Polypeptides

Abstract

The present disclosure relates to a class of engineered polypeptides having a binding affinity for albumin. It also relates to new methods and uses that exploit binding by these and other compounds to albumin in different contexts, some of which have significance for treatment or diagnosis of disease in mammals including humans.

Claims

1. An albumin binding polypeptide comprising the amino acid sequence selected from: TABLE-US-00007 i) (SEQIDNO:204) LAX.sub.3AKX.sub.6X.sub.7ANX.sub.10ELDX.sub.14YGVSDFYKRLIX.sub.26 KAKTVEGVEALKX.sub.39X.sub.40ILX.sub.43X.sub.44LP wherein independently of each other X.sub.3 is selected from E, S and Q; X.sub.6 is selected from E and S; X.sub.7 is selected from A and S; X.sub.10 is selected from A, S and R; X.sub.14 is selected from A, S, C and K; X.sub.26 is selected from D and E; X.sub.39 is selected from D and E; X.sub.40 is selected from A and E; X.sub.43 is selected from A and K; X.sub.44 is selected from A, S and E; L in position 45 is present or absent; and P in position 46 is present or absent; and ii) an amino acid sequence which has at least 95% identity to the full-length sequence defined in i) with the proviso that X.sub.7 is neither L, E nor D; wherein the albumin binding polypeptide does not comprise a hormone polypeptide.

2. The albumin binding polypeptide according to claim 1, wherein X.sub.6 is E.

3. The albumin binding polypeptide according to claim 1, wherein X.sub.3 is S.

4. The albumin binding polypeptide according to claim 1, wherein X.sub.3 is E.

5. The albumin binding polypeptide according to claim 1, wherein X.sub.7 is A.

6. The albumin binding polypeptide according to claim 1, wherein X.sub.14 is S.

7. The albumin binding polypeptide according to claim 1, wherein X.sub.14 is C.

8. The albumin binding polypeptide according to claim 1, wherein X.sub.10 is A.

9. The albumin binding polypeptide according to claim 1, wherein X.sub.10 is S.

10. The albumin binding polypeptide according to claim 1, wherein X.sub.26 is D.

11. The albumin binding polypeptide according to claim 1, wherein X.sub.26 is E.

12. The albumin binding polypeptide according claim 1, wherein X.sub.39 is D.

13. The albumin binding polypeptide according to claim 1, wherein X.sub.39 is E.

14. The albumin binding polypeptide according to claim 1, wherein X.sub.40 is A.

15. The albumin binding polypeptide according to claim 1, wherein X.sub.43 is A.

16. The albumin binding polypeptide according to claim 1, wherein X.sub.44 is A.

17. The albumin binding polypeptide according to claim 1, wherein X.sub.44 is S.

18. The albumin binding polypeptide according to claim 1, wherein L in position 45 is present.

19. The albumin binding polypeptide according to claim 1, wherein P in position 46 is present.

20. The albumin binding polypeptide according to claim 1, which binds to albumin such that the k.sub.off value of the interaction is at most 510.sup.5 s.sup.1.

21. The albumin binding polypeptide according to claim 20, which binds to albumin such that the k.sub.off value of the interaction is at most 510.sup.6 s.sup.1.

22. The albumin binding polypeptide according to claim 1, whose wherein the amino acid sequence is selected from the group consisting of SEQ ID NO:1-144 and SEQ ID NO:164-203.

23. The albumin binding polypeptide according to claim 22, wherein the amino acid sequence is selected from the group consisting of SEQ ID NO:4-5, SEQ ID NO:7-8, SEQ ID NO:10-11, SEQ ID NO:13-14, SEQ ID NO:16-17, SEQ ID NO:19-20, SEQ ID NO:22-23, SEQ ID NO:25-26, SEQ ID NO:28-29, SEQ ID NO:31-32, SEQ ID NO:34-35, SEQ ID NO:37-38, SEQ ID NO:41-42, SEQ ID NO:49-50, SEQ ID NO:164-170 and SEQ ID NO:192-203.

24. The albumin binding polypeptide according to claim 22, wherein the amino acid sequence is selected from the group consisting of SEQ ID NO:1-144.

25. The albumin binding polypeptide according to claim 24, wherein the amino acid sequence is selected from the group consisting of SEQ ID NO:4-5, SEQ ID NO:7-8, SEQ ID NO:10-11, SEQ ID NO:13-14, SEQ ID NO:16-17, SEQ ID NO:19-20, SEQ ID NO:22-23, SEQ ID NO:25-26, SEQ ID NO:28-29, SEQ ID NO:31-32, SEQ ID NO:34-35, SEQ ID NO:37-38, SEQ ID NO:41-42 and SEQ ID NO:49-50.

26. The albumin binding polypeptide according to claim 1, further comprising one or more additional amino acid residues positioned at the N- and/or the C-terminal of the sequence defined in i).

27. The albumin binding polypeptide according to claim 26, in which the additional amino acids comprise at least one serine residue at the N-terminal of the polypeptide.

28. The albumin binding polypeptide according to claim 26, in which the additional amino acids comprise a glycine residue at the N-terminal of the polypeptide.

29. The albumin binding polypeptide according to claim 26, in which the additional amino acids comprise a cysteine residue at the N-terminal of the polypeptide.

30. The albumin binding polypeptide according to claim 26, in which the additional amino acids comprise a lysine residue at the C-terminal of the polypeptide.

31. The albumin binding polypeptide according to claim 26, in which the additional amino acids comprise a glycine residue at the C-terminal of the polypeptide.

32. The albumin binding polypeptide according to claim 26, in which the additional amino acids comprise a cysteine residue at the C-terminal of the polypeptide.

33. The albumin binding polypeptide according to claim 26, wherein the amino acid sequence is selected from any one of SEQ ID NO: 145-150 and SEQ ID NO:162-163.

34. The albumin binding polypeptide according to claim 1, comprising no more than two cysteine residues.

35. The albumin binding polypeptide according to claim 34, comprising no more than one cysteine residue.

36. The albumin binding polypeptide according to claim 1, which binds to human serum albumin.

37. A fusion protein or conjugate comprising i) a first moiety consisting of the albumin binding polypeptide according to claim 1; and ii) a second moiety consisting of a polypeptide having a desired biological activity, wherein the fusion protein or conjugate does not comprise a hormone polypeptide.

38. The fusion protein or conjugate according to claim 37, in which the second moiety having a desired biological activity is a therapeutically active polypeptide.

39. The fusion protein or conjugate according to claim 38, in which the second moiety having a desired biological activity is selected from the group consisting of human endogenous enzymes, growth factors, chemokines, cytokines and lymphokines.

40. The fusion protein or conjugate according to claim 39, in which the second moiety is selected from the group consisting of interleukin-2 (IL-2), interleukin-1 receptor antagonist (IL-1RA), keratinocyte growth factor (KGF), ancestim, cytotoxic T lymphocyte-associated protein 4 (CTLA-4), Factor VII, Factor VIII and Factor IX.

41. The fusion protein or conjugate according to claim 38, in which the second moiety having a desired biological activity is a non-human biologically active protein selected from the group consisting of bacterial toxins, enzymes and activating proteins.

42. The fusion protein or conjugate according to claim 37, in which the second moiety having a desired biological activity is a binding polypeptide capable of selective interaction with a target molecule.

43. The fusion protein or conjugate according to claim 42, in which the binding polypeptide is selected from the group consisting of antibodies and fragments and domains thereof retaining antibody binding activity; microbodies, maxybodies, avimers, other small disulfide-bonded proteins; binding proteins derived from a scaffold selected from the group consisting of staphylococcal protein A and domains thereof, domain GM of protein G from Streptococcus strain G148, lipocalins, ankyrin repeat domains, cellulose binding domains, crystallines, green fluorescent protein, human cytotoxic T lymphocyte-associated antigen 4, protease inhibitors, Kunitz domains, PDZ domains, SH3 domains, peptide aptamers, staphylococcal nuclease, tendamistats, fibronectin type III domain, transferrin, zinc fingers and conotoxins.

44. The fusion protein or conjugate according to claim 43, in which said target molecule is selected from the group consisting of amyloid (A) peptide; toxins, bacterial toxins, snake venoms; blood clotting factors, von Willebrand factor; interleukins, interleukin-13 (IL-13); myostatin; pro-inflammatory factors, tumor necrosis factor alpha (TNF-), TNF- receptor, IL-1, IL-23, IL-8; complement factors, complement component 3 (C3), C5; hypersensitivity mediators, histamine, immunoglobulin E (IgE); tumor antigens, cluster of differentiation molecule 19 (CD19), CD20, CD22, CD30, CD33, CD40, CD52, CD70, oncogene MET (cMet), epidermal growth factor receptor 1 (HER1), HER2, HER3, HER4, carbonic anhydrase IX (CAIX), carcinoembryonic antigen (CEA), IL-2 receptor, mucin 1 (MUC1), prostate-specific membrane antigen (PSMA), tumor-associated glycoprotein 72 (TAG-72), granulocyte-colony stimulating factor (G-CSF), granulocyte-macrophage colony-stimulating factor (GM-CSF), growth hormone (GH), insulin, and somatostatin.

45. The fusion protein or conjugate according to claim 37, comprising a further moiety consisting of a polypeptide having a further, desired biological activity, which may be the same as or different from that of the second moiety.

46. The fusion protein or conjugate according to claim 45, wherein the second moiety is a therapeutically active polypeptide, and the further moiety is a binding polypeptide capable of selective interaction with a target molecule.

47. The fusion protein or conjugate according to claim 37, in which the second moiety is conjugated to the albumin binding polypeptide via the thiol group of any cysteine residue present at position X.sub.14 of the polypeptide.

48. The albumin binding polypeptide, according to claim 1, further comprising a cytotoxic agent.

49. The albumin binding polypeptide, according to claim 48, wherein the cytotoxic agent is selected from calicheamycin, auristatin, doxorubicin, maytansinoid, taxol, ecteinascidin, geldanamycin, methotrexate and their derivatives, and combinations thereof.

50. The albumin binding polypeptide, according to claim 1 further comprising a label.

51. The albumin binding polypeptide, according to claim 50, in which said label is selected from the group consisting of fluorescent dyes and metals, chromophoric dyes, chemiluminescent compounds and bioluminescent proteins, enzymes, radionuclides and particles.

52. The albumin binding polypeptide, according to claim 51, comprising a chelating environment provided by a polyaminopolycarboxylate chelator conjugated to the albumin binding polypeptide via a thiol group of a cysteine residue or an amine group of a lysine residue.

53. The albumin binding polypeptide, according to claim 52, wherein the polyaminopolycarboxylate chelator is 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid or 1,4,7,10-tetraazacyclododecane-1,4,7-tris-acetic acid-10-maleimidoethylacetamide.

54. The albumin binding polypeptide, according to claim 52, wherein the polyaminopolycarboxylate chelator is diethylenetriaminepentaacetic acid.

55. A polynucleotide encoding the albumin binding polypeptide according to claim 1.

56. A method of producing a polypeptide, comprising expressing the polynucleotide according to claim 55.

57. An expression vector comprising the polynucleotide according to claim 55.

58. A host cell comprising the expression vector according to claim 57.

59. A method of producing the albumin binding polypeptide according to claim 1 by non-biological peptide synthesis using amino acids and/or amino acid derivatives having protected reactive side-chains, the non-biological peptide synthesis comprising step-wise coupling of the amino acids and/or the amino acid derivatives to form the albumin binding polypeptide, deprotecting the protected reactive sidechains of the albumin binding polypeptide, and folding of the albumin binding polypeptide in aqueous solution.

60. The method of producing a polypeptide according to claim 59, further comprising conjugating the albumin binding polypeptide with a therapeutically active polypeptide.

Description

FIGURES

(1) FIG. 1A-1F is a listing of the amino acid sequences of examples of albumin binding polypeptides of the present disclosure (SEQ ID NO:1-152, SEQ ID NO:155-203), the GA3 domain from protein G of Streptococcus strain G148 extended by a N-terminal glycine residue (SEQ ID NO:153) and an albumin binding polypeptide derived from G148-GA3 as previously described by Jonsson et al (supra, SEQ ID NO:154).

(2) FIG. 2 shows the result of binding analysis performed in a Biacore instrument for investigating the binding of the albumin binding polypeptide PEP07912 (SEQ ID NO:157) to human serum albumin. Three different concentrations of purified protein (40 nM, fat gray line; 10 nM, black line; and 2.5 nM, gray line) were injected over a surface with 955 RU of immobilized human serum albumin.

(3) FIGS. 3A-C show the result of binding analysis performed by ELISA for investigating the binding of the albumin binding polypeptides PEP07913 (SEQ ID NO:153), PEP06923 (SEQ ID NO:154), PEP07271 (SEQ ID NO:155), PEP07912 (SEQ ID NO:157), PEP07554 (SEQ ID NO:156), PEP07914 (SEQ ID NO:158), PEP07968 (DOTA-conjugated PEP07911, SEQ ID NO:159) and PEP07844 (SEQ ID NO:161), to IgG molecules present in 126 individual normal human sera, where A) shows the average OD-value, B) shows the percentage of negative sera (defined as OD<0.15), and C) shows the percentage of positive sera (defined as OD>1.0).

(4) FIGS. 4A-B are chromatograms showing analysis of purified, chemically produced albumin binding polypeptide PEP07834 (SEQ ID NO:160), where A) shows the absorbance signal at 220 nm, blank subtracted, and B) shows the absorbance signal at 280 nm, blank subtracted. Two peaks appeared at both wavelengths.

(5) FIGS. 5A-B are spectrograms showing masspectrometric analysis of the two peaks identified in FIGS. 4A) and B). A) is the spectrogram of the first peak, i.e. the monomer of PEP07834 (SEQ ID NO:160), and B) is the spectrogram of the dimer of PEP07834.

(6) FIGS. 6A-C are diagrams showing an immunogenicity assessment of albumin binding polypeptides PEP07913 (SEQ ID NO:153), PEP07912 (SEQ ID NO:157), PEP07914 (SEQ ID NO:158) and PEP07968 (DOTA-conjugated PEP07911, SEQ ID NO:159) in a CD3.sup.+ CD4.sup.+ T cell proliferation assay. A) shows the number of individuals responding to the albumin binding polypeptides compared to recombinant human albumin in a cohort of 52 Caucasian donors. B) shows the average stimulation indices (SI) for PEP07913, PEP07912, PEP07914 and PEP07968 compared to the negative control containing recombinant human albumin. C) shows the number of responding individuals against all proteins in the study as compared to the buffer control.

(7) FIGS. 7A-C shows the result of binding analysis performed in a Biacore instrument for investigating the binding of the albumin binding polypeptides A) PEP07986 (SEQ ID NO:163), B) PEP08296 (DOTA-conjugated PEP08185, SEQ ID NO:148) and C) PEP06923 (SEQ ID NO:154) to albumin from different species. The sensorgrams shown correspond to protein injected at a concentration of 40 nM over surfaces immobilized with albumin from human (1130 RU), thin gray line; cynomolgus monkey (1046 RU), thick gray line; rat (831 RU), thick light gray line; dog (1053 RU), thin black line; and mouse (858 RU), thick black line.

(8) FIG. 8 shows the inhibitory effect of Z.sub.X-PP013 (open circles), Z.sub.Y-PP013 (open squares) and Z.sub.neg-PP013 (closed triangles) on cytokine induced TF-1 cell proliferation in the presence of five times molar excess of HSA.

(9) FIG. 9 shows the maximum binding responses obtained by Biacore analysis of PEP07986 (SEQ ID NO:163) stored at 4, 25 or 40 C. for one week, two weeks, one month and three months as indicated, at a concentration of 2 mg/ml, injected over immobilized HSA (704 RU) at a concentration of 10 nM. Non-treated samples from time=0 are shown as references.

(10) FIG. 10 shows the result of binding analysis performed in a Biacore instrument for investigating the binding of the albumin binding polypeptide PEP08296 (DOTA-conjugated PEP08185, SEQ ID NO:148) to human serum albumin before and after heat treatment. Two concentrations of PEP08296 (0.8 nM, grey lines; 4 nM, black lines) were injected over a surface with 724 RU of immobilized human serum albumin. Solid lines are before heat treatment and hatched lines after heat treatment for 10 minutes at 90 C.

(11) FIGS. 11A-B show the overlay of two CD spectra of PEP08296 (DOTA-conjugated PEP08185, SEQ ID NO:148) before and after heat treatment for 12 min at 90 C. A) Sample incubated in PBS pH 7.2. B) Sample incubated in PBS pH 4.0.

(12) FIG. 12 shows the maximum intensity projection (MIP) image of the whole body distribution of .sup.68Ga-PEP08296 in a healthy rat, summed during 1.5 h of data collection immediately following intravenous injection (tail vein). Circulating radioactivity in the major vessels (e.g. the jugular (long arrow) and femoral (short arrow)), the heart (H), liver (L), spleen (S), kidney (K) and bladder (B) are readily delineated.

(13) FIG. 13 shows a gel filtration chromatogram of PEP07986 (SEQ ID NO:163) injected at a concentration of 42 mg/ml, black solid line. A chromatogram of ovalbumin (Mw 43 kDa) injected at a concentration of 5 mg/ml, gray broken line, is included for comparison, confirming that the peak for PEP07986 is not an aggregate, which would have been expected in the void volume eluted at an earlier time point than ovalbumin.

(14) The invention will now be illustrated further through the non-limiting description of experiments conducted in accordance therewith. Unless otherwise specified, conventional chemistry and molecular biology methods were used throughout.

EXAMPLES

Example 1

Cloning, Expression, Purification and Characterization of Albumin Binding Polypeptides

(15) In this example, ten different albumin binding polypeptides, PEP07913 (SEQ ID NO:153), PEP07912 (SEQ ID NO:157), PEP07914 (SEQ ID NO:158), PEP07968 (DOTA-conjugated PEP07911, SEQ ID NO:159), PEP06923 (SEQ ID NO:154), PEP07271 (SEQ ID NO:155), PEP07554 (SEQ ID NO:156), PEP07844 (SEQ ID NO:161), PEP07986 (SEQ ID NO:163) and PEP08296 (DOTA-conjugated PEP08185, SEQ ID NO:148), the amino acid sequences of which are set out in FIG. 1 and in the appended sequence listing, were cloned, purified and characterized.

(16) Material and Methods

(17) Cloning of Albumin Binding Polypeptide Variants

(18) Mutations in G148-GA3 were generated using site directed mutagenesis with the appropriate oligonucleotides to obtain the desired albumin binding polypeptide variants. The gene fragments were amplified by PCR with primers adding specific endonuclease sites as well as an N-terminal MGSS sequence preceding the albumin binding polypeptide variants. The fragments were cleaved with Ndel and Notl, purified and ligated to a cloning vector, the plasmid pAY02556 (containing an origin of replication from pBR322, a kanamycin resistance gene and a T7 promoter for expression of the gene of interest), restricted with the same enzymes. Ligations were transformed to electrocompetent E. coli TOP10 cells. The transformed cells were spread on TBAB plates (30 g/I tryptose blood agar base) supplemented with 50 g/ml of kanamycin, followed by incubation at 37 C. overnight. The colonies were screened using PCR and sequencing of amplified fragments was performed using the biotinylated oligonucleotide and a BIGDYE Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems), used in accordance with the manufacturer's protocol. The sequencing reactions were purified by binding to magnetic streptavidin coated beads using a Magnatrix 8000 (NorDiag AB), and analyzed on ABI PRISM 3100 Genetic Analyzer (PE Applied Biosystems). All albumin binding polypeptide variants were subcloned as monomers and the constructs encoded by the expression vectors were MGSS-[PP###], where PP### corresponds to the amino acid residues constituting the sequence of the albumin binding polypeptide.

(19) In addition, the gene fragments of G148-GA3, PP007 (SEQ ID NO:7), PP013 (SEQ ID NO:13) and PP037 (SEQ ID NO:37) were amplified by PCR with primers adding specific endonuclease sites as well as a hexahistidin sequence, a TEV protease site and a glycine residue before the amino acid residues constituting the sequence of the albumin binding polypeptide. The polypeptides PEP07913 (SEQ ID NO:153), PEP07912 (SEQ ID NO:157), PEP07914 (SEQ ID NO:158) and PEP07968 (SEQ ID NO:159) correspond to the albumin binding polypeptides G148-GA3, PP007 (SEQ ID NO:7), PP013 (SEQ ID NO:13) and PP037 (SEQ ID NO:37) with glycine residues added. The fragments were cleaved with XbaI and NotI, purified and ligated to a cloning vector, the plasmid pAY02512 (containing an origin of replication from pBR322, a kanamycin resistance gene and a T7 promoter for expression of the gene of interest. The cloning site is preceded by a sequence encoding a peptide containing a hexahistidine tag followed by a cleavage site for the TEV protease), restricted with the same enzymes. Ligation, transformation and sequence verification were performed as described above. The four albumin binding polypeptide variants G148-GA3, PP007, PP013 and PP037 were subcloned as monomers and the constructs encoded by the expression vectors were MGSSHHHHHHLQSSGVDLGTENLYFQG-[PP###] (SEQ ID NO:205).

(20) The expression vector encoding MGSSHHHHHHLQSSGVDLGTENLY-FQG-[PP013] (SEQ ID NO:206) was further modified by site directed mutagenesis using oligonucleotides, resulting in the insertion of a serine residue before the amino acid residues constituting the sequence of the albumin binding polypeptide, to obtain the construct MGSSHHHHHHLQSSGVDLGTENLYFQ-GS-[PP013] (SEQ ID NO:207). This construct was further modified by 1) site directed mutagenesis to replace the serine residue at position 14 (within PP013) with a cysteine residue, generating MGSSHHHHHHLQSSGVDLGTENLYFQGS-[PP049] (SEQ ID NO:208), and 2) addition of a glycine residue C-terminally, generating MGSSHHHHHHLQSSGVDLGTENLYFQGS-[PP049]-G (SEQ ID NO:209).

(21) The addition of glycine C-terminally was accomplished by PCR amplification with primers including nucleotides encoding the glycine residue and specific endonuclease sites. The fragment was cleaved with Xba I and Not I, purified and ligated to a cloning vector, the plasmid pAY02641 (containing an origin of replication from pBR322, a kanamycin resistance gene and a T7 promoter for expression of the gene of interest), restricted with the same enzymes. Ligation, transformation and sequence verification were performed as described above.

(22) Protein Expression

(23) The albumin binding polypeptide variants were expressed in E. coli BL21 (DE3) either with an N-terminal MGSS-extension or with an N-terminal His.sub.6-tag followed by a TEV-protease recognition site and a glycine residue. A colony of each albumin binding polypeptide variant was used to inoculate 4 ml TSB+YE medium supplemented with kanamycin to a concentration of 50 g/ml. The cultures were grown at 37 C. for approximately 5 hours. 3 ml from each of the cultures was used to inoculate 800 ml TSB+YE supplemented with kanamycin to a concentration of 50 g/ml in parallel bio reactors (Greta system, Belach Bioteknik AB). The cultivations were performed at 37 C., with aeration at 800 ml/minute and an agitation profile to keep dissolved oxygen levels above 30%, to an OD600 of 2, which was followed by addition of IPTG to a final concentration of 0.5 mM. Cultivation was continued for five hours after which the cultivation was cooled to 10 C., aeration was stopped and agitation lowered to 300 rpm. Cell pellets were harvested by centrifugation (15600g, 4 C., 20 minutes) and stored at 20 C. until purification.

(24) Purification of Albumin Binding Polypeptide Variants with a His.sub.6-Tag and a TEV-Protease Site

(25) Frozen cell pellets harboring soluble hexahistidine-tagged polypeptides PEP07913 (SEQ ID NO:153), PEP07912 (SEQ ID NO:157), PEP07914 (SEQ ID NO:158), PEP07968 (SEQ ID NO:159), PEP07986 (SEQ ID NO:163) and PEP08185 (SEQ ID NO:148) were suspended in 35 ml binding buffer (20 mM sodium phosphate, 0.5 M NaCl, 20 mM imidazole, pH 7.4) with an addition of 1000 U BENZONASE (1.01654.001, Merck) and disrupted by ultrasonication. For each of the polypeptides, the ultrasonicated suspension was clarified by centrifugation (1 h, 37000g, 4 C.) and the supernatant was loaded onto a His GRAVITRAP column (11-0033-99, GE Healthcare). The column was washed with 10 ml washing buffer (20 mM sodium phosphate, 0.5 M NaCl, 60 mM imidazole, pH 7.4), before eluting the polypeptide with 3 ml elution buffer (20 mM sodium phosphate, 0.5 M NaCl, 0.5 M imidazole, pH 7.4). The buffer was exchanged to a cleavage buffer (50 mM Tris-HCl, 150 mM NaCl, pH 8) using PD-10 desalting column (17-0851-01, GE Healthcare). The amount of polypeptide product was determined by measuring the absorbance at 280 nm before adding DTT to a final concentration of 5 mM. His.sub.6-tagged TEV protease was added to the cleavage buffer at a 1:10 mass ratio relative to the polypeptide product. The cleavage was performed over night under slow mixing at 4 C. Imidazole was added to the cleavage mix, to a concentration of 20 mM, before loading the mix onto a His GRAVITRAP column (11-0033-99, GE Healthcare) for removing cleaved His.sub.6-tags, His.sub.6-tagged TEV protease and His.sub.6-tagged uncleaved product.

(26) For each variant, the flow-through, containing the albumin binding polypeptide variant, was further purified by reversed phase chromatography (RPC), as follows. The flow-through fraction was loaded on 1 ml Resource 15 RPC column (GE Healthcare), previously equilibrated with RPC A Buffer (0.1% TFA in water). After column wash with 10 column volumes (CV) RPC A Buffer, bound polypeptides were eluted with a linear gradient of 0-50% RPC B Buffer (0.1% TFA in acetonitrile) during 10 CV. The flow rate was 2 ml/min and the absorbance at 280 nm was monitored. Fractions containing albumin binding polypeptide variant were identified by SDS-PAGE analysis and pooled.

(27) The RPC-purified albumin binding polypeptide variants were further purified by gel filtration on 120 ml Superdex 75 (GE Healthcare) packed in an XK16 column (GE Healthcare). The running buffer was 1PBS, and the flow rate 2 ml/min. Fractions containing pure albumin binding polypeptide variant were pooled and concentrated to approximately 1.3 mg/ml. Finally, the concentrate was purified from trace amounts of remaining endotoxins by using 1 ml columns of AffinityPak Detoxi-Gel Endotoxin removing gel (Pierce, prod#20344), according to the manufacture's recommendations.

(28) The albumin binding polypeptide variants PEP07911 and PEP08185 were conjugated with Mal-DOTA before the RPC-purification step, as follows. The buffer of the flow-through fraction from the IMAC-FT purification step was exchanged to 0.2 M NaAc, pH 5.5, using a disposable PD-10 desalting column (GE Healthcare). Maleimido-mono-amide-DOTA (Macrocyclics, cat. no. B-272) was added at 5-fold molar excess and incubated for 60 minutes at 30 C. under continuous shaking. The resulting polypeptide were denoted PEP07968 and PEP08296, respectively.

(29) Purification of Albumin Binding Polypeptide-Variants Without His.sub.6-Tag

(30) Frozen cell pellets harboring soluble albumin binding polypeptide variants PEP06923 (SEQ ID NO:154), PEP07271 (SEQ ID NO:155), PEP07554 (SEQ ID NO:156) and PEP07844 (SEQ ID NO:161) were suspended in 20 mM Tris-HCl, pH 8 and disrupted by ultrasonication. For each of the polypeptide variants, the ultrasonicated suspension was clarified by centrifugation (30 min, 32000g, 4 C.) and the supernatant was loaded onto a HSA-Sepharose column (GE Healthcare). After washing with TST-buffer (25 mM Tris-HCl, 1 mM EDTA, 200 mM NaCl, 0.05% Tween 20, pH 8.0), followed by 5 mM NH.sub.4Ac, pH 5.5, bound albumin binding polypeptide variant was eluted with 0.5 M HAc, pH 3.2.

(31) The albumin binding polypeptide variants were further purified by reversed phase chromatography (RPC), as follows. For each of the variants, the eluate from the HSA-affinity purification step was loaded on 1 ml Resource 15 RPC column (GE Healthcare), previously equilibrated with RPC A Buffer (0.1% TFA in water). After column wash with 10 CV RPC A Buffer, bound polypeptides were eluted with a linear gradient of 0-50% RPC B Buffer (0.1% TFA in acetonitrile) during 10 CV. The flow rate was 2 ml/min and the absorbance at 280 nm was monitored. Fractions containing pure albumin binding polypeptide variants were identified by SDS-PAGE analysis and pooled. Finally, the buffer was exchanged to 1PBS (2.68 mM KCl, 137 mM NaCl, 1.47 mM KH.sub.2PO.sub.4, 8.1 mM Na.sub.2HPO.sub.4, pH 7.4) using a disposable PD-10 desalting column (GE Healthcare).

(32) Characterization of Purified Albumin Binding Polypeptide-Variants

(33) The concentration was assessed by measuring the absorbance at 280 nm using a NANODROP ND-1000 Spectrophotometer. The proteins were further analyzed with SDS-PAGE and LC-MS.

(34) For the SDS-PAGE analysis, approximately 10 g of each albumin binding polypeptide variant was mixed with NuPAGE LDS Sample Buffer (Invitrogen), incubated at 70 C. for 15 min and loaded onto NuPAGE 4-12% Bis-Tris Gels (Invitrogen). The gels were run with NuPAGE MES SDS Running Buffer (Invitrogen) in an XCell II SureLock Electrophoresis Cell (Novex) employing the Sharp Prestained Standard (Invitrogen) as molecular weight marker and using PhastGel BlueR (GE Healthcare) for staining.

(35) To verify the identity of the albumin binding polypeptide variants, LC/MS analyses were performed using an Agilent 1100 LC/MSD system, equipped with API-ESI and a single quadruple mass analyzer. Approximately 10 g of each of the purified albumin binding polypeptide variants was loaded on a Zorbax 300SB-C8 Narrow-Bore column (2.1150 mm, 3.5 m, Agilent Technologies) at a flow-rate of 0.5 ml/min. Polypeptides were eluted using a linear gradient of 10-70% solution B for 15 min at 0.5 ml/min. The separation was performed at 30 C. The ion signal and the absorbance at 280 and 220 nm were monitored. The molecular weights of the purified albumin binding polypeptide variants were confirmed by MS.

(36) Results

(37) The expression levels of the albumin binding polypeptide variants were 10-30 mg product/g cell pellet, as estimated from SDS-PAGE analysis.

(38) For all variants, the purity, as determined by SDS-PAGE analysis, exceeded 95% and the LC/MS analysis verified the correct molecular weights. After purification, between 1 and 8 mg of pure polypeptide was obtained for each of the ten albumin binding polypeptide variants.

Example 2

Affinity Determination for Albumin Binding Polypeptides

(39) In this example, PEP06923 (SEQ ID NO:154), PEP07271 (SEQ ID NO:155), PEP07844 (SEQ ID NO:161), PEP07912 (SEQ ID NO:157), PEP07913 (SEQ ID NO:153), PEP07914 (SEQ ID NO:158) and PEP07968, (DOTA-conjugated PEP07911, SEQ ID NO:159), synthesized or expressed and purified in Example 1 were characterized for affinity to human serum albumin (HSA) using a Biacore instrument. PEP07913 corresponds to the amino acid sequence of G148-GA3 with addition of a N-terminal glycine residue, whereas PEP07271, PEP07844, PEP07912, PEP07914 and PEP07968 correspond to the albumin binding polypeptides of PP001 (SEQ ID NO:1), PP043 (SEQ ID NO:43), PP007 (SEQ ID NO:7), PP013 (SEQ ID NO:13) and PP037 (SEQ ID NO:37) with different N-terminal amino acid additions.

(40) Material and Methods

(41) Biosensor analysis on a Biacore2000 instrument (GE Healthcare) was performed with HSA (ALBUCULT, Novozymes), immobilized by amine coupling onto the carboxylated dextran layer of the surfaces of CM-5 chips (research grade; GE Healthcare) according to the manufacturer's recommendations. Surface 1 of the chip was activated and deactivated and used as a reference cell (blank surface) during injections, whereas surface 2 comprised HSA immobilized to 731 resonance units (RU) and surface 4 comprised HSA immobilized to 955 RU. The purified albumin binding polypeptide variants were diluted in running buffer HBS-EP (GE Healthcare) to 2.5 nM, 10 nM and 40 nM, and injected at a constant flow-rate of 50 l/min for 5 minutes, followed by injection of HBS-EP for 60 minutes. The surfaces were regenerated with one injection of 25 l HCl, 10 mM. The affinity measurements were performed in two sets; in the first set HBS-EP, PEP06923, PEP07271, PEP07912, PEP07913, PEP07914 and PEP07968 were injected (chip surface 2), and in the second set HBS-EP, PEP06923, PEP07844, PEP07912 and PEP07914 were injected (chip surface 4). PEP06923 was injected twice in each run as a control. The results were analyzed with a BIAEvaluation software (GE Healthcare). Curves of the blank surface were subtracted from the curves of the ligand surfaces.

(42) Results

(43) The Biacore 2000 instrument has a technical limitation, hindering measurements of very high affinity. Hence, the purpose of the Biacore study was not to determine the exact kinetic parameters of the albumin binding polypeptide variants' affinity for HSA. However, the results provide a quantitative estimation of the relative affinities of these polypeptides for albumin. After subtraction of reference surface and buffer injection, curves were fitted to a 1:1 (Langmuir) binding model using BIAevaluation software with correction for mass transfer and with RUmax set as a local parameter. Curves are shown in FIG. 2. The relative K.sub.D, k.sub.a (k.sub.on) and k.sub.d (k.sub.off) values were estimated and are presented in the Tables below.

(44) TABLE-US-00002 TABLE 1 Kinetic parameters (k.sub.a, k.sub.d and K.sub.D) of albumin binding polypeptides to HSA, 1st set k.sub.a (Ms.sup.1) k.sub.d (s.sup.1) K.sub.D (M) PEP07913 5.7 10.sup.5 9.3 10.sup.4 1.6 10.sup.9 PEP06923 (1) 2.9 10.sup.7 2.9 10.sup.5 9.9 10.sup.13 PEP06923 (2) 2.6 10.sup.7 2.8 10.sup.5 1.1 10.sup.12 PEP07271 3.9 10.sup.6 2.9 10.sup.5 7.5 10.sup.12 PEP07912 4.6 10.sup.6 2.8 10.sup.5 6.2 10.sup.12 PEP07914 3.5 10.sup.6 2.5 10.sup.5 7.2 10.sup.12 PEP07968 3.0 10.sup.6 2.7 10.sup.5 9.0 10.sup.12

(45) TABLE-US-00003 TABLE 2 Kinetic parameters (k.sub.a, k.sub.d and K.sub.D) of albumin binding polypeptides to HSA, 2nd set k.sub.a (Ms.sup.1) k.sub.d (s.sup.1) K.sub.D (M) PEP06923 (1) 2.0 10.sup.7 2.6 10.sup.5 1.3 10.sup.12 PEP06923 (2) 2.1 10.sup.7 2.5 10.sup.5 1.2 10.sup.12 PEP07912 5.4 10.sup.6 2.8 10.sup.5 5.2 10.sup.12 PEP07914 3.8 10.sup.6 2.6 10.sup.5 6.9 10.sup.12 PEP07844 5.4 10.sup.6 2.3 10.sup.5 4.4 10.sup.12

(46) As shown in Tables 1 and 2, PEP07271 (SEQ ID NO:155), PEP07844 (SEQ ID NO:161), PEP07912 (SEQ ID NO:157), PEP07914 (SEQ ID NO:158) and PEP07968 (DOTA-conjugated PEP07911, SEQ ID NO:159) all seem to have approximately the same affinity for HSA, widely exceeding the affinity of the parent G148-GA3 (PEP07913; SEQ ID NO:153). The HSA affinity of these polypeptides is slightly lower compared to PEP06923 (SEQ ID NO:154), despite similar off-rate.

Example 3

Determination of Melting Temperature (Tm) for Albumin Binding Polypeptides

(47) In this example, the albumin binding polypeptide variants PEP07913 (SEQ ID NO:153), PEP06923 (SEQ ID NO:154), PEP07271 (SEQ ID NO:155), PEP07554 (SEQ ID NO:156), PEP07912 (SEQ ID NO:157), PEP07914 (SEQ ID NO:158), PEP07968 (DOTA-conjugated PEP07911, SEQ ID NO:159), PEP07844 (SEQ ID NO:161) and PEP07986 (SEQ ID NO:163), expressed and purified as described in Example 1, and the albumin polypeptide variant PEP07975 (DOTA-conjugated PEP07834, SEQ ID NO:160), produced as described in Example 5, were analyzed by CD analysis. PEP07913 corresponds to the sequence of G148-GA3 having an N-terminal glycine residue, PEP06923 is an engineered high affinity derivative previously described by Jonsson et al, supra, whereas PEP07271, PEP07554, PEP07912, PEP07914, PEP07968, PEP07844 and PEP07975 are examples of the albumin binding polypeptides of PP001 (SEQ ID NO:1), PP007 (SEQ ID NO:7), PP013 (SEQ ID NO:13), PP037 (SEQ ID NO:37) and PP043 (SEQ ID NO:43) having different N-terminal amino acid additions according to the present disclosure.

(48) Material and Methods

(49) Purified albumin binding polypeptide variants were diluted in 1PBS, to final concentrations between 0.4 and 0.5 mg/ml. Circular dichroism (CD) analysis was performed on a Jasco J-810 spectropolarimeter in a cell with an optical path-length of 1 mm. In the variable temperature measurements, the absorbance was measured at 221 nm from 20 C. to 90 C., with a temperature slope of 5 C./min.

(50) Results

(51) The melting temperatures (Tm) of the different albumin binding polypeptide variants were calculated by determining the midpoint of the transition in the CD vs. temperature plot. The results are summarized in Table 3 below.

(52) TABLE-US-00004 TABLE3 DeterminedTmvaluesoftestedalbuminbinding polypeptidevariants N-terminal Variant SEQIDNO:# sequence.sup.3 Tm( C.) PEP07913 SEQIDNO:153 GL 61 PEP06923 SEQIDNO:154 GSSL 57 PEP07271 SEQIDNO:155 GSSL 65 PEP07554 SEQIDNO:156 GSSL 58 PEP07912 SEQIDNO:157 GL 53 PEP07914 SEQIDNO:158 GL 59 PEP07968 SEQIDNO:159.sup.1 GL 53 PEP07975 SEQIDNO:160.sup.1,2 AL 50 PEP07844 SEQIDNO:161 GSSL 65 PEP07986 SEQIDNO:163 GSL 61 .sup.1The peptide is conjugated with maleimide-DOTA at the cysteine .sup.2The peptide is amidated at the C-terminus .sup.3Leucine (underlined) is the residue in position 1 of the amino acid sequence of the albumin binding polypeptide as defined in the first aspect of the present disclosure

(53) The polypeptide PEP07968 is identical to PEP07912, except for the former having a cysteine residue in position 14 conjugated with maleimide DOTA, and the latter a serine residue. Thus, the DOTA modification should not affect the melting temperature. Also PEP07975 is conjugated with DOTA using C.sub.14, and is identical to PEP07968 except for the C-terminal amide (resulting from the peptide synthesis in Example 5) and for having an N-terminal alanine instead of a glycine. Furthermore, comparing PEP07912 and PEP07554 reveals that an N-terminal serine gives a higher melting temperature than a glycine in the same position (5 C. difference in Tm). Thus, all albumin binding polypeptide variants according to the present disclosure show Tm above 55 C., except PEP07912 and DOTA-conjugated variants. Taking into consideration the importance of the N-terminal portion, all the tested albumin binding polypeptides are superior to the prior art derivative of Jonsson et al, i.e. PEP06923.

Example 4

Serum Response Analysis

(54) The percentage of human serum containing IgG, capable of binding to a set of albumin binding polypeptides as disclosed herein was analyzed by ELISA. In total, 149 serum samples corresponding to 127 individuals were screened.

(55) Material and Methods

(56) ELISA plates (96-well, half area plates (Costar, cat. No. 3690)) were coated with 50 l/well of ALBUCULT (Novozymes) diluted to 8 g/ml in coating buffer (Sigma, cat. No. 3041). The plates were coated over night for three days at 4 C. On the day of analysis, the plates were washed twice with tap water and blocked for 2 hours with 100 l of phosphate buffered saline (PBS) containing 0.05% casein (PBSC). The plates were emptied and 50 l/well of the albumin binding polypeptides PEP07913 (SEQ ID NO:153), PEP06923 (SEQ ID NO:154), PEP07271 (SEQ ID NO:155), PEP07912 (SEQ ID NO:157), PEP07554 (SEQ ID NO:156), PEP07914 (SEQ ID NO:158), PEP07968 (DOTA-conjugated PEP07911, SEQ ID NO:159) and PEP07844 (SEQ ID NO:161), diluted to 2 g/ml in PBSC were added according to a pre-made plate layout. After incubation for two hours at room temperature (RT), the plates were washed in PBSC four times using an automated ELISA washer. The 149 serum samples from 129 individuals were diluted 50 times in PBSC by adding 24 l serum to 1174 l PBSC. 50 l of the diluted sera was added per well according to the pre-made plate layout. Each serum sample was tested as a singlet. Positive and negative controls were included on each plate and for each albumin binding polypeptide. Albumin binding antibodies (50 l, 0.5 l/ml immunoglobulin solution prepared in house from sera from primates immunized with PEP06923 (SEQ ID NO: 154)) was added as a positive control and 50 l PBSC was used as a negative control. The plates were incubated for one hour at RT and subsequently washed four times in PBSC using an automated ELISA washer. The bound IgG was detected with 50 l/well of anti-human IgG (Southern Biotech, cat no 2040-05) diluted 10 000 times in PBSC. After washing four times in PBSC using an automated ELISA washer, 50 l/well of substrate was added (Pierce cat. No. 34021). The reaction was stopped after 10-15 minutes by the addition of 50 l H.sub.2SO.sub.4 to each well, prior to measuring the absorbance using a multi-well plate reader (Victor3, Perkin Elmer).

(57) Results

(58) Of the 149 sera screened for IgG binding to the albumin binding polypeptides, 23 were negative for all eight polypeptides (OD-value<0.1), i.e. showed no IgG bound to the polypeptides. The analysis was performed with the 126 sera that were positive for one or more albumin binding polypeptides. The average absorbance was calculated (FIG. 3A) and the percentage of sera with OD-values values either <0.15 (FIG. 3B) or >1.0 (FIG. 3C). The highest average OD-value and the highest percentage of serum with IgG binding were obtained with PEP07913 (SEQ ID NO:153), PEP06923 (SEQ ID NO:154) and PEP07844 (SEQ ID NO:161), whereas least reactivity was found against PEP07968 (DOTA-conjugated PEP07911, SEQ ID NO:159), PEP07914 (SEQ ID NO:158) and PEP07554 (SEQ ID NO:156).

(59) Thus, the most reactive albumin binding polypeptides were the parental G148-GA3 (PEP07913, SEQ ID NO:153) and the previously affinity improved derivative (PEP06923, SEQ ID NO:154), having helix 1 retained from G148-GA3. The third of the more reactive polypeptides (PEP07844, SEQ ID NO:161) contains the original lysine in position 14 in helix 1. This residue is intended for conjugation, and will therefore not be exposed in the final context. The identical albumin binding polypeptide variant, except for having an alanine in position 14 (PEP07554, SEQ ID NO:156), is one of the least reactive.

Example 5

Chemical Synthesis of a DOTA-Conjugated Albumin Binding Polypeptide

(60) Material and Methods

(61) The albumin binding polypeptide PEP07834 (SEQ ID NO:160) was synthesized by solid phase peptide synthesis (SPPS, as described by Quibell, M. & Johnson, T., in Fmoc Solid Phase Peptide Synthesis-A Practical Approach, W. C. Chan, P. D. White Eds, Oxford University Press 2000, 115-135) in a 433 A Peptide Synthesizer reactor (Applied Biosystems, Foster City, Calif.) on a 0.1 mmol scale, i.e. with a theoretical possible yield of 0.1 mmol peptide, using standard Fmoc chemistry. An acid-labile Fmoc amide resin was used as solid support throughout the synthesis (Rink Amide MBHA Resin LL (100-200 mesh), loading 0.39 mmol amide/g resin (Novabiochem)).

(62) 47 amino acid residues according to the sequence below were coupled to the amide resin by acylation reactions in the reactor for 10 minutes at room temperature (RT) and mixing. The acylation reactions were performed with a ten-fold molecular excess of Fmoc protected amino acids in NMP (N-methylpyrrolidone, Merck), activated with 1 eq of 2-(1H-benzotriazole-1-yl)-1,1,3,3-tetramethylaminium hexafluorophosphate (HBTU, IRIS Biotech), 1 eq of 1-hydroxybenzotriazole (HOBt, IRIS Biotech) and 2 eq of diisopropylethylamine (DIEA, Applied Biosystems). In addition, all reactive amino acid side chains were protected with standard side chain protection groups (tert-butyl (tBu) for Asp, Glu, Ser, Thr and Tyr, tert-butyloxycarbonyl (Boc) for Lys, 2,2,4,6,7-pentamethyldihydrobenzofuran-5-sulfonyl (Pbf) for Arg, and trityl (Trt) for Asn and Cys) prior to activation and coupling. In order to diminish the amount of incomplete couplings leading to truncated peptides, a minor amount of selected amino acid residues were subjected to coupling by acylation twice, without Fmoc deprotection as described below between the first and second coupling. The amino acid sequence of the synthesized albumin binding polypeptide PEP07834 was ALASAKEAAN AELDCYGVSD FYKRLIDKAK TVEGVEALKD AILAALP-NH.sub.2 (SEQ ID NO:160-NH.sub.2).

(63) The underlined amino acid residues were double coupled. Any remaining unreacted amino groups on the resin bound peptides were capped with acetic anhydride (0.5 M acetic anhydride (AlfaAesar), 0.125 M DIEA, 0.015 M HOBt in NMP) for 5 min. Following every coupling, deprotection of the N-terminal Fmoc group on the resin bound peptides were performed by treatment with 20% piperidine (Sigma-Aldrich) in NMP for 10 min.

(64) After completed synthesis, the peptides were cleaved from the solid support and simultaneously the side chain protection groups were cleaved off by treatment with TFA/EDT/H.sub.2O/TIS (94:2.5:2.5:1) (TFA: trifluoroacetic acid (Apollo), EDT: 1,2-ethanedithiol (Aldrich), TIS: triisopropylsilane (Aldrich)) at RT for 2 h with occasional mixing. After TFA treatment, the peptides were extracted three times using 20% acetonitrile (Merck) in water and tert-butyl methyl ether (Merck). The aqueous phases were combined, filtered and lyophilized.

(65) The crude peptides were analyzed and purified by semi-preparative RP-HPLC (Reprosil GOLD C18 300, 250*10 mm, 5 m particle size) and a gradient of 32-55% B (A: 0.1% TFA-H.sub.2O; B: 0.1% TFA-CH.sub.3CN) during 25 min at a flow rate of 2.5 ml min.sup.1, followed by lyophilization.

(66) The synthetic yield was determined by calculation of the integrated areas under the peaks from the 220 nm signal from the crude analysis on RP-HPLC. The correct molecular weight was verified using liquid chromatography electrospray ionization mass spectrometry (LC-ESI-MS) on a 6520 Accurate Mass Q-TOF LC/MS (Agilent Technologies). The purity of the product was verified using RP-HPLC (Reprosil GOLD C18 300, 250*4.6 mm, 3 m particle size) using a gradient of 35-55% B over 25 min at a flow rate of 1.0 ml min.sup.1.

(67) DOTA Conjugation

(68) 3 mg of PEP07834-amide (SEQ ID NO:160-amide) was reduced with 20 mM DTT at 40 C. for 30 minutes. Excess DTT was removed by buffer exchange on a PD-10 column (GE Healthcare) to 0.2 M ammonium acetate, pH 5.5. The coupling was performed with a 5-fold molar excess of chelator, maleimido-mono-amide-DOTA (Macrocyclics, Cat. No. B-272) solution in water (1 mg/ml). The mixture was incubated for 1 hour at 30 C. under continuous shaking. Purification from non-conjugated chelators was made on a semi-preparative RPC column (Zorbax 300SB C18, 9.4250 mm, 5 m). The coupling degree of the purified material was analyzed by HPLC-MS on a Zorbax 300SB C8 1502.1 mm, 3.5 m analytical column. Only maleimide-DOTA-conjugated PEP07834, denoted PEP07975, was detected by the method.

(69) Results

(70) Based on the elution profile of the crude material, the synthetic yield of the albumin binding polypeptide PEP07834-amide (SEQ ID NO:160-amide) was determined to be 8%. The found molecular weight was 4952.9 Da, which is in good agreement with the theoretical molecular weight calculated to 4952.6 Da. When analyzing the purified product, approximately 10-15% of the protein was found to be a disulfide linked homodimer (FIGS. 4 and 5). The binding activity of the DOTA-conjugated peptide (PEP07975) was confirmed as described in Example 2 (data not shown), and the melting temperature determined as described in Example 3.

Example 6

Immunogenicity Testing of Albumin Binding Polypeptides

(71) PEP07913 (SEQ ID NO:153), PEP07912 (SEQ ID NO:157), PEP07914 (SEQ ID NO:158), and PEP07968 (DOTA-conjugated PEP07911, SEQ ID NO:159) were screened for their ability to induce T cell proliferation in peripheral blood mononuclear cells (PBMC) from 52 human Caucasian individuals (obtained from CRI-Labo Medische Analyse, Gent, Belgium). PEP07913 corresponds to the sequence of G148-GA3 having an N-terminal glycine residue, whereas PEP07912, PEP07914 and PEP07968 are examples of the albumin binding polypeptides of PP007 (SEQ ID NO:7), PP013 (SEQ ID NO:13) and PP037 (SEQ ID NO:37) having different N-terminal amino acid additions according to the present disclosure.

(72) Materials and Methods

(73) PBMCs, prepared according to standard cell biological methods, were added to a tissue culture (TC) treated 96-well round bottom plate (Falcon) in an amount of 300 000 PBMCs/well. The cells were stimulated by addition of 100 l/well of albumin binding polypeptides PEP07913, PEP07912, PEP07914 and PEP07968 in AIMV medium (Invitrogen) additionally containing 900 g/ml (3-fold molar excess) of recombinant human albumin (ALBUCULT, Novozymes). This corresponded to a final concentration of albumin binding polypeptide of 30 g/ml. The stimulation was done in eight-plicates, i.e. the same albumin binding polypeptide were added to eight wells in identical amounts and under the same conditions. In positive control wells, the cells were stimulated with either 30 g/ml Keyhole Limpet Hemocyanin (KLH, Calbiochem) or 30 g/ml tetanus toxoid (TT, Statens Serum Institut). In negative control wells, only AIMV medium with or without 900 g/ml of albumin were added.

(74) Cell proliferation was assessed after seven days of culturing using Alexa Fluor 488 Click-iT EdU flow cytometry assay kit (Invitrogen). 1 M/well of EdU incorporation marker was added on day six. On day seven, cells were washed, dissociated from the plate, washed again and stained for 30 minutes with anti-CD3-PerCP reagent (Becton Dickinson) and anti-CD4-Alexa647 reagent (Becton Dickinson). Following staining, the cells were washed, fixed (BD cellfix, BD biosciences), permeabilized (using saponin) and stained for EdU by addition of Click-iT reagent according to the manufacturer's protocol (Invitrogen). After completed staining, cells were washed again and analyzed using flow cytometry (FACSCantoll, BD Biosciences). To assess the number of proliferating cells, a fixed number of fluospheres (Invitrogen) was added to each well before analysis. All staining procedures and washes were performed directly in the 96-well plate.

(75) The raw FACSCantoll data were gated hierarchically on CD3.sup.+ CD4.sup.+ T cells and the number of gated cells as well as their fluorescence intensity of EdU-Alexa Flour 488 incorporation marker were recorded. The mean values of the number of proliferating cells/eight-plicate of protein treated wells were compared to the positive and negative controls and the resulting ratios, described as stimulation indices (SI), were calculated. Based on the SI and the variation between replicates, threshold SI-values were set to 2.0 and 0.5 for stimulation and inhibition, respectively.

(76) Results

(77) The albumin binding polypeptides PEP07913, PEP07912, PEP07914 and PEP07968 were assessed for their immunogenic potential in the presence of 3-fold excess of recombinant human albumin in a target human population using an in vitro PBMC proliferation assay. Compared to the albumin control, PEP07913 induced CD3.sup.+ CD4.sup.+ T cells proliferation in 6 of 52 donors, PEP07912 in 5 of 52 donors and PEP07914 and PEP07968 in 1 of 52 donors (FIG. 6A).

(78) The mean stimulation index (SI) for all 52 donors was not significantly different for PEP07914 and PEP07968 compared to the negative control containing recombinant human albumin (p=0.79 and 0.48 respectively, FIG. 6B). The SI for PEP07913 was significantly higher (p=0.002) whereas the SI for PEP07912 was higher but not significant (p=0.03, FIG. 6B).

(79) As compared to buffer only, the number of responding individuals was 10 for PEP07912, 7 for PEP07912, 2 for PEP07914, 1 for PEP07968, 2 for recombinant human albumin, and 49 and 51 for the two positive controls TT and KLH, respectively (FIG. 6C). The albumin binding polypeptides were ranked according to their immunogenicity in the following order: PEP07913>PEP07912>PEP07914>PEP07968. Both PEP07914 and PEP07968 were defined as non-immunogenic. The above results thus demonstrate that the immunogenic potential of the albumin binding polypeptides of the present disclosure is low, as compared to the positive controls.

Example 7

Albumin Binding Polypeptides' Affinity to Albumin from Different Species

(80) In this example, PEP06923 (SEQ ID NO:154), PEP07986 (SEQ ID NO:163) and PEP08296, (DOTA-conjugated PEP08185, SEQ ID NO:148), expressed and purified as described in Example 1, were characterized for affinity to albumin from human (HSA), cynomolgus monkey (CSA), rat (RSA), mouse (MSA) and dog (DSA) using a Biacore instrument.

(81) Material and Methods

(82) Biosensor analysis on a Biacore2000 instrument (GE Healthcare) was performed with HSA (ALBUCULT, Novozymes), CSA (purified in-house from cynomolgus serum), RSA (Sigma-Aldrich, Cat. No. A6272), MSA (Sigma-Aldrich, Cat. No. A3559) and DSA (MP Biomedicals, Cat. No. 55925), immobilized by amine coupling onto the carboxylated dextran layer of the surfaces of CM-5 chips (research grade; GE Healthcare) according to the manufacturer's recommendations.

(83) On chip 1, surface 1 was activated and deactivated and used as a reference cell (blank surface) during injections, whereas surface 2 comprised HSA immobilized to 1130 resonance units (RU), surface 3 comprised CSA immobilized to 1046 RU, surface 4 comprised RSA immobilized to 831 RU. On chip 2, surface 1 was used as blank surface, whereas surface 3 comprised MSA immobilized to 858 RU. On chip 3, surface 1 was used as blank surface, whereas surface 2 comprised DSA immobilized to 1053 RU. For analysis of affinity for HSA, CSA, and RSA (chip 1), the purified albumin binding polypeptide variants were diluted in running buffer HBS-EP (GE Healthcare) to 40 nM, 10 nM and 2.5 nM; for analysis of affinity for MSA (chip 2) the albumin binding polypeptide variants were diluted to 1280 nM, 640 nM, 160 nM and 40 nM and for analysis of affinity for DSA (chip 3) albumin binding polypeptide variants were diluted to 1280 nM, 640 nM, 160 nM, 40 nM and 10 nM. The albumin binding polypeptides were injected at a constant flow-rate of 50 l/min for 5 minutes, followed by injection of HBS-EP for 60 minutes. The surfaces were regenerated with one injection of 25 l HCl, 10 mM. All samples were run in duplicates.

(84) The results were analyzed with a BIAevaluation software (GE Healthcare). Curves of the blank surface were subtracted from the curves of the ligand surfaces.

(85) Results

(86) The Biacore 2000 instrument has a technical limitation, hindering measurements of very high affinity. Hence, the purpose of the Biacore study was not to determine the exact kinetic parameters of the albumin binding polypeptide variants' affinity for HSA, CSA, RSA, MSA and DSA respectively. However, the results provide a quantitative estimation of the relative affinities of the enclosed polypeptides for albumin from these different species. After subtraction of reference surface and buffer injection, curves were fitted to a 1:1 (Langmuir) binding model using BIAevaluation software with correction for mass transfer and with RUmax set as a local parameter. Representative binding curves are shown in FIG. 7.

(87) PEP07986 and PEP08296 (DOTA-conjugated PEP08185) bind with high affinity (K.sub.D in the range from below picomolar to below nanomolar) to human serum albumin as well as to albumin from the frequent preclinical model species rat, cynomolgus monkey, mouse and dog. The relative affinities for the different species can be ranked as RSAHSA/CSA>MSA/DSA, i.e. the K.sub.D values ranked as K.sub.D-RSAK.sub.D-HSA/K.sub.D-CSA<K.sub.D-MSA/K.sub.D-DSA. The affinities in terms of K.sub.D values are the same or slightly lower (but in the same order of magnitude) as the affinity obtained for PEP06923 (non-inventive polypeptide).

Example 8

In Vitro Activity of Protein Z Variants Fused to an Albumin Binding Polypeptide

(88) In this example, polypeptides comprising cytokine-specific protein Z (derivative of domain B of staphylococcal protein A) variants genetically fused to the albumin binding polypeptide variant PP013 (SEQ ID NO:13) were tested for their functionality, this being to block cytokine-induced proliferation of TF-1 cells in the presence of human serum albumin. Proliferation of TF-1 cells is dependent of the presence of any of several different types of cytokines and the proliferative response can be inhibited by blocking reagents such as the corresponding cytokine-specific protein Z variant. PP013 fused to a protein Z variant with specificity for an irrelevant protein was used as negative control.

(89) Materials and Methods

(90) Cloning of Z-PP013 Fusion Proteins

(91) Gene fragments of protein Z variants with specificity for cytokine X or Y respectively, or for an irrelevant protein (negative control), were amplified by PCR using primers adding PstI and AccI specific endonuclease sites. The fragments were cleaved with PstI and AccI, purified and ligated into an expression vector, the plasmid pAY02747, restricted with the same enzymes. pAY02747 contains an origin of replication from pBR322, a kanamycin resistance gene and a T7 promoter for expression of the gene of interest. The cloning site is preceded by a sequence encoding the amino acids MGSSLQ and succeeded by a sequence encoding VDSS-PP013, where PP013 is the disclosed albumin binding polypeptide with SEQ ID NO:13. Ligation, transformation and sequence verification were performed as described above. The encoded proteins were: 1) MGSSLQ-Z.sub.X-VDSS-PP013 (denoted Z.sub.X-PP013) (SEQ ID NO:210) 2) MGSSLQ-Z.sub.Y-VDSS-PP013 (denoted Z.sub.Y-PP013) (SEQ ID NO:211) 3) MGSSLQ-Z.sub.neg-VDSS-PP013 (denoted Z.sub.neg-PP013) (SEQ ID NO:212)
Protein Expression

(92) Z.sub.X-PP013, Z.sub.YPP013 and Z.sub.neg-PP013 were expressed in E. coli BL21 (DE3) cells. Colonies from the transformations of each fusion variant were used to inoculate starter cultures of 50 ml TSB+YE medium supplemented with kanamycin to a concentration of 50 g/ml. The cultures were grown at 37 C. over night with agitation, 100 rpm. The starter cultures were then used to inoculate 900 ml TSB+YE medium supplemented with kanamycin to a concentration of 50 g/ml. The cultures were grown for approximately 1.5 h to an OD600 of >1.1, upon which IPTG was added to a final concentration of 0.2 mM. Cultivation was continued for five hours. Cell pellets were harvested by centrifugation (15600 g, 4 C., 20 minutes) and stored at 20 C. until purification.

(93) Protein Purification

(94) Frozen cell pellets harboring soluble fusion protein variants Z.sub.X-PP013, Z.sub.YPP013 and Z.sub.neg-PP013 were resuspended in 50 mM Tris-HCl, 150 mM NaCl, pH 8 and 1000 U BENZONASE (Merck Cat. No. 1.01654.0001) was added. The cells were disrupted by ultrasonication and for each of the fusion protein variants, the ultrasonicated suspension was clarified by centrifugation (15 min, 37000 g, 4 C.). 20TST-buffer (20[25 mM Tris-HCl, 1 mM EDTA, 200 mM NaCl, 0.05% Tween 20, pH 8.0]) was added at a volume resulting in 1TST buffer in the clarified suspension. Each sample of fusion protein variant was loaded onto a HSA-Sepharose column (GE Healthcare). After washing with TST-buffer, followed by 5 mM NH.sub.4Ac, pH 5.5, bound fusion protein variant was eluted with 0.5 M HAc, pH 2.5.

(95) The fusion protein variants were further purified by reversed phase chromatography (RPC), as follows. For each of the variants, the eluate from the HSA-affinity purification step was loaded on a 1 ml Resource 15 RPC column (GE Healthcare) previously equilibrated with RPC A Buffer (0.1% TFA in water). After column wash with 10 CV RPC A Buffer and 5 CV of RPC B Buffer (0.1% TFA in acetonitrile), bound fusion proteins were eluted with a linear gradient of 10-50% RPC B Buffer over 20 CV. The flow rate was 2 ml/min and the absorbance at 280 nm was monitored. Fractions containing pure fusion protein variants were identified by SDS-PAGE analysis and pooled. Finally, the buffer was exchanged to 1PBS (2.68 mM KCl, 137 mM NaCl, 1.47 mM KH.sub.2PO.sub.4, 8.1 mM Na.sub.2HPO.sub.4, pH 7.4) using a disposable PD-10 desalting column (GE Healthcare). To verify the identity of the fusion protein variants, SDS-PAGE and LC/MS analyses were performed as described in Example 1.

(96) In Vitro Cell Assay of Z-PP013 Fusion Proteins

(97) The cell line TF-1 (CLS Cat. No. 300434) was propagated as recommended by the provider in RPMI 1640 medium+10% fetal calf serum (Gibco) with the addition of 2 ng/ml of rhGM-CSF (Miltenyi). At the day of experiment, the cells were washed in RPMI 1640 medium+10% fetal calf serum to remove GM-CSF.

(98) The ability of Z.sub.X-PP013 and Z.sub.YPP013 to block cytokine induced proliferation was analyzed by mixing the molecules Z.sub.X-PP013, Z.sub.Y-PP013 and Z.sub.neg-PP013 with cytokines X and Y respectively, and with a five times molar excess of HSA (ALBUCULT, Novozymes). The molecules were titrated in a 2-fold dilution series with a fixed concentration of cytokine (4.9 M) and a five times molar excess of HSA. The titration was performed in 96-well plates in a volume of 100 l. 25 000 cells were added per well (100 l) and plates were incubated at 37 C., 5% CO.sub.2 for three days. To measure the proliferation, 19 l of CCK-8 cell proliferation reagent (Sigma) diluted two times in RPMI 1640 medium+10% fetal calf serum, was added per well. The color reaction was monitored after 4 hours using 96-well plate reader (Victor3; PerkinElmer).

(99) Results

(100) As shown in FIG. 8, both Z.sub.X-PP013 and Z.sub.YPP013 inhibited the respective cytokine induced proliferation in the presence of HSA whereas Z.sub.neg-PP013, the negative control, did not affect proliferation of TF-1. Thus, the experiment shows that the function of the Z molecules was retained when incorporated into a fusion protein containing the albumin binding polypeptide, and also when the fusion proteins were bound to albumin.

Example 9

Long-Term Stability of an Albumin Binding Polypeptide

(101) In this example, the stability of PEP07986 (SEQ ID NO:163), expressed and purified as described in Example 1, was investigated after storage at 4, 25, and 40 C. for up to three months. The status of the polypeptide after storage was investigated by measuring its binding to HSA using a Biacore instrument.

(102) Material and Methods

(103) Lyophilized PEP07986 was dissolved in sterile NaPi buffer (20 mM sodium phosphate, 150 mM sodium chloride, pH 7.2) at a concentration of 2 mg/ml. A reference sample (time=0) was removed and stored at 80 C. Aliquots of 105 l were stored in sterile screw-cap eppendorf tubes sealed with parafilm at 4, 25, and 40 C. After one week, two weeks, one month and three months, a sample stored at each temperature was cooled to 4 C., centrifuged for 5 min at 13000 rpm and then stored at 80 C. awaiting Biosensor analysis.

(104) Biosensor analysis was performed essentially as described in Example 2 but with HSA (ALBUCULT, Novozymes), immobilized to 704 resonance units (RU) and the albumin binding polypeptide variant was diluted to 10 nM and injected at a constant flow-rate of 20 l/min for 10 minutes, followed by injection of HBS-EP for 10 minutes.

(105) Results

(106) The binding to HSA of PEP07986 (SEQ ID NO:163) was retained after storage at 4, 25, and 40 C. for at least three months. The maximum binding responses to HSA obtained for PEP07986 stored at the various conditions are shown in FIG. 9.

Example 10

Stability of an Albumin Binding Polypeptide Under Extreme Conditions

(107) In this example, biosensor and circular dichroism (CD) analysis of the albumin binding polypeptide PEP08296 (DOTA-conjugated PEP08185, SEQ ID NO:148) after heat treatment (90 C.) in low pH (40) buffer is described. Since such extreme reaction conditions have to be used for example for .sup.68Ga labeling of DOTA-modified proteins, the influence of high heat and low pH treatment on the structural identity of the polypeptide and its capacity to bind HSA was investigated by measuring the melting temperature (Tm), refolding properties and binding to HSA.

(108) Material and Methods

(109) Biosensor Analysis of Heat Stability

(110) Biosensor analysis on a Biacore 2000 instrument (GE Healthcare) was performed with HSA (ALBUCULT, Novozymes) immobilized by amine coupling onto the carboxylated dextran layer of the surfaces of CM-5 chip (research grade; GE Healthcare) according to the manufacturer's recommendations. Surface 1 of the chip was activated and deactivated and used as a reference cell (blank surface) during injections, whereas surface 2 comprised HSA immobilized to 724 resonance units (RU). PEP08296 (50 l, 100 g) in a 15 ml Falcon tube was diluted with 450 l 0.2 M sodium acetate (NaAc) pH 5.5 to a final peptide concentration of 0.2 mg/mL. After addition of 1.5 ml 0.05 M HCl (resembling the conditions and volume used for eluting a .sup.68Ge/.sup.68Ga generator) the sample was incubated for 10 minutes at 90 C. or RT (control) and then transferred to RT. 6 ml 0.1 M sodium citrate was added to neutralize the pH. The heat treated PEP08296 (0.8 and 4 nM) was injected at a constant flow-rate of 50 l/min for 5 minutes, followed by dissociation in HBS-EP for 15 minutes. The surfaces were regenerated with one injection of 25 l 10 mM HCl. The results were analyzed with BIAevaluation software (GE Healthcare). Curves of the blank surface were subtracted from the curves of the ligand surfaces.

(111) Determination of the Melting Temperature (Tm)

(112) PEP08296 was dissolved in PBS to a final concentration of 0.5 mg/ml. PBS with a pH of approximately 4.0 was prepared by adding 9.5 l 100 mM HCl to 100 l PBS. Circular dichroism (CD) analysis was performed as described in Example 3.

(113) CD Analysis of Heat Stability

(114) To investigate structural reversibility of PEP08296 after heat treatment, two CD spectra between 195 and 250 were recorded per sample at 20 C. After the first spectrum, a VTM cycle with heating to 90 C. was run as described above followed by collection of the second CD spectrum between 195 and 250 nm at 20 C. In addition, PEP08296 was incubated in PBS pH 4.0 buffer or PBS pH 7.2 buffer for 12 minutes at 90 C. in a thermomixer (500 rpm, interval mixing 10 s on, 30 s off). After incubation, the samples were cooled on ice followed by centrifugation at 13000 rpm for 1 minute, and a CD spectrum between 195 and 250 nm was recorded at 20 C.

(115) Results

(116) Biosensor analysis was used to investigate if heat treatment in combination with low pH, i.e. common conditions needed for .sup.68Ga-labeling of polypeptide, would affect the capacity of PEP08296 to bind to HSA. FIG. 10 shows the result of this binding analysis performed with a Biacore 2000 instrument. Two different concentrations of PEP08296, 0.8 nM and 4 nM, were injected over a surface with 724 RU of immobilized human serum albumin. Heat treatment for 10 min at 90 C., pH 4.0, slightly reduced the binding capacity of PEP08296 to HSA, indicating a potential structural change of the molecule.

(117) CD was used to further investigate the potential structural change of the molecule. Similar CD spectra before and after heating would prove a sample to be structurally reversible. In the first experiment, the samples were heated with a temperature gradient from 20 C. to 90 C. The CD spectra before and after heat treatment were similar in the Tm determination experiment with the typical minima at 207 and 221 nm indicating -helicity, i.e. short time heating to 90 C. in either pH 4 or pH 7.2 buffer had no effect on the structure of PEP08296.

(118) However, pretreatment of PEP08296 for 12 minutes at 90 C. showed a slightly reduced alpha helix content of PEP08296 if incubated at pH 4.0, but no change in alpha helix content if incubated at pH 7.2. Typical overlays of two CD spectra before and after heating are shown in FIG. 11.

(119) The results from the melting temperature (Tm) determination are summarized in Table 4.

(120) TABLE-US-00005 TABLE 4 Tm of PEP08296 Designation Tm ( C.) PEP08296 at pH 7.2 59 PEP08296 at pH 4.0 62

Example 11

Blood Pool Imaging Using a 68Ga-Labeled Albumin Binding Polypeptide

(121) In the experiments making up this example, whole body distribution of .sup.68Ga-labeled PEP08296 (DOTA-conjugated PEP08185, SEQ ID NO:148) in rats was followed by dynamic imaging over 1.5 hours. Due to the strong association between the labeled polypeptide and serum albumin, the labeled polypeptide can be used for example to study blood pool and tissue permeability.

(122) Material and Methods

(123) .sup.68Ga-Labeling of PEP08296

(124) .sup.68Ga was eluted as .sup.68GaCl.sub.3 from the .sup.68Ge/.sup.68Ga generator (Obninsk, Russia) with 0.1 M HCl, converted to .sup.68GaCl.sub.4 with concentrated HCl, trapped on an anionic exchange column (Chromafix-HCO.sub.3) and subsequently eluted with 18 MO water, as previously described (Velikyan et al (2008), Nucl Med Biol 35:529-536).

(125) The labeling was performed essentially as described in Tolmachev et al. (EJNMMI 37:1356-1367, 2010). The concentrated .sup.68Ga-eluate (150-200 l) was added to PEP08296 (100 g in 0.2 M sodium acetate buffer pH 5.5) and the pH was adjusted to 3.5-4 using sodium acetate (1.25 M) or HCl (0.1 M). The labeling mixture was incubated at 90 C. for 15 min before cooling, and the labeled protein was isolated by size exclusion purification on a NAP-5 column eluted with physiologically buffered saline.

(126) The radiochemical purity and identity of the .sup.68Ga-labeled protein was assessed by radio-HPLC using UV (210 nm) and radioactivity detectors in series and a Superdex Peptide 10/300 GL column (GE Healthcare) eluted with physiologically buffered saline.

(127) Small Animal PET

(128) A rat (277 g) was anesthetized with isoflurane (initially 5%, then 2% blended with 7:3 air/O.sub.2), controlled by an E-Z vaporizer using Microflex non-rebreather masks from Euthanex Corporation, and was kept on a heating pad (37 C.) while lying within a microPET Focus120 system (Siemens, CTI Concorde Microsystems). .sup.68Ga-PEP08296, 33 MBq, was dispensed in a syringe, diluted with saline to 0.5 ml and injected via the tail vein. Data were acquired from the whole body by moving the bed in a constant bed motion protocol for 1.5 h. Data were processed with MicroPET Manager and corrected for randoms, dead time and decay. Images were reconstructed by standard 2D filtered back projection using a ramp filter and evaluated using Inveon Research Workplace (Siemens Medical Solutions) software.

(129) Results

(130) Basic distribution patterns (FIG. 12) for PEP08296 were very similar to that of albumin labeled with radioisotopes such as .sup.68Ga-DOTA, .sup.64Cu-DOTA and .sup.11C (see e.g. Hoffend et al (2005), Nucl Med Biol 32:287-292 and Lu et al (2008), [1-.sup.11C]Butanol and [Methyl-.sup.11C]Albumin for Blood Flow and Blood Pool Imaging, poster at the XIth Turku PET Symposium, 24-27 May 2008). In brief, high radioactivity concentrations were observed in major blood vessels throughout the scan. Organs with large blood volumes (liver, spleen and kidney) were also clearly delineated, as was the cardiac blood pool radioactivity. Radioactivity in the urinary bladder increased during the observation period, this observation of renal elimination being consistent with previous observations with labeled albumin-based tracers and with that of the metabolism of albumin itself.

(131) The general distribution pattern of radioactivity and very slow plasma clearance after intravenous injection of .sup.68Ga-PEP08296 is consistent with its expected very rapid and strong binding to albumin. These results therefore support further applications of the radiotracer as an in vivo blood pool imaging agent for use with positron emission tomography studies of tissue permeability, both during the development of disease and during therapeutic intervention.

Example 12

Solubility of an Albumin Binding Polypeptide

(132) The solubility of PEP07986 (SEQ ID NO:163) in physiological buffer was investigated by consecutive concentrations of the sample using ultrafiltration, followed by concentration measurement and investigation of aggregation status. Concentrations determined by direct absorbance readings at 280 nm were consistent with concentrations determined by gel filtration, showing a solubility of more than 42 mg/ml with no aggregates detected.

(133) Material and Methods

(134) Lyophilized PEP07986 was dissolved in NaPi buffer (20 mM sodium phosphate, 150 mM sodium chloride, pH 7.2) at a concentration of 3 mg/ml. Amicon Ultra centrifugal filter units, cut off of 3 kDa, (Millipore, Cat. No. UFC800324) were prerinsed with 2 ml NaPi buffer by centrifugation at 4000 g for 20 min in a swinging bucket rotor centrifuge (Multifuge, Heraeus). 1620 l of 3 mg/ml PEP07986 was applied to a first centrifugal filter unit and centrifugation was performed at 4000 g, 20 C., for 7 min. A 25 l sample was removed (UF sample 1) for further analysis and the rest of the sample was transferred to a second centrifugal filter unit. The centrifugation and sample removal were repeated three times with spinning times of 8, 9 and 20 min respectively (UF sample 2, 3 and 4 respectively). Absorbance readings were performed using a NANODROP ND-1000 Spectrophotometer and by diluting UF samples 1-4 in NaPi buffer 2, 4, 6 and 12 times respectively. The concentrations were calculated using the extinction coefficient 1 Abs 280=1.955 mg/ml. Gel filtration was performed on a 1100 HPLC system (Agilent Technologies) using a Superdex75 10/300 GL column (GE Healthcare) which had been equilibrated in NaPi buffer. 10 l of each UF sample were applied to the column; NaPi buffer was used as running buffer and the flow rate was 0.5 ml/min. A chromatogram of the molecular weight standard ovalbumin (GE Healthcare), injected at a concentration of 5 mg/ml was collected as well. Concentrations were determined by integrating the area under the curve.

(135) Results

(136) Concentrations determined by direct absorbance readings at 280 nm and concentrations determined by gel filtration are shown in Table 5. The solubility of PEP07986 (SEQ ID NO:163) is at least 42 mg/ml in physiological buffer (20 mM sodium phosphate, 150 mM sodium chloride, pH 7.2). No aggregates were detected by gel filtration, as shown by FIG. 13.

(137) TABLE-US-00006 TABLE 5 Concentrations determined after consecutive concentration of PEP07986 (SEQ ID NO: 163) Concentrations (mg/ml) determined by Sample Spectrophotometer Gel filtration UF2 12.1 12.4 UF3 22.2 22.1 UF4 42.7 42.6