of SEQ ID NO: 2. in another aspect, the variant comprises the substitution K538A,H of the mature poiypeptide of SEQ ID NO: 2, in another aspect, the variant comprises the substitution T640S of the mature polypeptide of SEQ. iD NO; 2. in another aspect, the variant comprises the substitution KS41 A,D,G,N,E of the mature polypeptide of SEQ ID NO: 2. In another aspect, the variant comprises the substitution E542P.D of the mature polypeptide of SEQ ID NO: 2. in another aspect, the variant comprises: the substitution D55SN of the mature polypeptide of SEQ ID NO; 2. In a n o t h e r a s ρ e c t, t h e v a r t a n t c o mi p r I s e s the substitution of the mature poiypeptide of SEQ ID NO: 2. in a nether: aspect, the variant comprises the substitution K574N of the matu re poiypeptide of SEQ i ΟΝΟ: 2, In another aspect, the variant comprises the substituiton Q580K of the matyre poiypeptide ef SEQ ID NO: 2. to another aspect, the variant comprises: the substitution L57SF of the mature poiypepfide of SEQ iD NO: 2. in another aspect, the variant:comprises the substitution Α077Τ,Ε of the mature poiypepfide of SEQ ID NO: 2. in another aspect, the variant comprises the substitution A57BR.S of the mature poiypeptide of SEQ ID NO: 2, in another aspect, the variant comprises the substitution SS78C,T of the mature polypeptide of SEQ ID NO: 2. in another aspect, the variant comprises the substitution QS80K of the mature polypeptide of SEQ ID NO: 2, in another aspect, the Variant comprises the substitution A581D of the mature poiypeptide of SEQ ID NO: 2, in another aspect, the variant comprises the substitution A582T of the mature polypeptide of SEQ iD NO: 2 In another aspect, the variant comprises the substitution G584A of the mature polypeptide of SEQ ID NO; 2. in one aspect, the variant comprises: an alteration dt p position corresponding to position 417, In another aspect, the: amirio acid at a posliion corresponding to position 417 is 'Substituted with Aid, Arg, Asn, Asp, Cys, Gin, Glu, Giy,: His, lie, Leu, Lys, Met, Hhe, Pro, Ser, Thr, Trp:, Typor Vai, preferably· with Aia or His. In another aspect, the variant comprises the substitution Q417A, H of the malum polypeptide of SEG ID NO; 2. in another aspect, the variant comprises an. alteration at a position corresponding to position 446. in another aspect, the amino acid at a position corresponding to position 440 is substituted with Aia, Arg, Asm Asp, Cys, Bin. Glu, Giy, His, lie, leu».lys, Met, Phe, Pro, Ser, Thr, Tip, Tyr, or Vai, preferably with Ala. in another aspect, the variant comprises the substitution H440Q of the mature polypeptide of SEQ ID NO; 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 464. in another aspect, the amino acid at a position corresponding to position 464 is substituted with Ala, Arg, Ash, Asp, Cys. Gin, Glu, Gly:: His, lie. Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr. or Vai, preferably With Ala. in another aspect, the variant comprises the substitution H464Q of the mature polypeptide of SEQ ID NO; 2.

In another aspect the variant comprises an alteration at a position corresponding to position 490 In another aspebt, the amino acid at a position corresponding to position 490 is substituted with Ala, Arg, Asm Asp, Gys, Gin, Glu,: Giy, His, tie, Leu, Lys. Met, Phe, Pro, Set, Thr, Trp, Tyr, or Mai, In another aspect, the variant comprises the substitution A490G of the mature poiypeptide ef SEG ID 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 substituted with Aia, Arg, Asn, Asp,: Cys> Gin, GiU, :Gly,: His,: lie, De% Lys, Met, Phe,: Pro, Ser, Thp Trp, Tyr, or Vai, preferably with Giy. in another aspect, the variant comprises the substitution E492G of the mature polypeptide of SEG ID NO; 2.

In another aspect, the vananf eompnses an alteration at a position corresponding to position 493, in another aspect, the amino acid at a:'posit!on correspondingtO: position 493 Is substituted with Aia, Arp, Asn, Asp, Cys, Gin, Glu, Gfy, His* lie, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, or Vai, preferably with Fro, in another aspect, the variant comprises the substitution V493P of the mature polypeptide ofSEQJQ NO: 2. in another aspect, the variant comprises:; an aileratioh at a position corresponding to position 494. In another aspect, the amino acid at a; position corresponding lo position 494 is substituted With Ala, Arg, Asn, Asp, Cys, Gih, Glu, Giy,, His, lie. Leu, Lys, Met, Phe, Pro, Ser, Thr. Trp, Tyr, or Vai, preferably with Asn, Gin or Ala, In another aspect, the variant comprises the substitution D494N.Q, A of the mature poiypeptide of SEG ID NO: 2, in another aspect, the variant comprises:an alteration at a position edrresponding to position 495, in another aspect, the: amino add at a position corresponding to position 495 is substituted with Ala, Arg, Ash, Asp, Cys, Gim Giu, Gly, His, lie, Leu, Lys, Met* Phe, Pro, Ser, Thr, Trp;, Tyr, or Vai, preferably with Gin or Ala - In another aspect, the variant comprises the substitution E495G or A of the mature polypeptide of SEO. 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 substituted with Aia, Arg, Asn, Asp, Cys, Gin, Glu, Gly, His, lie, Leu, Lys, .Met, Phe, Pro, Ser, Thr, Trp, Tyr, or Vai, preferabiy with Aia. in another aspect, the variant comprises the substitution T498A of the mature 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 substituted with Ala, Arg, Asn, Asp, Cys, Gin, Giu, eiy!: His; lie, .Ley, Lys, Met, Pile, Pro, Ser, Thr, Trp,: Tyr. or Vai, preferabiy with Ala. In another aspect, the variant comprises the substitution P499A of the mature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 600. in another aspect, the amine add at a position corresponding 'to position 600 Is subsmuied with Aia, Arg, Asn, Asp, Cys, Gin, Giu< Gly. His, lie, Leu, Lys, Met, Phe, Pro, Ser, Thr,: Trp, Tyr. or Vai, preferabiy with Aia, In another aspect, the variant comprises: the substitution K5G0E,6,D, A,S,G, P, H, F,N,vy ,T,Μ,: V; V, Q * L,i, R of the mature poiypeptlde of SEQ ID NO: % in another aspect, the variant comprises an alteration at a position corresponding to position 601. in another aspect, the amino acid at a position corresponding to position 601 is substituted with Aia, Arg, Asn, Asp, Cys. Gin, Glu, Gly, His, lie, Leu, Lys, Met, IP he. Pro, Ser, Thr, Trp, Tyr, or Vai, preferabiy with Aia or Gin to reduce affinity and Pro to Increase affinity, in another aspect, the variant comprises the substitution E561A, Q, P of the mature polypeptide of SEQ ID NO: 2. in another aspect the variant comprises an aiferation at a position corresponding to position 603. In another aspect, the amino acid at a position corresponding to position 503 is substituted with Aia, Arg, Asn, Asp, Cys, Gin, Giu, Gfy, Bis, lie, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, or Vai, preferably with Asp or Lys or His. in another aspect, the variant comprises the substitution N503D. K, H of the mature polypeptide of SEQ ID NO: 2. in another aspect, the variant comprises an a iteration at a position corresponding to position 504. in another aspect, the amino acid at a position corresponding to position 604 is substituted with Aia, Arg, Asn, Asp, Cys, Gin, Giu, Gly, His, lie. Leu, Lys, Mot. Phe, Pro, Ser, Thr, Trp, Tyr, or Vai. in another aspect, the variant comprises the substitution A504 of the mature polypeptide of SEQ iD NO: 2. in another aspect, the variant comprises an alteration at a position eorrespondirig to position 595. in another aspect the amino acid at a position corresponding to position 505 is substituted with Ala, Arg, Asa, Asp, Cys, Gin, Giu, Giy, His, li#, Leu, Lys, Met, Ebe, Pro, Ser, Thr, frp, Tyr, or Val in another aspect, the variant comprises the substitution E605Q 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 ado at a position corresponding to position 506 is substituted with Ala, Arg, Asn, Asp, Cys* Bin, Gkp Gty, His, lie, leu, lys, Met, Phe, Fro, Ser* Thr, Trp, Tyr, or Val. i n another aspect,: the variant: corn prises the su bstitution T506S*F of the mature polypeptide: of SEQ ID NO: 2.

In another aspect, the variant comprises an, alteration at a position -corresponding to position 610, In another aspect, the amino acid at a position corresponding to position 510 is substituted with Ala, Arg, Ash, Asp, Cys. Din, Giu, Sly, His, lie. Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr,, or Val, preferably with Gin. In another aspect, the variant comprises the substitution H510Q of the feature polypeptide of SEQ id NO: 2, !h another aspect, the variant comprises an alteration at a position corresponding to position 535, In another aspect, the afeino: acid at a position comgspdridlng to position 535 Is substituted with Aia, Arg, Ash,, Asp, Gys, Gin, Giu,: Gly, His, tie, Leu, Lys, Mat, Phe, Pro, Ser, Thr, Trp, Tyr, or Val, preferably With Gin, In anptheii aspect, the variant comprises the substitution H535G Of the mature 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 substituted with Aia, Arg, Asn, Asp, Cys. Gin, Giu, Gty, His, lie, leu, tys, yet, Phe, Pro, Ser, Thr, Trp, Tyr, or Val. preferably with Ala. in another aspect, the variant comprises the substitution K538A of the mature 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 eorresponding to position S3? is substituted with Aia, Arg, Asn, Asp, Cys, Gin, Giu, Gty, His, lie, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, or Val, preferably wife Ala. in another aspect, the variant comprises the substitution E637A of the mature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 533, In another aspect, fee amino acid at a position corresponding to 'position 538 is substituted with Aim Arg, Asn, Asp, Cys, Gin, Giu, Gly, His, lie, iep, tys. Met, Phe, Pro, Ser,. Thr, Trp, Tyr,.or Val, preferably with Ale, in another aspect, the variant comprises the substitution K538M, A of the mature polypeptide of SEQ ID NO: 2, in another aspect, the variant comprises an alteration at a position corresponding to position S4Q, in another aspect, fee amino acid at a position corresponding to position 540 is substityfed with Ala, Arg, Asn, Asp, Cys, Gin, Glu, Giy, His, lie, Leu, Lys, Met, Pbe, Pro, Ser, Thr, Trp, Tyr, or Vat In another aspect, the variant comprises the substitution T64SS of the mature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 341, in another aspect, the amino acid at a position corresponding to position 541 is substituted with Ala, Arg, Asn, Asp, Cys, Gin, Glu, Giy, His, lie, Leu, Lys, Met, Fhe, Pro, Ser, Thr, Irp, Tyr, or Vai, 'preferably with Giy, Asp: or Ala, In another aspect, the variant comprises the substitution K;S4iG} 0 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 642 is substituted with Aia, Arg, Asn, Asp, Cys, Gin, Gio, Giy, His, lie, Leu, Lys, Met, Pbe, Pro, Ser, Thr, Tfp¥: Tyr,: or Vai, preferably with Asp or Pro. In another aspect, the variant comprises the substitution E542D, P of the niature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to posltipri 550, in another aspect* the amino acid at a position corresponding to position 550 is substituted with Ala, Arg, Ash, Asp, Cys, Glh, Gib* Giy, His, He, Leu, Lys, Mat, Phe, Pro, Ser, Thr, Trp, Tymor Vai, 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 673, in another aspect, the amino acid at a position corresponding to position 573 is substituted with Aia, Arg, Asn, Asp, Cys. Gin, Glu, Giy. His, iie, Leu, Lys, Met, Phe, Pro, Sec Thr, Trp, Tyr. or Vai, .preferably with Tyr, Trp, Pro. His. Phe, Vai, lie, Thr, Asn, Ser, Giy, Met, Cys, Aia, Glu, Gin, Arg, Leu, Asp, In another aspect, the variant comprises the substitution ΚΘΤδΥ^,Ρ,Η,Ρ,νΑΤ,Ν,Ε,Ο,Μ,Ο,Α,Ε,Ο,Η,Ι,Ο of the mature poiypeptide of SEQ ID NO: 2, in another aspect, the variant comprises an alteration at a position corresponding to position 674, in another aspect, the amino acid- at a position corresponding to position 574 is substituted with Aia, Arg, Asn, Asp, Cys, Gin, Glu, Giy, His* lie, Leu, Lys, Met* Phe, Pro, Ser, Thr, Trp, Tyr, or Vai, preferably with Asn, in another aspect, the variant comprises the substitution K574N of the mature poiypeptide of SEQ ID NO: 2, in another aspect, the variant comprises an a iteration at a position corresponding to position 575. in another aspect, the amino acid at a position corresponding to position 575 is substituted with Ala, Arg, Ash, Asp, Cys, Gin, Glu, Giy, His, lie, Leu, Lys, MebiPhe, Pro, Ser, Thr, Trp, Tyr, of Vai, preferably with Phe, Id; another aspect, the: variant comprises the substitution L575F of the mature poiypepfide of SEQ1D NO: 2. in another aspect, the variant comprises an alteration at a position corresponding to position 677, in another aspect, the amino acid at a position corresponding to position 577 is substituted with Ala, Arg, Asn, Asp, Cys, Gin, Glu, Gfy, His* lie, Leu,, tys, Met, Phes:Pro, Ser, Thr, Trp, Tyr,: or VaL preferably with Thr or Glu, In another aspect the variant comprises tho subsfitution A577TE of the mature polypeptide of SEC ID NO: 2. in another aspect, the variant comprises an. alteration at a position, corresponding to position 578, In another aspect, the amino add at a position corresponding to position 578 is substituted with Ala, Arg, Asn, Asp, Gys, Gin, Glu* Giy, His, lie, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, or Val, preferably with Arg or Sen in another aspect, the variant comprises ire substitution A578fl,S of the mature polypeptide of SEQ ID NO: 2.

In another aspect, the variant comprises an alteration at a position corresponding to position 679, In another aspect, the amino acid at; a position corresponding to position 579 is substituted with Ala, Arg, Asn, Asp, Cys, Gin, Glu, Giy, N is, He, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, or Vai, preferably with Gys or The in another aspbcf, the variant comprises the substitution S579C,T Of the mature polypeptide of SEQ iD NO: 2. in another aspect, the variant comprises an afteratiph at a position corresponding to position 580, in another aspect, the amino acid at a position correspending to position 580 is substituted with Ala, Arg, Asn, Asp, Gys, Gin, Glu, Giy, His, He, Leu, Lys, Met, Phe, Pro, See The Tip, Tyr, or Val, preferably with Lys. In another aspect, the variant comprises the substitution GS80K of the mature polypeptide of SEQ 10 NO: 2. in another aspect, the variant comprises an alteration at a positioncorresponding to position 581, In another aspect, tool amino add at a position corresponding to position 581 Is substituted with Ala, Arg, Asp, Asp, Cys, Gin, Glu, Giy, His, He, Leu, Lys, Met, Phe,: Pro, Ser, Thr, Trp, Tyr, or Val, preferably with Asp. in another aspect, the variant comprises: the substitution A581G of the mature polypeptide of SEQ1D NO: 2. fa another aspect, the vananf eomprises an alteration at a position corresponding to position 582, in another aspect, the amino acid at a position corresponding: to position 582 is substituted with Aia, Arg:, Asn, Asp, Gys, Gin, Glu, Giy, His* fie. Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, or Val, preferably with Thr, in another aspect, the variant comprises: the substitution A582T of the mature polypeptide of SEQ 10 NO: 2, in another aspect, the variant comprises an alteration at a position corresponding to position 584. in another aspect, the amino acid at a position corresponding to position 584 is substituted with Aia, Arg, Asn, Asp, Cys, Gin, Glu, Giy, His, He, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, or Val, preferably with Aia, in another aspect, the variant comprises the substitution G584A of the mature polypeptide of 8EG iD NO: 2, in another aspect, the variant comprises an alteration at positions corresponding to positions 494 and 498 in SEQ ID NO: 2, such as those described above.

In another aspect, the variant comprises alterations at positions corresponding to positions 492 and 493 in 8EG SO: NO: 2, such as those described above.

In another aspect, the variant comprises .'alterations at positions corresponding to positions 494 and 417 In SEQ ID NO: 2, such as those described above. in another aspect, the variant comprises alterations at positions corresponding to positions 492 and 503: In SEQ ID MO: 2, such as those described above. in another aspect, the variant comprises alterations at positions corresponding to positions 492 and 573 In SEQ. iD NO: 2, such as those described above. in another aspect the variant comprises alterations at positions corresponding to positions 492, 503, and 573 in SEQ ID NO; 2, such as those described above.

In one: embodiment the variant albumin or fragments thereof, or fusion polypeptides comprising; the variant albumin or fragments thereof according to the invention contains one substifufion at a position corresponding to a position in HSA selected from the group consisting of 417, 440, 464, 490, 492, 493, 494, 495, 496, 499, 500, 501, 503, 504, 506, 506, 510, 535, 536, §37, 538, §40, §41, 542, 550, 573, 574,575, 577, §78, §79, 580,681, 582 and §84 in SEQ ID NO: 2 provided that the variant aibumin is not the variant consisting of SEQ ID NO: 2 with the substitution P494N, £501K, K541E, D550G.A, K573E or K574N. The variant aibumin, fragment thereof or fusion polypeptides comprising variant aibumin or a fragment thereof according to the Invention may comprise additional substitutlonsE insertions or deletions at one or mere (several) positions corresponding to other positions: in HSA.

In another embodiment the: variant albumin or fragments thereof, or fysion polypeptides comprising variant aibumin 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 in HSA selected from the group consisting of 417,448, 464,490,492,493,494,495,496, 499, §00, 501, 593, §04, 505, §06, §40, 535, 638, 537, §38, §40, 541, 542, 550, 573, 574, 575, 577, 578, 579, 580, 581, 582 and 584 of SEQ lip NO; 2, The variant albumin or fragments thereof, or fusion polypeptides comprising variant albumin or fragments thereof according to the Invention may comprise additional substitutions, insertions or deletions at positions corresponding to other positions in HSA. in a further embodiment the variants albumin or fragments thereof, pr fusion polypeptides comprising variant aibumin or a fragment thereof according to the invention have a plasma half-life that is longer than the plasma half-life of the parent aibumin fragment thereof dr 'fusson polypeptide comprising the parent albumin or a fragment thereof. Examples according ip this embodiment include variants of albumin or fragments thereof, or fusion polypeptides comprising variant albumin or a fragment thereof comprising a substitution in the position corresponding to 492, 503, 542: 550, 673, §74!: 580, 581, 582 or 584 in SEQ ID NO: 2, 'Preferred substitutions according to this embodiment of the invention include the substitution: of the amino aek! residue in the position corresponding to 492 in SEQ ID NO: 2 with a G residue:, substitution of the amino acid residue in the position corresponding to 683 in SEQ· ID N0: 2 with aHoraK residue, substitution of the amino acid residue: in the position -corresponding i&amp; 550 in SEQ ID NO: 2 with an E residue, the substitution of the amino acid residue in a position corresponding to: 873 In SEQ .ID NO; 2 with an or a D, the substitution of the amino acid residue in a position corresponding to 574 in SEQ ID NO: 2 with an N residue, or the substitution of the amino acid residue in the position corresponding to 580 in SEQ ID NO: 2 with an K residue, Other preferred variants:have :a substitution in the position corresponding to 492 ip SEQ ID NO: 2 with a G residue and a: substitution in the position corresponding to 573 in SEQ ID NO: 2 with an A pr a P residue. Other preferred variant has a number of substitutions corresponding to posiieh: 492 in SEQ IDNQ: 2 with ran Hmesidue In position 503 in SEQ ID NO: 2.

Other preferred variants have a substitution in the posmon correspondingto 492 in SEQ ID NO: 2 with a G residue and a substitution in the position corresponding to position 503 in SEQ 10 NO: 2 corresponding to a H or a K and a substitution in position 573 In SEQ iD NO: 2 with an A or a P residue. in a further embodiment the variants of albumin or fragments thereof, or fusion polypeptides comprising variant albumin or fragments thereof according to the invention have a plasma half-! if© that is: shorter than the plasma hatf-life of the parent albumin fragment thereof or fusion polypeptide comprising the parent albumin or a fragment thereof. Examples according to this embodiment include variants of albumin or fragments thereof , or fusion polypeptides comprising variant albumin or a fragment thereof comprising a substitution in the position corresponding to 417, 440,494,495, 496, 499, 500, 501, 586, 537, 538, 541, 494+498 or 492*493 in SEQ ID NO: 2, Preferred substitutions include- the substitutions oorresponding to Q417A, H440Q, P494E+Q417H, D424N!Q,A, E495Q,A, T496A, 0494Ν*?495Α or, P499A. K50OA, E501A , E501Q, K536A, P537A , K538A. K541 £3, K541A K541D or D550N in SEQ ID NO: 2.

In another embodiment of the Invention the variants of albumin or fragments thereof, or fusion polypeptides comprising variant albumin or a fragment thereof according to the invention have lost their ability to bind FcRn. In this connection variants of albumin or fragments thereof, or fiiSion polypeptides comprising variant albumin or fregments ihered is considered to have lost the ability to bind FcRn if the measured resonance units for the variant In the SPR assay described below is less than 10% of the measured resonance units for the corresponding parent albumin or fragment thereof, Examples according to this embodiment include variants of albumin dr fragments thereof, or fusion polypeptides comprising variant albumin or fragments thereof comprising a substitution at a position corresponding to 464, 500* $10 or 63$ in SEQ ID NO: 2, Preferred substitutions include- the substitutions corresponding 10 H484Q;, ^ΟΟΑ,Ρ,β,β,Α,Ο.Θ HS1O0 Or H53SQ in SEQ iD NO: 2. in addition: to the one or more substitutions at one or more posi tions corresponding to positions 417,464,490, 492,493, 494, 495, 490,499, 500, SOI, 603, 504, 506, 506, 510, 535, 536, 537, 538,540,541,542, 550, 673, 574,530 531,582 and 584 in SEQ ID NO: 2 the variant albumin or fragments thereof, or fusion polypeptides comprising variant albumin or fragments thereof according to the invention may contain additionai substitutions, deletions or Insertions in other positions of the molecules. Such additionai substitutions, deletions or insertions may bo useful In order to alter other properties of the molecules such as but not: limited to altered glycosyiatlon; introduction of reactive groups of the surface such a thibi groups, removing/genemtiug a earbamoylatlon site; etc.

Residues that might be altered in order to provide reactive residues on the surface and which advantageously couid be applied to the present invention has been disciosed in the unpublished patent application WO 2910/092135 (fhctuded by reference). Particular preferred residues 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 positions in other albumins in order to provide a reactive thiol group on the surface includes alterations corresponding to following alterations in SEQ ID NO; 2' 15850:, DI G, A20, 05820, A384C, A504C, E505C, T79C,: i860, D1290,: D549C, A581G, D121G;, E82C, S27QC, A578C. L595LC, D1DC, A2AG, DSS2DC, A364AC, AS04AC, E505EC, T79T0, E8SEC, D129DC, D549DC, A5S1AC, A581AC, D121DC, E82EC, S270SC, A579AC, G360\ C3160 C7S3 Cl 68', 0558s', C361\ G91s, 012.4*, 0189s and G56?\ Alternatively a cysteine residue may be added to the N or C terminal of albumin.

The present: invention also relates to Isolated polynucleotides that encode any of: the variants of the present invention.

Nucleic Acid Constructs

The present invention also felaies to nucleic acid constructs comprising a polynucleotide encoding a variant of the present invention operahly linked to one or more (several) control sequences that direct the expression of the coding sequence in a suitable host coil under conditions compatibie with the control sequences. A polynucleotide may be manipulated in a variety of ways to provide for expression of a variant Manipulation of the polynucleotide prior to its insertion info a vector may be desi rable or necessary depending on the expression vector. The techniques for modifying polynucleotides utilizing recombinant DHA methods are welt known In the art.

The conlroi sequence may he a promoter sequence, which is recognized by a host ceil for expression of the polynucleotide. The promoter sequence contains transcriptional control sequences that mediate the expression of the variant. The promoter may be any nucleic acid sequence that shows transcriptional activity in the host ceti including mutant, truncated, and hybrid promoters, and may be obtained from genes encoding extracellular or intracellular polypeptides either homoiogous or heterologous to the host cell. in a yeast host, useful promoters are obtained from the genes for Saccharomyces cerevisiae enolase (ENO-1), Saccharomyces cerevisiae protease A (RRA1), Saccharomyces mrmisiae protease B (PRB1), Saccharomyces cerevisiae transiation elongation factor (TEF1), Saccharomyces cerevisiae translation eibngation factor {TEF2}, Saecharomyoes cerevisiae gaiactokinase '(0AL1), $8CCharo?nyc&amp;s eer&amp;ymiae alcohol dehydrogenase/gtyceraidehyde-3-phosphate dehydrogenase (ADH1. ADH2/GAP), Saccharomyces cerevisiae those phosphate isomerase (TPI), Saccharomyces cerevisiae metailothionein (CUP1), and Saccharomyces cerevisiae 3~phosphogiycerate kinase- Other useful promoters for yeast host cells are described by Romanos of af, 1992, Yeast 8:423-488,

The control sequence may also be a suitable transcription terminator sequence, which is recognized by a host ceil to terminate transcription, The terminator sequence is operabiy linked to the 3’4ermirais of the polynucleotide encoding the variant. Any terminator that Is functional in the host ceil may be used.

Preferred terminators for yeast hosf eeiis are obtained from the genes for Saccharomyces cerevisiae enolase, Saccharomyces cerevisiae: cytochrome G (GYG1}, Saccheromyces cerevisiae alcohol dehydrogenase (ADHi} and Saccharomyces cerevisiae giycera!dehyde~3~phosphaie dehydrogenase. Other useful terminators for yeast host cells are described by Romanos et af;, :1982, supra.

The control sequence may also be a suitable leader sequeneev a nontransiated region of an mRRA that is Important for transiation by the host cell. The leader Sequence is operabiy linked to the S'Aerminus of the polynucleotide encoding thevariant Any loader sequence that is functional in the host ceil may be used.

Suitable leaders for yeast host ceils are obtained from the genes ter Saccharomyces cerevisiae enolase (ENO-1), Saccharemyms cerevisiae Sqshosphogiycerate kinase,

Savcharomyc&amp;s cerevisi&amp;e alpha-factor. and Saccharomyces c&amp;mvisiae alcohol dehydrogenase/giycerBidehyde-S-phosphate dehydrogenase (ADH2/GAP).

The control sequence may also be a polyadehyiailon sequence, a sequence opembly linked Id the 3'-fermlnus ofthe variant-encoding sequence and, when transcribed, is recognized by the bast cell as a signal to add pdyadenoslhe residues id transcribed ffiRNA. Any polyadenylalion sequence that is fun etlonal in the host ceil may be used.

Useful poiyadenyiation 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 signal peptide linked to the N-terminus of a variant and directs the variant Into the celts secretory pathway.; The S’-end of the coding sequence of the polynucleotide may inherently contain a signal pepde Coding region naturqliy linked in translation segtoepief ·&amp;©:^inQ region that encodes the variant. Alternatively, the 5’-end of the coding sequence may contain a Signal peptide coding region that is foreign to .the coding: sequence; The foreign signal peptide coding region may be required where the coding sequence does hot naturally contain a signal peptide coding region. Alternatively,: the foreign signal peptide coding: region may simply replace the natural signal peptide boding: region In: order to enhance secretion of the variant. However, any signal peptide coding region that directs the expressed variant into the secretory pathway of a host ceil may be used.

Useful signal peptides for yeast host ceils are obtained from the genes for Saccharomyces cemvmlm alpha-factor and Saccharomyces cerevisiae invertase. Other useful signal peptide coding sequences are described by Romanos eial... 1992, supra.

Where both signal peptide and propeptide regions are present at tie N~terminus of a variant^ the propeptide:: region is positioned next to the N-terminus of the variant and the signal peptide: region is positioned next to the N-ierminus of the propeptide region,

Methods of Production

The variants of the present'invention can be prepared using techniques well known to the skilled person. One convenient way is by cloning nucleic acid encoding the parent albumin ora fragment thereof or fusion polypeptide comprising albumin or a fragment thereof; modifying said nucleic acid to Introduce the desired substitution's) at one or more (several) positions Corresponding to positions 417, 484, 490, 492, 49:3« 494, 495, 498, 499, 500, SOI, 503, 504, SOS, SOS, 510, 535, 538, 537, 538, 540, 541, 542, 550, 573, 574 and 588 in SEQ ID NO: E where the variant is not the variant consisting of SEQ ID NO:2 with the substitution D484M, E501K, K541E, D550GA K673E or R:574N., preparing | suitable genetic Construct where the modified nucleic acid i$ placed in operative connection with suitable regulatory genetic elements, such as promoter, terminator* activation sites* ribosome; binding sites etc,, Introducing the genetic construct Into a suitable dost organism, culturing the transformed host organism under conditions leading to expression of the variant and recovering tee variant. Ail these techniques are known in tee art and ills within the skills of the average practitioner to design a suitable method for preparing a particular variant according to the invention.

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

Techniques for preparing vacant polypeptides have also been disclosed in W0 2009019314 (inciuded 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 hosts including fungi (including but not limited to Aspergillus (WG06086395), Kiuyveromyces (Fleer 1981, BhAmfUKiiogy % 963-976), Prchla (Kobsyashi 1-896; "'fh&amp;mpeotic Aphmmsis 2, 267--262) and iSaccdammycds |$ieep 1990, BioMoftnoiogy 8, 42-46}}, bacteria (Pandjaiipb 290&amp; immungt 166, 279-285)), animals (Barash 1993, Trmsgenia Rmearch 2* 268-276} and plants (including but not limited to potato and tobacco (Simons 1990, BioffeGhnology 6, 217 and Farran 2002, Tr&amp;mgmm Research 11, 337-348), The variant polypeptide of the Inventten is preferabiy produced 'recbmbinantiy in a suitable host cell. In principle any host cell capable of producing a polypeptide in suitable amounts may be used and it is within the skills of the average practitioner to select a suitable host ceil aocordihg to the Invention. A preferred host organism is yeast, preferably Selected among Saechammycaeae, more preferred Saccharomyces cerevtsiae.

The variant polypeptides of the invention may be recovered and purified from the growth medium using a combination of known separatfon techniques suoh as filtration, centrifugation, chromatography, and affinity separation techniques etc. It is within the skis of the average practitioner to purify the variants of the invention using a particular combination of such known separation steps. As an example of purification techniques that may be applied to the variants of tee present invenflbn can be mentioned fee teaching of W00044772.

The variant polypeptides of the invention may be used for delivering a therapeutically beneficial compound to an animal or a human individual in need thereof. Such therapeuflcaiiy beneficial compounds Include, but are not limited, to labels and readily detectable compounds for use in diagnostics, such as various imaging techniques; pharmaceutical active compounds such as drugs, or specifically binding moieties such as antibodies. The variants of the invention may even be connected ^.fv^^prmore^itfef^t^herapeutically.l^^icfel compounds, e,g„ an antibody and a drug, which gives tile combined moiecuie the ahliity to bind specifically to a desired target and thereby provide a high concentrafton of the connected drug at that particular target

Fusion polypeptides

The variants of aihumsn or fragments thereof according to the invention may also be fused with a non-aibumin polypeptide fusion partner. The fusion partner may in principle be any polypeptide but generally lifs preferred that the fusion partner is a polypeptide having therapeutic or diagnostic properties. Fusion polypeptides comprising albumin or fragments thereof are known in the art. It has been found that such fusion polypeptide comprising albumin or a fragment thereof and a fusion partner polypeptide have a longer plasma: half-life compared to the unfused fusion partner polypeptide. According to the invention it: is possible to alter the plasma half-life of the fusion polypeptides according to the invention compared to the corresponding fusion polypeptides of the prior art.

One or more therapeutic polypeptides may be, fused to the N-ierminus, the Grterrnlnus of albumin, inserted Into a loop: in the albumin structure or any combination thereof, it may or if may not comprise linker segueoses separating the various-components of the fusion polypeptide-

Teachings relating to fusions of albumin or a fragment thereof are known in the art and the skilled: person will appreciate that such teachings can also be applied te the present Invention, WO 2001/79271 A and WQ 2093/59934 A also contain examples of therapeutic polypeptides that may be fused to albumin or frag meets thereof, and these exampies apply also to the present invention.

The variants of albumin or fragments thereof according to the invention may be conjugated to a second molecule using techniques know within the art. Said second molecule may comprise a diagnostic moiety fa in this embodiment the conjugate may be useful as a diagnostic tool such as in imaging: or the second moiecuie may be a therapeutic compound and in this embodiment the conjugate may be used for therapeutic purposes where the conjugate will have the therapeutic properties of the therapeutic compound as well as the long plasma haifdife of the albumin, Conjugates cf album in and a therapeutic molecule are known In the art and it has bean verified that such conjugates have long plasma haffdife compared with the nomconjugated, free therapeutic molecul®: as: such. The conjugates may conveniently be linked via a free thio group present on the surface 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 a beneficial therapeutic compound and the conjugate is used for treatment of a condition in a patient in need thereof, which condition is responsive to the particular selected therapeutic compound. Techniques for conjugating such a therapeutlcalSy compound to the variant albumin or fragment thereof are known in the art. WO 2009/010314 discloses examples of techniques suitable for conjugating a therapeutically compound to a polypeptide which techniques can also be applied to the present invention. Further WO 2009/019314 discloses examples of compounds and moieties that may be conjugated to substituted transferrin and these examples may aiso be appiled to the present invention. The teaching of WO 2009/019314 is included herein by reference. HSA contains in its natural form one free thiol group that conveniently may be used for conjugation. As a particuMr embodiment within this aspect the variant albumin or fragment thereof may comprise further modifications provided to generate additional free thiol groups on the surface. This has the benefit that the payload of the variant aibumin or fragment thereof is increased so that more than one molecule of the therapeutic compound can be conjugated to each molecule of variant albumin or fragment thereof, or two or more different therapeutic compounds may be conjugated to each molecule of variant albumin or fragment thereof, e.g,, a compound having targeting properties such as an antibody specific for example a tumour; and a cytotoxic drug conjugated to the variant aibumin or fragment thereof thereby creating a highly specific drug against a tumour. Teaching of particular residues that may he modified to provide for further free thiol groups on the surface can be found in copending patent application WO 2010/092135, which is incorporated by reference.

Associates

The variants erf aibumin or fragments thereof may further be used in form of "'associates", in this connection the term "associate" is intended to mean a compound comprising a variant of albumin or a fragment thereof and another compound bound or associated to the variant albumin or fragment thereof by non-covaient binding. As an example of such an associate can be mentioned eh{associa^ogeliltihg.vari$rit aibumin and a lipid associated to aibumin by a hydrophobic Interaction. 'Such;'associates are known in the art and they may be .prepared using welf known iechnigues. As an example of a preferred associate according tp the inyehtioh can be mentioned an associate comprising variant albumin and paciitaxei

Other uses

The variant albumin or fragments thereof or fusion polypeptides comprising variant aibumin or fragments thereof according to the invention have the benefit that their piasma half-life is altered compared to the parent aibumin: or fragments thereof or fusion polypeptides comprising parent albumin or fragments hereof. This: has the advantage that the piasma half-life erf conjugates comprising variant albumin or a fragment thereof or fusion polypeptide comprising variant aibumin or a fragment thereof, or an associate comprising variant albumin or a fragment thereof according to the invention can bo selected in accordance .-with the particular therapeutic purpose.

For example for a^epfij!L(gate,^associafeQr:ftίSίoπ polypeptide used for imaging purposes in animals or human beings, where the imaging moiety has an; very abort iialMife and a conjugate or a fusion polypeptide comprising: HSA has a piasma haff-ffe that is far longer than needed for the imaging purposes it won id be advantageous to use a variant albumin or fragment thereof of the invention having a shorter plasma; hail-life than the parent albumin or fragment thereof, to provide conjugates effusion polypeptides having a plasma: haff-iffe that Is sufficiently long for the imaging purpose but :sufficientiy short: to be cleared form the body of the particular patient on which if is applied. in another example for a conjugate, an associate or fusion poiypepiide composing a therapeutic compound effective to treat or alleviate a particular condition in a patient In need for such a treatment it would be advantageous to use the variant albumin or fragment thereof having a longer plasma haif-iife than the parent albumin or fragment thereof, to provide associates or conjugates or fusion polypeptides having longer plasma halfdives which wouid have the benefit that the administration of the associate or conjugate or fusion polypeptide of the invention would be needed less frequently or reduced dose with less side affects cbmparedtethe situation where the parent 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 fusion polypeptide comprising variant albumin or fragment thereof according to the invention. The compositions are preferably pbarmabeutica! compositions. The composition may be prepared using techniques known in the area such as disclosed in recognized handbooks within the pharmaceutical field. in a pedicular embodiment the compositions: comprise a variant albumin or a fragment: thereof according to the invention and a compound comprising aipharmaceutically beneficial moiety and an albumin: binding domain |ABD), According fo: foe invention ABD means a site, moiety or domain capable of binding to circulating albumin "in·· vivo and thereby conferring transport in the circulation of the ABD and any compound or moiety bound to said ABD, ABD’s are known In the art and have been shown to bind very tight to albumin so a compound comprising an: ABD bound to aibumin will to a: Pertain extent behave as a single molecule. The Inventors have realized by using the variant albumin or fragment thereof according to the Invention together with a compound comprising a pharmaceutically beneficial moiety and ari ABD makes it :possible to alter the plasma halfoife of the compound comprising a pharmaceutically beneficial moiety and an ABD compared to the situation where said compound were Injected as such in a patient having heed thereof or administered in a formulation comprising natural albumin or a fragment thereof.

The variant albumin or fragments thereof, conjugates comprising vacant albumin or a fragment thereof or fusion polypeptide comprising: variant albumin or a fragment thereof,: or an associate comprising variant albumin or a fragment thereof according' to the invention may also be incorporated into nano-· or microparticles using techniques well known within the art. A preferred method for preparing nano- or microparticles that may be applied to the variant albumins or fragments thereof according to the invention is disclosed in WO 2004/97153®. which is incorporated herein by reference.

The present invention is also directed to the use of: a. variant of albumin or a fragment thereof or fusion poiy peptides: eorrigrising variant albumin or 'fragments thereof, or a coniugate comprising a variant of aihunlih of a fragment thereof, of an associate comprising: a variant of albumin or a fragment thereof for the manufacture of a pharmaceutical composition, where in the variant of aibumin or a fragment thereof or fusion polypeptides comprising variant aibumin or fragments thereof, or a conjugate:comprising a variant of albumin:or a fragment thereof, gr an associate comprising a variant of aibumin or a fragment thereof has ah altered plasma half-life compared with HSA or. the corresponding fragment thereof or fusion polypeptide comprising HSA or fragment1 thereof or conjugate comprising HSA. in this connection the corresponding fragment of HSA is intended to mean a fragment of HSA that aligns with and has seme number of amino acids as the fragment of the-variant-albumin with which it is compared. Similarly the corresponding fusion polypeptide comprising HSA or conjugate comprising HSA is intended to mean molecules having same size and amine acid sequence as the fusion polypeptide of conjugate comprising variant aibumin. with which It is compared.

Preferably tire variant of aibumin or a fragment thereof or fusion polypeptides comprising variant aibumin or fragments thereof, fragment thereof, or a conjugate comprising a variant of aibumin or a fragment thereof has a plasma half-life that is higher than the plasma half-life of HSA or the corresponding fragment thereof or fusion polypeptide comprising HSA or fragment thereof.

Afternatlveiy, this may be expressed as the variant erf albumin or a fragment thereof or fusion polypeptides comprising variant aibumin or fragmerits thereof, fragment thereof, or a conjugate comprising a variant of aibumin or a fragment thereof has a KD to FcRn that is lower that the corresponding KD for HSA or the corresponding fragment thereof or fusion polypeptide comprising HSA or fragment thereof. Preferably, is KD for the variant of albumin dr a fragment thereof or fusion polypeptides comprising variant aibumin or fragments thereof, fragment thereof, or a conjugate comprising a variant of albumin or a fragment thereof less than 0.9X KD for HSA, more preferred less than G.5X KD for HSA, more preferred less than 0.1 X KD for HSA. even more preferred less than Ο,ΘδΧ KB for HSA, even more preferred less than 0.02X KD for HSA and most preferred less than 0.01 X KD for HSA.

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

The present Invention Is further described by the following examples that should not be construed as limiting the scope of the invention.

Examples IVIaferlais and lief hods ELISA:

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 foom temperature;. The. wells were then washed four times with P:B8/0,005% TWEEN® 20 (P88/T) pH 6:0 before giutaihione'S-transferase (GST) -fused (fshFcRn (0,8 pg/mi) as described id PEBS J. 2008 Aug;278(10):4007-110, pre-lnoebaled with an horseradish peroxidase (HREj-cenjugated polyclonal anti-GST from goat (1 :5000; GE Healthcare), diluted in 4% skimmed milk PBS/D.005% TWEEN® 20 (PBS/T) pH 6.0 was added to each well and incubated for 1.5 h at room temperature followed by washing four times with PBS/T pH 6.0. One hundred pi of the substrate totrameihyibenzidino (TMS) (Calbfochem) was added to each well and incubated for 45 min before 100 pi of 0.25 M HCi was added. The absorbance was measured at 450 nm using a Sunrise TECAN spectrophotometer (TEGAN, hlaennedorf, Switzerland).

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

Surface PSasmon Resonance (SPR): SPE experiments were carried out using a Biacore 30QQ instrument (GE: Healthcare), Flow cells of CMS sensor chips were; coupled with shFoRn-GST (-1400-5000RU) using amine coupling chemistry as described in tics protocol provided by the manufacturer, The coupling was performed by injecting tOgg/ml of the protein in 10:τ·Μ sodium acetate pH 5,0 (GE healthcare). Phosphate buffer (67mM phosphate buffep 0.15M Had, 0.005% TWEEN® 20) at pH 6.0) was used as running buffer and dilution buffer. Regeneration of the surfaces were achieved using injections of HBS-EP buffer (0.01 Al HEPES, 0.15M NaGi, 3miV? EOTA, 0.005% surfactant P20)vat pH 7.4 (Biacore AB), For binding to immobilized shFcRn-GST, 1.0-0.5 μΜ of each HSA variant was injected over the surface at constant flow rate (40 ul/m!) at 25 G. In ail experiments, data was zero adjusted and the reference ceil subtracted. Data evaluation was performed using BlAevaiualion 4.1 software (BiAcore AS).

The same SPR assay was repeated with HBS-EP buffer pH 7,4.

Per the purpoees of this patent unless otherwise stated HSA, WT HSA, rHA refer to Recombinant human serum albumin commercially available under the registered tradename RECGMBUMIN (available from Novoxymes Biopharma UK Ltd, Nottingham UK) was used for the examples.

Serum albumin from other species: The albumins ware recombinant whores stated, produced using sequences provided from publicly available databases, Or purchased fern commercial suppliers.

FeRn Expression and purification of soluble Human (shFcRn) and Mouse (smFcRn) FcRn : Methods for the generation of shFcRn and srnFcRn expression plasmids, expression and purification of each heferodirner can be found in Berntzen et al (2005) J. Immunol. Methods 208:03-104), Alternatively ShFcRn FcRn heterodimer was produced by Gene Art AG (Germany). Sequences for the two sub uhifsibf th#. heiecM»lpr can be found in SEQID NO: 3 and SEG ID NO: 4. The soluble receptor was expressed In HEK293 cells and purified from culture supernatant using NI-HiTrap chromatography columns*

Example 1, (Preparation of variants

Preparation of specific HSA muteln expression plasmids

Methods for the expression of HSA mutant variants and HSA fusion variants were produced using several techniques. Standard molecular biology techniques were employed throughout such as described in SambrooiA 3- and D. W. Russell,: 2001, Molecular Cloning; a laboratory manual, 3rd ed. Cold;Spring Harbor Laboratory Press, Cold Soring Harbor, N.Y.

Method 1. Amino acid substitutions In HSA detailed in Table)

Synthetic DNA feol/SacI fragments (859 bp) were generated by gene assembly (GeneArt AG. Germany) containing point mutations within the HSA-eneoding gene (SEQ 10 NO: 1) to introduce the desired amino acid substitution in the translated protein. 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 amino acids (Le, wild type) was Identical to that in pQB2243 (described in WO 00/44772). The synthetic nucleotide fragments wore ligated into /V'coi/Sachdigested pD82243 to produce piasmids pDB3876 - pDB3888 (Table 1). For the production of expression plasmids, pPB3876 pDB3888 (see Table 1) were each digested With Νοά and the DM A fragments were separated through a 0.7% |w/y)::¥ae gel, and 2992bp fragments (Wofi cassettes’ including PR 81 promoter, ON A encoding the fusion leader (FL) sequence (disciosed in WO 2010/092135), nucleotide sequence encoding HSA and ADH1 terminator; see Figure 1) were purified from the agarose go! using a Giageo Gel Extraction Kit Mlowing the manufacturers Instrudions, Wfi cassettes’ were ligated into a RM/Shrimp Alkaline Phosphatase (Roche)- treated "disintegration” plasmid pSAC35, disciosed in EP-A-288 424 and described by Sleep, D., &amp;t aL (1991) Blo/Teehnoiogy 9, 183 - 187. Ligation mixtures were used to transform chemicaliy-competent £. coil DH5a, Expression plasmids pDB3S87 - pDB3S97, pSAC35-<terivafixes containing the ’Moil cassettes”, were identified using standard techniques. Disintegration plasmids pDS3887 -- pDB3897 and pDS224:4 (For the expression of wild type HSA, described in WO 00/44772) (Table t ) were used to. transform £, ceraws/ae ΒΧΡ10ο!£ (as previously described WD/2001 /079480 as described below.

Table 1: Plasmid, amino add substitution introduoed Into HSA

* , :·

n/a ~ Not applicable. pDB38?:6“pOB3886 are sub^cfodlng piasmids-Table 2: Codons used to introduce amino add substitutions inb.H&amp;A

ilethod 2. Production of HSA variants D994N*E495CHT496A and £495Q±Τ.4ΜΛ A PCR-based method, using a QuickChange Lightening Kit (Statagene), was employed to introduce pdiht mutations into HSA, Oligonucleotide pairs xAPQ94 (SEO li^ NO: S)/xAPQ96 (SEQ ID NO: 8) and xAP896 (SEQ1B NO: 7pAP097 (SEQ ID NO: 8} were used to generate two HSA variants (D494N+E49SQ+T498A and E49S0+T496A, respectively). Plasmid pD83927.(disdosed in WQ 2013/092135) was used as template DNA and the methodology recommended by the manitfacterer of the kit was followed. The resulting plasmids were named pDB3995 and pD8399e (contain HSA D494N+E4950-91498A and E495Q+T496A expression cassettes, respectively), pD83995 and pD83996 were digested with BsMWBsrBl and the linearised DMA molecules were purified using standard techniques. One hundred ng of each SsfEli/BsrBi digested DMA, purified using a Qiagen PCR~Punfication kit following the manufacturer’s instructions, was mixed individually with 10C)ng Acs6Si/8smHIndigested pDB3936) (disclosed in WO 2010/092135) and used to directly transform: S. cemvisiae BXPIOcir0 using the Sigma Yeast Transformation kit described below.

Method 3, Amino acid substitutions in HSA detailed in Tabie 3

Plasmid pDB392? (disclosed In WO 2010/082185) (containing an identical nucleotide sequence encoding HSA as in pDB2243) was manipulated: to amino acid substitutions within the mature HSA protein. Synthetic DMA fragments were generated (GeneArt AG, Germany or DNA2,b Inc, USA) (NcollBmjMk AvdiiSphl or SaciiSphl fragments}* containing: point mutations within the HSA-encoding gene to: introduce the desired amino acid substitution's): into the translated protein sequence. Table 2 details the codons used to Introduce the amino acid substitutions into the HSA-encoding gene. The nucleotide sequence of the synthetic Pagmentenooding unchanged amino acids: (ke, wild type)(was(identical to those in pDB3927. Synthetic DMA fragments were sub-cloned infoJycel/SsoSeij AvAUSphl·, SacliSph -digested 0DB392? (described In PCI 11 527.204-WO) to generate pDS40C6~pD84010, pDB4083-pD841D;1 and pD84103-pD84111 and pD84194, 0884200:,0084202 (see Table 3).

Sirrillariy, BamHUSaH fragments containing point mutations in the nucleotide sequence encoding HSA were generated by gene assembly (PMA2.0 inc, USA) and ligated into SamHi/Sa/i-digested pDB3964 (described in WO 2010/092135} to produce plasmids pD83888~pDB3iS9 {Tabie 3).

The C-lerminal string of amino adds from position 673-585 (KKLVAASQAALGL) (SEG ID MO: 9) in HSA were mutated to those in macaque (PKFVAASQAALA) (SEQ ID NO: 10), mouse (PNLVTRGI4DALA) (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 amino acid substitution are given in Table 2, Synthetic DMA fragments (Saci/SpPI) were generated (PNA20 Inc. USA) by gene assembly'{the nucleotide sequence of the synthetic fragment encoding unchanged amino acids (i.e. wild type) was identical to that In pD83927) and were sub-cloned into SadiSphi-digested pDB3827 to produce plasmids pD84114-4117 (Tabie 3).

Plasmids pD83883 (Table 1), pDB4094 and pD84095 (Table 3) were digested with Ncol/Sacl and 85?bp fragments from each digest were purified before being ligated into Ncoi/Sael-digosted pDB4008 Or p084110 (8.888kb) (Table 3} to produce pDB41S6-p0B4101.

Expression plasmids were generated in vim (i.e. via homologous recombination in S. cerevisiaai a technique referred to as gap repair or in vivo cloning- see Οπ-Weaver A Szosiak. 1983. Proc. NatL Acad. Sci. USA. 80:4417-4421}. Modified piasmids listed in Table 3 were digested with SsO/BsrBI and the linearised DMA molecules were purified using standard techniques. One hundred ng of each BstEil/BsfBi digested DNA, purified using a Qlagen PCR-Purilication kit following the manufacturers instructions. was mixed individually with tOOfig 4cc86l/BamHI-digested pOB3938 (disclosed in WO 2310/092135) and used to directly transform S. cerems&amp;e SXP10cirf! using the Sigma Yeast Transformation kit described below.

Table 3,

Table 4: KSQD primers and plasmids

Table δ: K573 primers and plasmids

Method 4. HSA K5Q0 and K573 permutation library PC ft was used to produce two permutation libraries in which the codons encoding amino add 500 or 573 of mature HSA were changed (mutated) to alternative non-wiid type amino adds and a termination codons (K5XXSTOP). Mutagenic oiigonudetides (Table 4 and Table 5). were designed to amplify HSA- encoding DMA and incorporate the desired changes, That is, for the changes at position 500, pDB4082 (Figure 1) was used as a template DNA. pDB4082 is a derivative of pDB2305 (disclosed in EP1788084) and was produced as follows. pDB230S (Figure 2) was digested with NsiliSpel and the yielded 8.779kb Nsi) fragment was seif-iigated to produce pD84005 (Figure 3). A synthetic DNA fragment (Ssal/S'phi) was generated by gene assembly (DNA2.0 Inc, USA) (SEQ ID NO: 1) (containing 3' region of the PR81 promoter, modified fusion leader sequence, nucleotide sequence encoding HSA and 5’ region of the modified ADH"\ terminator), and ligated into H/ndiil/Sp&amp;l-digested pDB4O05 (Figure 3) to produce pDB4082. Note, The MedlSi site in PR81 promoter site has been removed and a Seell site within the nucleotide sequence encoding HSA has been introduced,

For the permutation library for position 600 of HSA, the nucleotide sequence encoding HSA corresponding to that between the Sail / HinM sites (see plasmid map pDB4082, Figure 1) was generated using the New England Bioiabs Phusion kit (Table 6) and oligonucleotides listed in Table 4. Table 7 describee the PGR method employed.

The permutation library at amino acid position 573 in HSA was generated using pD83S27 as template DNA and involved amplifying the albumin-encoding DNA corresponding to that between tie Nco\ and Bsu36l sites using oligonucleotides detailed in Table 5,

Table 6:. PGR ingredients

Table 7: PCR conditions:

For the albumin variants based at positions 500 and 573, each PCR-prcduct was purified using a Qiagen PCR-ciean up kit (according to the manufactures instructions), digested with SalVHindWi (position 500 library) or Afcol/8su38i (position 573 library). The digested DN As were then purified using a Qiagen PCR-ciean up kit and ligated into Sa/i/ HhuM ~ or Nco\iB$uZ§\ -digested pDB4082 or ρΰ&amp;3927, respectiveiy, replacing the eguivaient native sequence. Ligations were transformed into E. coli DHSa, subsequent plasmids isolated from transformants using a Qiagen minlprep kit (according to the manufacturers instructions) and the correct; constructs identified by restriction analysis. This produced a collection of plasmids. pD84204 -- pDB4222 (position 8d0 library) p;DB4173 to pQB4190 (position 573 library), containing albumin genes which differed' only in their sequence corresponding to the codpn fpr the amino dcid at position 500 or 573 Table 4 and 5, respectively). Thespedfle ehangesin each plasmid wore confirmed by sequencing.

The resultants plasmids wdre used to, generate expression plasmids and albumin fusion producing yeast· by m vryp cloning as described above. That is, cefbwsfp© was tmnsformed using the Sigma Yeast; Transformation kit (described below). Using a mixture of a 100 ng BstEiUBsrS 1 -diaeste HSA variant containing plasmid and 100 ng AcCSSl/BamHI digested pBB3938,

Transformation of 8, cerevtssae 8, cBXP10 oids (as previously described tA(O/2001/079480) or Strain A ctH (described in VV(®/20(5B/06.17i 8) was streaked on to YEPQ: plates (1% (w7v) yeast extract, 2% (w/v) Bactopeptone, 2% (w/v) glucose), 1.5%: agar) and allowed to grow for 4 days at 3(ΓΒ prior to transformation. One gg of whole plasmid (i.e, circular plasmids) or, for gap repair,: 100 ng BStEII/SsrSI- or -digested HSA variant or HSA vacant fusion containing plasmid and 100 ng AccSSilBamHI digested pDB3S36 were used to transform &amp; mmim® using a Sigma Yeast Transformation kit using a modified lithium acetate method (Sigma yeast transformation: kit, YEAST-1, protocol 2; Ito ©fal (1983) a. BacferioL, 153, 16; Eibie, (1992) Biotechniques, 13, 18). The protocol was amended slightly by incubating the transformation at room temperature for 4 h prior to heat shock, Following heal, shock, the cells were briefly centrifuged before being resuspended in 200μΙ 1M sorbitol then spread over 8MMD agar plates, the composition of BMMD is described by Sleep et a/., (2001), Yea si 18, 403. Plates were incubated at 30'C for 4 days before individual colonies were patched on to fresh BMMD plates. Yeast strain numbers are detailed in Table 1.

Stocks were prepared for each yeast strain as follows: SMMD broth was inoculated with a heavy loop of each yeast patch and grown for 24h at 3CTC with orbital shaking at 2G0rpm. Cells were harvested by centrifugation at 1300 * g for S min In a Sorval RT600 centrifuge, 1 SmL supernatant was removed and replaced by trehalose 40% (w/v). The ceils were resuspended and transferred to cyrovials (ImL) tor storage at ·8ϋ:Ό.

Shake flask growth of c&amp;evisme BMfvID (recipe Q<1?% (w/v) yeast nitrogen base without amino acid and ammonium sulphate (Difeo), 37.8mM ammonium sulphate, 29mM citric acid, 142mM dlsodium hydrogen orthophosphate dehydrate pH6.5, 2% (w/v) glucose) media (lOmi) was inoculated with each yeast strain and grownfdr 12h at 30®C with orbital shaking at 200rphv An a lip uot of each starter culture prut) was used to Inoculate 2 * £00mL BfvfMD media and grown for 36h at 30*€ with orbital shaking at 20brpm. Cells were harvested by filtration through 0.2pm vacuum filter membranes (Stericup, yillippre) including a GF-D prefltter (Whatman) and the supernatant retainedfor purification.

Primary concentration

Retained culture supernatant was concentrated using Tangential Flow Filtration using a Pail Filtron IV system fitted with a Omega 10KD (0Xi83sq,m2) filter (t¥ CentramateTM cassette, Pall Fittron) with a transmembrane pressure of 20psi and a recirculation rate of liOmb.mln1,

Fermentation

Fed-batch fermentations were carried out in a 10 L Sartorlus Biostat C fermenter at 30:>C; pH was monitored and adjusted by the addition of ammonia or sulphono acid as appropriate. The ammonia a iso provided the nitrogen source for the cultures: The level of dissolved oxygen was monitored and linked to the stirrer speed, to maintain the level at >20% of saturation. Inocula were grown In shako flasks In buffered minimal media (recipe). For the batch-phase the cultures was inoculated into fermenter media (approximately 30% of the fermentbr volume) containing 2% (w/v) sucrose. The feed stage was automatically triggered by a sharp rise in the level of dissolved oxygen. Sucrose was kepi at growth-lirnitibg: concentrations by controlling the rate of feed to a set nominal growth rate.: The food consisted of: fermentation media containing 50% (w/v) sucrose, all essentially as described: by Goins. (Collins,, S,H.:, {1980): Production of secreted proteins in yeast, In: TJ.R. Harris (Ed,) Pmtein production:by biotechnology, Elsevier, London, pp, 61-77). GP-HPLC quantitation

Purified albumin variants, fusions and conjugates were analysed by GP-HPLC and quantification as follows, injections of 25pL were made onto a 7.8mm id χ 300mm length TSK G300QSVviXL column (Tosoh Bioscience), with a 6.0mm id x 40mm length TSK SW guard column (Tosoh: Biesckance}, Samples were chromatographed in 25mM sodium phosphate, 1Q0mM sodium sulphate, 0.05% (w/v) sodium azide, pH 7.0 at Imt/min, Samples were quantified by UV detection at 280nm, by peak area, relative to a recombinant human albumin standard of Known concentration (lOmg/mLlund corrected for their relative extinction coefficients.

Purification of albumin variants from shako flask

Albumin; variants were purified from shake flask (either culture supernatant or concentrated culture supernatant) using a single chromatographic step using an albumin affinity matrix (AihuPure™ - ProMetic BioSeiences, Inc.), Chromatography was performed at a constant linear velocity of 240cm/h throughout. Culture supernatant was applied to a 6cm bed height, 2.0rnL packed bed prereguliibratod Vyith 50mM sodium acetate pHSJ, Following load the column was washed with 10 column volumeTQV) of eguiiibration buffer, then 50mM ammonium acetate pH8.0 (fOGV); Product was aimed with either 60mM ammonium acetate 10mM octanoafe pH8.0, 50niM Ammonium Acetate 3omM Sodium Octanoafe 2Q0mM Sodium Chloride pH7.0 or 200mM Potassium thiocyanate, The eeiumn was cleaned with Q.5M NaOH (3cv) and 20mM NaOH (3.5cv). ESuafe fraction from each albumin Variant were concentrated and diaflitered against ID volumes of 50mM sodium chloride |Vivaspin2D 10,000 MW CO PES with optional diafiiiration cups, Sartorius). Purified albumin variants: were guantified: by GP-HPLC as described above.

Purification of albumin-fusion variants from shake flask

AlhuminTuslen variants were purified from shake flask: culture supernatant using a single chromatographic step using an aibumln affinity matrix (AlbuPure™ ~ ProMetic SioSolences, Inc,), Chromatography was performed at a constant linear velocity of 240cm/h throughout. Culture supernatant or concentrated culture supernatant was applied to a 8cm bed height, 2.0mL packed bed pre-equilibrated· with SOmM sodium acetate pH5,3. Following toad the column was washed with 10 column volume (cv) equilibration buffer then 50mM ammonium acetate ρΗβ.Ο (lOcv). Product was eluted with either SOmM ammonium acetate 10mM octanoate pNS.Q, 50mM Ammonium Aoretate 39m!V1 Sodium Octanoate 2Q0niM Sodium Chloride pH7.0, 50 mM Ammonium Acetate IDOmM Sodium Octanoate pH9.Q of 2D0mM Potassium thiocyanate. The column was Cleaned with 0.5Μ NaOH |3ev] and 20mM NaOH (3.5cv), Eiuate fraction from each albumin variant-fusion were concentrated and diaflitered against 10 volumes ot 25 mM Trie, 150 mM NaCI, 2 mM K'Ci, pH 7.4 (Vivaspin2G 10,000 MWCD PES with optional diafiltration cups, Sartdrius), Purified aibumin-fpsion variants were quantified by GP~HPLC as described above.

Purification of albumin variants from fermentation

Albumin variants were purified from high ceil density fed batch fermentation supernatants after separation by centrifugation, using a Sorvafi RC 3C centrifuge (DuPont), Outturn supernatant was chromatographed through an 11cm bed height column 8.8ml packed feed packed with a custom synthesised albumin affinity matrix (AibyPure™ - Profeteie BidSciences, Inc.) as described above, Product was eluted using elution buffers describe above at allow rate of 120cm(H« The eluate fraction(s) was analysed by ©P-HPLG, (above) .and fecludftg SDS-PASE for purity: and If required concentrated: (Vivaspin2010,008 MWCQ PES) and applied to a 24x96om column packed with Superdex ?S run at a flow rate of 39cm/h in 25 mM Tris, 150 mM MaGi, 2 mfe1 KCi, pH 7A . The peak was fractionated, assayed by GPHPLC and pooled in order to generate the monomeric protein of interest. Pooled fractions were concentrated (Vivaspin20 10,000 MWCO RES, Sartorius).

Ali proteins to be assayed for receptor (FcRn) binding properties and or other analysis were quantified by GP HPLC as described above corrected for their relative extinction coefficients.

Example 2. Determination of receptor fshFcRn) binding properties of bipod derived HSA and recombinant human albumin

Essentially fatty acid-free HSA (Sigma-Aid rich) was further purified by size exclusion chromatography as described in Andersen et a/ (2010). J.Bioi.Chem. 285, (7)4826-4836. Ten of monomeric HSA and rHA were analysed using SPR as described above and the data presented in Figure 4,

Direct comparison of HSA (blood derived) with recombinant human aibumin (Recornbumin) at the same concentration (10μΜ) (F!gure 4A and 48) shows for both samples binding to immobilized shFCRn (pH8.0, pH7.4 respoccvely) was reversibie and pH dependent, in addition, comparison of HSA vs recombinant Hyman albumin by Bbsse ef a/ (2085)., 3. Clin, Pharmacol, 4S:; 57-67, demonstrated equivaient half life te 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 major differences between the assays: In the ELISA system HSA is coated direotly in weils and sHFcRn-GST is added in solution whereas in the SPR assay shFcRn-GST is immobilized to a Ofv15 chip and HSA 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 am disclosed in Figure 5. The ELISA values represent the mean of duplicates.

The variants were analpeij using SPR analysis at pH 6.0 and pH 7)4. Results are disclosed for a representative numberof variants in Figure 0 using a concentration of the variants of 0.2 μΜ and 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 of variants at each concentration is shown in Figure 8 A and B respectively.

The conclusions of the analysis are that ai! tested variants have the oharacteristlc binding to the receptor at pH 6,0 but no binding at pH 7,4. The variants P494N,Q,A, E495CkA> Τ4Θ6.Α, and D494N+T406A show reduced binding to the receptor compared to .'HSA,

Example 4, Determination of receptor fshFcRn/smFeRn) binding properties Of albumin variants USing the SPR analysis method below the association constant Ka the dissociation constant Kd and the binding constant KD calculated for HSA and mouse serum albumin (MSA) binding to human and mouse FcRn {Table 8). $PR analyses- SPR analyses were performed on a BIAoere 8000 instrument (GE Healthcare) using CMS chips and immobilization of smFcRn-etBT end ShfcRn-GST variants or srnPcRn was performed using the amihe coupling kit (GE Healthcare). Protein samples (10 pg/mi) were injected in 10 miVi sodium acetate at pH 4.5 (GE Healthcare), ail as described by the manufacturer. Unreacted moieties on the surface were blocked with 1 M eihanoiamine, For all experiments, phosphate buffer (87 mM phosphate buffer. 0H6 M NaCi, 0,005% TWEEN® 20) at pH 6.0 or pH 7.4, or HBS-P buffer (0.01 M HEPES, 0.15 M NaCI, 0.005% surfactant P2G) at pH 7.4 were used as running buffer or dilution buffer. Kinetic measurements were performed using a iow density immobilized sudace (100-200 resonance units (RU)). Serial dilutions of higG1 (2000.0-31.2 nM). migGI (1000.0-15.6 nM), MSA (20.0-0,3 μΜ) and HSA (200.0-3.1 μΜ) were injected at pH 6.0 or pH 7.4. at a flow rate 50 μΙ/minuie at SSEC. Additive binding was recorded by injecting HSA (10 μΜ). MSA (δ μΜ), higGI (100 nM) or migGI (100 nM) aione or two at a time at 2530 at 20 pi/minute at pH 6,0 over immobilized shFcRn (-600 RU) or srnFcPn (-600 RU). Competitive binding was measured by Injecting shFcRn (50 nM) or smFc-Rn (100 nM) alone or together with different amounts of HSA or MSA (10.0-0.05 μΜ) over immobilized HSA (-2600 RU) or MSA (-2000 RU). in ail cases, to correct for nonspecific binding and bulk buffer effects, responses obtained from the control surfaces and blank Injections were subtracted from each Interaction curve. Kinetic rate values were calculated using predefined: models (Langmuir 1:1 ligand model, heterogeneous ligand model and steady state affinity model) provided by the SiAevaluation 4,1 software. The closeness of the fit, described by the statistical value that represents the mean square, was lower than 2.0 In ail affinity estimations^

Table 8: Binding constants of HSA and MSA shFcRn and snlFeRn,

The KD's were generated using the SIAevaludiian: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: Net acquired

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

Commercially available animal albumin (either Sigma-Aldrich or Calbiociiem) were further purified as described in Andersen ef ai |2010). J.BiohChem, 285, (7},4826-4836, The binding of donkey serum albumin, bovine serum albumin, goat serum albumin, sheep serum albumin, rabbi serum albumin,: dog serum albumin, hamster serum albumin, guinea pig albumin, rat serum albumin and: chicken serum albumin to shFcRn was determined using the techniques described in Materials' and Methods,, The ELISA results are disclosed in Figure 9 A-D and the relative bindings summarized in Figure § E-

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

Table 10: Cross-species aiburoin-FcRn binding

Relativebinding responses am categorized from strongest (++++) to weakest {+} and no binding (- h a: Not determined (ND). A hierarchy ranging from strongest to weakest binding is as 'follows; guinea pig ~/> rabbit > hamstebdog > rabmouse > donkey > human > bovine > goal/sheep > chicken. This data shows that animal albumins have different affinities for siiPcRn,

Example 8, Kinetics of the HSA variant for shFcKn

The binding constants for variants according to the invention were determined according to the methods described in Materials and IVtethods,

Table 11: Binding constants of HSA variants for shFcRn

The KO's were generated using the BiAevatualion 4.1 softwares A Langmuir 1:1 ligand mode! was used throughout. The kinetic values represent the average of tripilcates, NO means: Not determined.

The results correspond with theconclusions made in Exampie 3 based on SPR and ELISA databut In addition shows that E492I3 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; was performed using the methods described In example 4, Results are shown In Figure 15¾

The results show that the variant Ε492Θ, unlike E492H E492P and E492G+V483P, has stronger binding to shFcRn than HSA..

Example 8, Analysis of G41? substitutions

Using the: method of Exampie 1 variants of HSA having the substitutions Q417A and D494£*Q41?H were constructed. The kinetic properties of these variants were tested using the methods In Materials and Methods: and are shown in Table 12.

Table 12: Binding constants of HSA variants for shFcRn

a: Dilutions of HSA variants were injected over immobilized shFcRn .{H500 R.U). b: :theAinsite.raiecponstanis· were obtained using a simple first-order (1:1) bimolecuiar interaction model, o: The steady state affinity constant was obtained using an equilibrium (Req) binding model supplied by the BiAevaiuation 4.1 software. The kinetic values represent the average of triplicates, d: Not determined (ND).

The data show that variants Q417A and D494E*G417H bind weaker to the receptor than the wild-type HSA,

Example9, Analyst of HSA variants in position 499, §00, 536, 537, §33 and §73

Using the method of Example 1 variants of HSA having the substitutions P49BA, KSO0A, K536A, PS37.A, K538A and K573A were constructed. The receptor binding properties of these variants were tested as described in Materials and Methods. Results are shown in Figure 11.

The data demonstrated that variants P499A, K538A, P537A and K538A had a reduced binding affinity to shFcRn relative to HSA, Variant KS00A had almost completely lost its ability to bind to shFcRn and K573A had an increased binding affinity to shFcRn both relative: to HSA.

Example 10, Analysis of variants in position §01 of HSA

Using the method of Example 1 variants of HSA having: the substitutions ES01A and ESO IQ were constructed,: The kinetic properties of these variants were tested as described in Materials and Methods,

Table 13 Binding constants of HSA variants for shFcR

a; Dilutions of HSA variants wore injected over immobilized shFcRn (-1500 RU). br ibe kinetic rate constants were obtained using a slmpie first-order: { 1:1 yfeimotequlaf Interaction model. o; The steady state affinity constant was obtained using an equilibrium (Reg) binding model supplied by the BiAevaiuation 4,1 software, Tire kinetic values: represent th§-avera^0f^||^al@i§ d: Not determined :ND) fhe$^-;#iqws that Vigi&amp;ntsi ESOI A and E501Q:haves slightly decreased binding affinity to shFcRn relative to H8A,

Example 11, Analysis of HSA variants m position 573

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 of these variants were tested as described in Materials and Methods :bat with SPR analysis, performed at pH5.5. Results are shown In the table 14 below and Figure 12 and 13. .T .. 1 fΛ..: Ja.i ί! ^. ί?ί. . Λ?. § ϊ.0.9JP Λ..........................................................

a: Dilutions of H8A variants weretlnjected overimmoblltzed shFoRn (H600 RU). h: the kinetip rate constants were obtained using a simple first-order (i :1} blmoieeyiar interaction model, c: The Steady state affinity constant was obtained using an equilibrium (Reg) binding model supplied by the BIAevaluafiob 4,1 software. The kinetic values represent the average of duplicates, d: Not determined (NO).

The results show that all variants having substitution in position 573 have improved binding to sbFoRn compared with Wt HSA, in particular the-variants K573F, K§73HS K673P, K573W and K573Y have more than 10 fold. lower KD to shFc-Rn than the parent HS.A. The variant K5738TC5P is a truncated albumin having a stop codon in position 573, The sensorgraro for the K5738TQP variant show significantiy reduced binding compare to the WT HSA and generated a high KD. The increased affinity that we have-shown for the variant K573E, a-natural variant characterized by Dtagiri (2009). BIoLPharm. Bull. 32(4} 627*634,is: predicted to have increased haif-iife in vjm.

Example 12, Analysis of fu rther HSA variants

Using the method of Example 1 variants of HSA having the: substitutions E492i3;: E492GfN503H:( N503H, D550E, E492G+N503K, E542P, MA40Q, K541G, K541D, D550N Ε492@*Κ538Μ+Κ541 N+E542D, E492T+N503K+K541A, E492P*N503K+K541 G+E542P,

E492H+E5O1P^NSO3H^E5O5D+T508S+T54OS«41E< A48O0+E492T+V493L*E3O1P +N5O3D+A504E'rE50SK:'f-T5O6F'rK541 D, E492G+V493P+K538H+K541N+0542D were constructed. The receptor binding properties of these variants wore tested as described in Materials and Methods, and the results are shown in Tabid 15 and Figure 14,

Tabie 15: Binding constants of HSA variants for shFcR

a: Dilutions of HSA variants wore injected over immobiiized shFcRn p-1500 RU). h: The kinetic rate constants were obtained using a simple first-order (1:1 J bimoieduiar iniemofion model, c; The steady state affinity constant was obtained using an -equilibrium (Reg) binding model supplied by the BIAevaluation 4,1 software. The kinetic: values represent the average of triplicates, d: Not determined |ND),

The results show that for position 560, a substitution to Έ results in an increased affinity whilst a substitution to- bASdoced affinity for shFcRn at p.H8,0, When this analysis was repeated 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 the substitution for an uncharged amide amino acid reduces binding at pH6,0, Based on this observation, we would-predict for this position that substitutions to baslcfamlno 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: H44GG. B484Q, HS10Q and HS35Q,. Figure :15 Shows SPR sensorgradiS: Of those variants interacting with shFcRn as described in Materials and Methods..

It was found that the variant H440Q bound with comparable affinity as HSA. in contrast H484Q, H510Q and H535Q had significantly reduced affinity to shFcRn. This supports the previously published observations that mutagenesis of these Histidine residues significantly reduced HSA binding to shFcRn (Wu ef a/ (2010). PEDS,23(10)789-798). Wu et ai show a reduced half-life for a.diabody fusion proteins (scFv-DIH)2 in mice with an order of removal from slowest to fastest; Dh-Glij WT>H535A>H510 A>B464A>Db. Based on affinity to shFcRn and when compared to smFoRn (example 5) we would predict the clearance order in humans to be (for glutamine {©) substitutions) W?> H440Q>H510Q>H464Q>H535Q,

Example 14. Further variants

The following variants were generated using the methods described in Example 1-: K5T4N and OS80R. in MSA. Binding of the variants to FcRn was tested using the SPR assay as described in Materials and Methods and the results are shown in Table: 16,

The results show that variants RS74N and QS80K hpurto stronger to shFcRn,

Tablel6: Following kinetic data was found for these variants:

Example I S, Analysis of HS A variants In position 500

Using the method of Example 1 variants of HSA having a substitution at position 500 were constructed. AH variants at position 500 were generated and the receptor binding properties of these variants were tested, Biacore X, Blaeore XiOO and Sensor Chip CMS were used for ail analyses, both supplied by G E Healthcare, shFcRn produced by GeneArt AG {Germany} (diiuted to lOpg/mL So 10mM sodium acetate pH5.G (G E Healthcare)} was immobilised on flow eel! 2 (FC2) to levels between 1600 - 2200 response units (RU) via standard amine coupiing as per manufacturers instruoions (G E Healthcare). A blank immobilisation was performed on flow ceil 1 (FCi) for it to serve as a reference cell- To stabilise the assay, 3-5 start up cycles were run first, with running buffer (6?mM phosphate buffer, 0.15M NaGL 0,005% Tween 20 at pH5.7S ±0,25) only, followed by regeneration, WT rHA and K500 library variants were Injected at various concentrations (ipM - fbOpiVt) for 80s at a constant How rate of (SOpi/min) at 25 ‘0 followed by regeneration of tee surface using HBS-EE buffer pH7.4 (G E Healthcare) until approximate initial baseline RU was restored (usually 12s pulse would suffice).

Results am shown In the Table 17 and Figure 16 Table 17: Kinetics Of HSA K500 single point mutants.

1 gi Mean of 4 values. b: The kinetic rate constants were obtained using a simple first-order (1:1) himoleculaf interaction mode!. c: The steady state affinity constant was obtained using an equilibrium (Req) binding model supplied by the SiAevaluation 4.1 software.

The resuits show for variants K6Q6R and RSOOI have increased and comparable affinity for sliPcRn compared to WT HSA respecriveiy. Variant K5G0E bound rightly to immobilised shFcRn but still demonstrated the cbaracteristic pH-dependenCy c4 the FcRn interaction. This complex was very stable, such that kinetic analysis vfas not possible (Figure 16). Ail other variants have reduced binding to shFcRn than wt fHA<

Ail 'Variants bound to shFcRn (to some extent) at pH5.5. No binding of K500 library variants to shFcRn was detectable at pH7,4.

The generation of albumin fusions containing albumin muteins

Plasmids containing expression cassettes for the production of scFv (vHvL) genetically-fused lO: HSA, at either the :N* or C-terminus or both, (described in, Evans &amp;t al.. 2010. Protein Expression and Purification. 73,113-424} were modified to allow the production of albumin fusions using m vivo cloning ...(describe above). That is, pDB30i7 (Figure 1?)ypDB3021 (Figure 18), pDB3056 (Figure 19} were digested with NsiVSpeI and Naif fragments corresponding 6.611¾ g,S89kb and 8-795kb, respectively, were purified using standard techniques. Purified Nsil fragments were self-ligated and used to transform chemically competent E mil OHb« to produce pDB4168. ODB4169 and pQB4170, respectively (Table 18),

Svrnlariy, pD83165 (containing the bivalent fusion ) (Figure 20) was digested with NoU and the expression cassette (4.508kb fragment) was purified before being ligated into Nod-digested pDB3927 to produce pDB4172 (Figure 21. Tabie 18).

Synthetic Sai\;8$u3§\ DMA fragments (269bp). which contain point mutations within the albumin encoding nucleotide sequence to introduce amino acid substitutions corresponding to fCSOOA, or D650N or R573P into the trarisiated albumin protein sequence, were generated by gene assembly (GerieArt .AG, Germany). The Saii/Ssi38i fragments were individually ligated Into Sa&amp;BsudBi-digested pDB4168-pDB4170 and pDB4T72 and used to transform : chemically competent E coli DH5a using standard: techniques to generate plasmids pDB4265 - pDB4276 (Table 18).

Table 18: Albumin variant fusions

Similarly* aDNA fragment was generated by PCR (using standard techniques), ίο introduce a K673A substitution in ite translated albumin protein sequence, PCR was performed using the New England Slolabs Pbusion kit using pOB428? (Figure 22} as template DNA and oligonucleotides xAP238 (SEQ ID NO: 53) and XAF239 (SEO ID NO: 54);

Table 19 describes PCR cycling.

Table 19: PCR cycling

The RCR-product was purified, digested with the fragment (269:bp) isolaied was ligated Mo <Sa/l/Sso30i“diqpsted pD34168~pDB41?0 and pQB4172 and used to transform chemically competent £ DHSo, Resulting plasmids (pP84277 - pDB4288} are listed in Table 18.

The nucleotide sequence encoding the FLAG tag; was removed fmm plasmids pD84188 and pDB4268-42 ΙΌ (plasmids for the expression of scFv N~term!naliy fused to HSA and HSA muteihs K500A, D560N and K573P, respectively. pDB4188 and pDB4268-4270 (Table 18) were digested with 8sb38i/Sphi to remove &amp; 231bp product comprising 3’ region of HSAteneeding gene, nucleotide sequence encoding FLAG tag and 5' region of ADH1 terminatdf, A B$u3®\i$phi fragment (287bp}f composing 3' region of HSA-encoding gene and S’ region of mADHI terminator (SRQ 1D1) from: pDR418l was ligated into Rsu36PSpilfedigsstecJ pDB4168 and pt)B4288~pD84270 using standard techniques, Ligati on mixtures were used to transform chemically competent £, mil DH5a using standard techniques to generate plasmids pDB4281-pOB4284 (Table 18} pDB4265~pDS4284 were digested with BsiBWBsiBi and the linearised ONA molecules were purified using etandard techniques, One hundred ng BsiEWBsrQl ONA samples were mixed with lOOftg Ac»6SlfBamHi--digested pDB3938 and used to transform S, cerevisia&amp; BAP 1 Qdih* using the Sigma Yeast Transformation kit described below. In each case the expression piasmid was generated in the past by homologous recombination (te \Mo dontegj between the aibuminduston containing piasmid (pDB4265-pDB428d} (Tabie 18} and pDB393£

Plasmids pDB3017, pDB3021, pDB3056 and pPB3165 (wild type HSA fusions, described by Evans e>( a/., 2010, Protein Expression and Purification, 73,113-124} were used te transform $. mmvisim Strain Acls·0 (described in WG/200S?0i1718) using the Sigma Yeas? Transformation kit described below.

The nucleotide sequence encoding human ILMIRA (intedeukin-l receptor antagonist} (accession number; CAAS9D87) could fee synthetically generated fey gene assembly. The nucleotide sequence of the 708bp synthetic fragment (8su38l/Spbi fragment) is given in SEQ ID MO: 56 and includes the 8'region of the gene encoding HSA, the hudleotide sequence encoding a GS linker, the nucleotide sequence encoding human iL*1RA (N84Q to abolish the N-iihked glycpsylatlon motif) and the 5' region of the ADH1 terminator. The synthetic DMA fragment could be ligated into 6su38i/Sqh!Mgesteet pD83927 to produce pDB268§,

Plasmids containing the expression cassettes for the production of it~1 RA genetically fused to the C-terminus pf HSA and the HSA variants K5Q0A, D550N, K573A and KS73P were prepared as follows. pD82588 was digested with 8su38l/Sphl and a TGSbpIragment containing.the:.=-3 region of the 'HSA encoding gone, nucleotide sequence encoding a SS linker, nucleotide sequence encoding human IL1-RA (N84Q) and the: S’ region of a modified 8. cerevlslae ADH1 terminator (SEQ iD3) was purified using standard techniques then ligated into BsuSdi/SphPdigesied pD840Q8 (containing HSA K573A expression cassette). pD840i0 (containing HSA D560N expression cassette),;, pDB4086 (containing HSA KSOOA expression cassette):» pD84iiO (containing HSA K573P expression cassette | to generate pDS42B7, pOB4288:, pDS42S6 and pDB42B8, respectively (for an example, see Figure 23). pOS4285~pDB4288 were digested with Nsii/Pvui and the linearised DNA rfioieOuSes were purified using standard techniques. One hund;red ng Nsii/Pvui-digested DNA sampies Were mixed With 1 OOng Acc65i/Bam H i-d igested pD83936 (9721 bp) (Le. in vivo cloning) and used to transform S, Cereviside fie. by in vivo cloning) using the Sigma Yeast Transformation kit described below.

Preparation of an S, cerevisia® strain expressingwild type HSA genetically fused to a C3S linker and IL1-RA (N84Q) (see Table 18) could also be generated following the methods described above.

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

Table 20: Kinetics of HSA fusion variants.

a: Dilutions of NBA 'variants were "injected over immobilized shFeRn |~i§00 RU), b: The kinetic rate constants were obtained using a simple first-order (1:1) bsmolecuiar interaction mode!.. The kinetic values represent the average of duplicates, c: Not .determined1 due to weak binding pDj, in exampie 8 if was shown that the K5GQA variant did not significantly bind shFcRn, in Exampie 10 it wee shown that the K573F and R573A variants bind shFcRn stronger than HSA and in Exampie 11 it was shown that the D550N variant binds FeRn weaker than HSA.

In the present example it is shown that these observed difference in binding properties also are reflected in fusion pobrpeptides in differeni configurations: C-terminat fusions with a small moiety (HSA-FLAG), C-terminai fusions with a larger polypeptide (NSA-iLIRA); N4ermina! fusions with polypeptide (scFv-HSA); N- and C-terminai fusions (scFv-HSA-FLAG and scFv-HSA-scFv-FLAG).

Exampie 17. Conjugation of Horseradish peroxidase protein to Albumin and the K573F variant.

For conjugation analysis, eommerciaiiy available recombinant aiburnin (Recombumin™) was used as a control molecule, For this exampie, a final 20§hicpmL albumin K573P variant of the invention was purified from a fed batch fermentation by means described in Material and Methods, A two step pu nfseation was carried out;

The first step used a column (bed volume approximately 4001111, bed height 11 cm) packed with AlbuPureTb! matrix (Profcletlc). This was equilibrated with SOmM sodium acetate, pH 5,3 and loaded with neat culture supernatant, at approximately pH 5,5-8,5.,: to ..approximately 20 mg/mL matrix. The column was then washed with approximately 5 column volumes each of 50mM sodium acetate, pH 5,3, 50mM sodium phosphate, pH i.O. SDmM sodium phosphate, pH 7.0 and SOmM ammonium acetate, pH 8.0, respectively. Bound protein was eluted using approximately two column volumes of 50mM ammonium acetate, 10m iVI oetanoate, pH 7,0, The flow rate for the entire purification was 154mt/min.

For the second step, the eluate from the first step was diluted approximately two fold with water to give a conductivity of 2,5±0>5 mS/cm after adjustment to pH 5.S+0.3 with acetic acid. This was loaded onto a DEAE-Sepharose Fast Flow (GE Healthcare): eoiumn (bed volume approxlteately 4O0;mL( bed height 11cm), eduillbrated with SOmfyl sodium acetate, 5m!H octenoaie, pH 5:,5. Lpadihg was appfpxlrhateiy 30mg protein/mL matrix. The: column was washed with approximately δ column volumes of 80mfvt sodium acetate, SmM octanoate, pH 5.5. Followed by approximately 10 column volumes ofIbJmfvi potassium tefrabpratej; pH 9:.2. The bound protein was ©luted using two column volumes of 110mlV1 potassium: tetraebteto, 1200mM sodium chloride, approximately pH 9.0. The flow rate was 183mL/mln during the load and wash steps, and 169ml/mih during the elution step.

The eluate was concentrated and diafiitered against 145mM NaCI, using a Pall Centfamate Omega 10,000 Nominal lyWGO mernbrane, to give a final protein concentration of approximateiy 200mg/mL.

Both 200mg/mL stock solutions of the rHA and K573P variant albumin were diluted down to 5mg/mL, using phosphate buffer saline (PBS), pH adjusted to pH 6.5-5,7. This ensured a favourable pH environment for tee maieimfde reactive group of the EZ~Unk® Maieimlde Activated Horseradish Peroxidase (Thermo Scientific) to react with the free suiphydryk to form a stable thioester bond. 2mg of tee EZ-Unk<D Valeimsue Activated Horseradish Peroxidase (HRP) was mixed with either ImL of the Smg/ML rHA or K573P yafiantalbumin. This mixture ensured an approximate 2 fold molar excess ofthe albumin, or KS73P variant albumin. This mixture was minimally Incubated at 4 :C. for 24 hours. The reaction mixtures were then checked for conjugation, 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 (Vivaspin26, 10,060 MWCQ PES, Sartorius), and then individually applied to a Tricorn Superdex™ 200:, 10/300 GL coipte'h f:GE: Health care), run at a flow rate of 46cm/hr in PBS. The elufioh peak was fractionated and GP-HPLC analysed;. Fractions containing the conjugated species were pooled. concentrated and diafiitered against :bOmb! NaCl and analysed by GPHdPLC to demonstrate (Figure 24}

These samples were then assayed using the Biacore method: described herein (Table 21), This example demonstrates that the KB73P maintains Its increased affinity for shFcRn compared the the WT HSA,

Example 18, Conjugation of Fluorescein in Albumin and the K573F variant

The two same albumin samples used in Example 17» were also the start materials for this example, Le, Approximately 2Qt)mg/mL rHA or the K573P albumin variant

Fluoresceih~5~!Vialelmides Thermo Scientific (F5y) was dissolved in dimethylformamlde» to give a final eohcentratlon of 26mg/mL This was then further diluted into 18mis of PBS, pH adjusted fb approximately pH 8.5, To this solution either 1ml of 200mg/mL rHA or 1ml .of· 200mg/ml KS73P variant was added. This gave an approximate 2{) foid final molar excess of F5M, These samples were incubated and atiowed to con|ugate overnight at 4^0, in the dark, to allow the maielmide groups on the FSM to react with predominantly the free sulfhydryl, present in both aibumin species.

Following overnight incubation aliquots of the reaction mixtures wore extensively diafilterod against SOmy NeCI to remove unconjugated F5M, (Vivaspin20, 10,000 MW CO PES: Sartorius). Conjugation was confirmed by ultraviolet visualization of conjugated Fluorescein:Aibumins Following standard SDS-PAGE (Figure 25).

These diafiitered samples were then assayed using the Biaeore method described herein (Table 21). This example demonstrates that the conjugation of a small molecule to either rHA or a variant. e.g. Κ573Ρ does not affect tee trend in binding affinities to shFcRn,

Table 21: Representative Biaeore assay KD vaiues of conjugated rHA or a variant (K573P) when binding to immobilized shFcRn.

Example 19. Further albumin variants.

The following variants: were generated using the methods described in Example i E492T, N503O, E49iT+N503Ds K636R E542D, D494N+E495Q+T496ATE49SC3+T496A, N403K, K541A and KS41N. SPR analysis was carried out as described in Example 1S and the results presented in Figure 26 and figure 27.

Figure 30A and 3QB shows the effect on shFcRn binding for the albumin variants.

Substitutions: N503D. D494N+E49SG+T4S6A Ε482Τ*Ν603α E495Q+T496A within HSA had a negative inpact on binding to shFcRn at pH5,5.

Example 20. Variants of albumin at the C-fermini,

The following variants were generated using the methods described in Example 1. Binding to the shFcRn was determined as described in Materials and Methods and the results are presented in Table 22.

Table 22: Kinetics Of the HSA Oierminai swapped variant interactions with shFcEn,

a; Dilutions of HSA variants were injected over immobilized shFcRn RfSDO RU). b: The kinetic rate constants Were obtained using a simple firsForder tlH) bimoiecular interaction model. c: Data from Table 2 d: Data from Table 3

Not determined due to weak binding (ND)

This example demonstrates that for all C~terminaJ swaps to human albumin tested an increase in binding over the donor albumin was observed- All donor sequences contain the KS73P substitution 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 variant albumins prepared as described in Example 1, were performed as described in Example 4, Results are presented in Figures 28-31.

The competitive binding hierarchy was Identical for the variants fusions of HSA~FLA0 and , N-r-d-terminal scFv H8A-FLAG lo the hierarchy of the individual HSA variants (unfusedland fused) affinity data. For the fEIRa variants KS73PvftS73A, and the KbOOAiwere as predicted, however the D55QN appears to inhibit more efficiently than the:WT fusion.

Example 22. Further HSA variants

The foiiowing vanants were generated using methods described in Example 1; HSA E492G*K573A, HSA Ε492β* M503K+ K573A, HSA £4820* M503H: * K573A, HSA £4920 + K573P, HSA E492G + NS03K * RS73P, HSA E492G + NS33H + NS73R SPR analysis was performed as described in Materials and Methods. Results (Figure 32) showed that aiS HSA variants bound more strongly to siiFeRn compared to wild type HSA at pH 5.5. No binding was Observed at pH 7.4. HSA S492G*KS73A, HSA E492G+ N503K* I4S73A, unlike HSA E492G-*- N503H + K573A, had marginally improved binding beyond that of HSA K573A. The combination variants containing KS73P did hot show improved binding over the K573P single variant

Claims (15)

1. A method for preparing a variant of albumin, a fragment thereof or a fusion polypeptide comprising said variant albumin or fragment thereof, comprising following steps: a. providing a nucleic acid encoding a parent albumin having at least 90% sequence identity to SEQ ID NO: 2; b. modifying the sequence of step a., to encode said variant of albumin, fragment thereof or fusion polypeptide comprising said variant albumin or fragment thereof having a substitution corresponding to a substitution in SEQ ID NO: 2 selected among: K500A,C,D,E,F,G,H,L,M,N, Q,S,T,V,W,Y; c. introducing the modified sequence of step b., in a suitable host cell; d. growing the cells in a suitable growth medium under conditions leading to expression of said variant of albumin, fragment thereof or fusion polypeptide comprising said variant albumin or fragment thereof; and e. recovering said variant of albumin, fragment thereof or fusion polypeptide comprising said variant albumin or fragment thereof from the growth medium; wherein said variant of albumin, fragment thereof or fusion polypeptide comprising said variant albumin or fragment thereof, has a shorter plasma half-life compared with the parent albumin, fragment thereof or fusion polypeptide comprising said variant albumin or fragment thereof.

2. An isolated variant of albumin, a fragment thereof or a fusion polypeptide comprising said variant albumin or a fragment thereof having a shorter plasma half-life compared with the parent albumin, fragment thereof or fusion polypeptide comprising said parent albumin or fragment thereof, comprising at least 90% identity to SEQ ID NO: 2 and a substitution at a position corresponding to position 500 in SEQ ID NO:2, where the variant is not the variant consisting of SEQ ID NO: 2 with the substitution K500I or K500R.

3. The variant of albumin, fragment thereof or fusion polypeptide comprising said variant albumin or fragment thereof according to claim 2, comprising one or more further alteration that generates thio group on the surface.

4. The variant of albumin, fragment thereof or fusion polypeptide comprising said variant albumin or fragment thereof according to claim 2 or claim 3, wherein the sequence identity of said variant of albumin, fragment thereof or fusion polypeptide comprising said variant albumin or fragment thereof to SEQ ID NO: 2 is more than 95%, more preferred more than 96%, even more preferred more than 97%, more preferred more than 98% and most preferred more than 99%.

5. The variant of albumin, fragment thereof or fusion polypeptide comprising said variant albumin or said fragment thereof according to any one of claims 2 to 4 being a variant fragment of albumin or a fusion polypeptide comprising a variant fragment of albumin wherein the fragment is at least 20 amino acids, preferably at least 50 amino acids, preferably at least 100 amino acids, more preferred at least 200 amino acids, more preferred at least 300 amino acids, more preferred at least 400 amino acids and most preferred at least 500 amino acids.

6. An isolated nucleic acid encoding the variant of albumin, fragment thereof or fusion polypeptide comprising said variant albumin or fragment thereof according to any one of claims 2 to 5.

7. A conjugate comprising: an isolated variant of albumin or fragment thereof having a shorter plasma half-life compared with the parent albumin, fragment thereof or fusion polypeptide comprising said parent albumin or fragment thereof, comprising a substitution at a position corresponding to position 500 in SEQ ID NO:2; and a beneficial therapeutic moiety; where the isolated variant of albumin is not the variant consisting of SEQ ID NO: 2 with the substitution K500I or K500R.

8. An associate comprising: an isolated variant of albumin or fragment thereof having a shorter plasma half-life compared with the parent albumin, fragment thereof or fusion polypeptide comprising said parent albumin or fragment thereof, comprising a substitution at a position corresponding to position 500 in SEQ ID NO:2; and a beneficial therapeutic moiety; where the isolated variant of albumin is not the variant consisting of SEQ ID NO: 2 with the substitution K500I or K500R.

9. A fusion polypeptide comprising: an isolated variant of albumin or fragment thereof having a shorter plasma half-life compared with the parent albumin, fragment thereof or fusion polypeptide comprising said parent albumin or fragment thereof, comprising a substitution at a position corresponding to position 500 in SEQ ID NO:2; and a fusion partner polypeptide; where the isolated variant of albumin is not the variant consisting of SEQ ID NO: 2 with the substitution K500I or K500R.

10. A composition comprising: an isolated variant of albumin or fragment thereof having a shorter plasma half-life compared with the parent albumin, fragment thereof or fusion polypeptide comprising said parent albumin or fragment thereof, comprising a substitution at a position corresponding to position 500 in SEQ ID NO:2; or a conjugate comprising said variant albumin, wherein the binding to the neonatal Fc receptor (FcRn) is weaker than for the corresponding albumin or fragment thereof or fusion polypeptide comprising said albumin or fragment thereof or a conjugate comprising said albumin; where the isolated variant of albumin is not the variant consisting of SEQ ID NO: 2 with the substitution K500I or K500R.

11. The composition according to claim 10, wherein the binding coefficient (KD) of the variant of albumin, fragment thereof or fusion polypeptide comprising said variant albumin or fragment thereof or a conjugate comprising said variant albumin to FcRn is lower than the KD for SEQ ID NO: 2 to FcRn.

12. The composition according to claim 10 or claim 11, further comprising a compound comprising an albumin binding domain (ABD) and a pharmaceutically beneficial moiety.

13. The composition according to any one of claims 10 to 12, being a pharmaceutical composition.

14. An isolated variant of albumin, a fragment thereof or a fusion polypeptide comprising said variant of albumin or said fragment thereof when prepared by the method according to claim 1.

15. Use of an isolated variant of albumin or fragment thereof having a shorter plasma half-life compared with the parent albumin, fragment thereof or fusion polypeptide comprising said parent albumin or fragment thereof, comprising a substitution at a position corresponding to position 500 in SEQ ID NO:2, to increase the half-life of a beneficial therapeutic moiety; where the isolated variant of albumin is not the variant consisting of SEQ ID NO: 2 with the substitution K500I or K500R.