Single domain binding molecule

Abstract

The present invention provides a single domain specific binding molecule having the structure
FW1-CDR1-FW2-HV2-FW3a-HV4-FW3b-CDR3-FW4
in which the Framework Regions FW1, FW2, FW3a, FW3b, and FW4, the Complementarity Determining Regions CDR1 and CDR3, and the Hypervariable Regions HV2, and HV4 have amino acid sequences as defined which provide a high affinity anti-human serum albumin (HSA) binding domain.

Claims

1. A fusion protein comprising: A) a single domain specific binding molecule comprising the Framework Regions FW1, FW2, FW3a, FW3b, and FW4, in which: i) FW1 comprises TRVDQSPSSLSASVGDRVTITCVLT (SEQ ID NO: 28), ARVDQSPSSLSASVGDRVTITCVLR (amino acids 1-25 of SEQ ID NO: 67) or TRVDQTPRTATRETGESLTINCVLT (SEQ ID NO: 1), ii) FW2 comprises TYWYQQKPGS (SEQ ID NO: 29), TCWYQQKPGK (amino acids 34-43 of SEQ ID NO: 67), or TYWYRKNPGS (SEQ ID NO: 2), iii) FW3a comprises GRYSESVN (SEQ ID NO: 30), or GRYVESVN (SEQ ID NO: 3), iv) FW3b comprises FTLTISSLQPEDFATYYCRA (SEQ ID NO: 31), FTLTISSLQPEDFATYYCGL (amino acids 65-84 of SEQ ID NO: 67) or FSLRIKDLTVADSATYICRA (SEQ ID NO: 4), and v) FW4 comprises GAGTKVEIK (SEQ ID NO: 32), CGQGTKVEIK (amino acids 103-112 of SEQ ID NO: 67) or GAGTVLTVN (SEQ ID NO: 5); or B) a single domain specific binding molecule comprising framework regions FW1, FW2, FW3a, FW3b, and FW4 in which: i) FW1 comprises a sequence having at least 50% identity to TRVDQSPSSLSASVGDRVTITCVLT (SEQ ID NO: 28), ARVDQSPSSLSASVGDRVTITCVLR (amino acids 1-25 of SEQ ID NO: 67) or TRVDQTPRTATRETGESLTINCVLT (SEQ ID NO: 1), ii) FW2 comprises a sequence having at least 50% identity to TYWYQQKPGS (SEQ ID NO: 29), TCWYQQKPGK (amino acids 34-43 of SEQ ID NO: 67, or TYWYRKNPGS (SEQ ID NO: 2), iii) FW3a comprises a sequence having at least 50% identity to GRYSESVN (SEQ ID NO: 30), or GRYVESVN (SEQ ID NO: 3), iv) FW3b comprises a sequence having at least 50% identity to FTLTISSLQPEDFATYYCRA (SEQ ID NO: 31), FTLTISSLQPEDFATYYCGL (amino acids 65-84 of SEQ ID NO: 67) or FSLRIKDLTVADSATYICRA (SEQ ID NO: 4), and v) FW4 comprises a sequence having at least 50% identity to GAGTKVEIK (SEQ ID NO: 32), CGQGTKVEIK (amino acids 103-112 of SEQ ID NO: 67) or GAGTVLTVN (SEQ ID NO: 5).

2. The fusion protein as claimed in claim 1, which is humanized.

3. The fusion protein of claim 1, wherein the single domain specific binding molecule has an amino acid sequence selected from the group consisting of: E06 (SEQ ID NO. 46), BB11 (SEQ ID NO. 85) and H08 (SEQ ID NO: 36), or a sequence having at least 50% identity thereto.

4. The fusion protein of claim 1, wherein the single domain specific binding molecule has an amino acid sequence selected from the group consisting of: huE06 v1.1 (SEQ ID NO. 68), huH08 v1.1 (SEQ ID NO. 75), huE06 v1.2 (SEQ ID NO. 69), huE06 v1.3 (SEQ ID NO. 70), huE06 v1.4 (SEQ ID NO. 71), huE06 v1.5 (SEQ ID NO. 72), huE06 v1.6 (SEQ ID NO. 76), huE06 v1.7 (SEQ ID NO. 73), huE06 v1.8 (SEQ ID NO. 77), huE06 v1.9 (SEQ ID NO. 78), huE06 v1.10 (SEQ ID NO. 74, AC9 (SEQ ID NO. 79), AD4 (SEQ ID NO. 80), AG11 (SEQ ID NO. 81), AH7 (SEQ ID NO. 82), BA11 (SEQ ID NO. 83), BB10 (SEQ ID NO. 84), BC3 (SEQ ID NO. 86), BD12 (SEQ ID NO. 87), BE4 (SEQ ID NO. 88), or BH4 (SEQ ID NO. 89), or a sequence having at least 50% identity thereto.

5. The fusion protein of claim 1, in which the fusion protein further comprises a biologically active agent fused to the single domain specific binding molecule.

6. The fusion protein of claim 5 wherein the biologically active agent is a pharmaceutically active agent.

7. The fusion protein of claim 5, wherein the biologically active agent is a chemical compound or polymer.

8. The fusion protein as claimed in claim 6, wherein the pharmaceutically active agent is selected from the group consisting of a toxin, anti-inflammatory drug, antibiotic, anti-cancer drug, and analgesic.

9. The fusion protein as claimed in claim 6, wherein the pharmaceutically active agent is a protein.

10. The fusion protein as claimed in claim 9, wherein the protein is an enzyme or fragment thereof, or a cytokine or fragment thereof.

11. The fusion protein as claimed in claim 8, wherein the toxin is cholera toxin or a non-protein toxin.

12. The fusion protein as claimed in claim 8, wherein the anti-inflammatory agent is a non-steroidal anti-inflammatory drug (NSAID) or steroid.

13. The fusion protein as claimed in claim 8, wherein the anti-cancer drug is cytotoxic or cytostatic.

14. The fusion protein as claimed in claim 7, where the chemical compound or polymer is a substance suitable to extend the half-life of the fusion protein in vivo.

15. The fusion protein as claimed in claim 14, wherein the chemical compound or polymer is polyethylene glycol (PEG) and/or cyclodextrin.

16. The fusion protein as claimed in claim 5, wherein the biologically active agent is linked via a linker moiety to the single domain specific binding molecule.

17. The fusion protein as claimed in claim 16, wherein the linker moiety is a peptide, peptide nucleic acid or polyamide linkage.

18. The fusion protein as claimed in claim 5, wherein the single domain specific binding molecule and biologically active agent are joined together by chemical means.

19. The fusion protein as claimed in claim 5, wherein the single domain specific binding molecule and biologically active agent are joined together by peptide bonds through protein synthesis.

20. The fusion protein as claimed in claim 5, wherein the single domain specific binding molecule and biologically active agent are N-, C- and/or N-/C-terminal fusion(s).

21. A nucleic acid encoding the fusion protein of claim 9.

22. A nucleic acid construct comprising a nucleic acid as claimed in claim 21.

23. A host cell comprising a vector as claimed in claim 22.

24. A process for the production of a fusion protein comprising the step of expressing the nucleic acid sequence of claim 21 in a host cell.

25. The process as claimed in claim 24, further comprising the step of purifying the expressed fusion protein.

26. The process as claimed in claim 25, wherein the fusion protein is humanised.

27. A process for the production of a fusion protein of claim 7, wherein the fusion is by chemical means.

28. The process as claimed in claim 27, wherein the chemical means include hydrogen bonds, salt bridges, chemical cross-linking or any combination thereof.

29. A process for the production of a fusion protein comprising the step of expressing the nucleic acid sequence encoding the single domain specific binding molecule of claim 1 in a host cell, and fusing the expressed protein to a chemical compound or polymer.

30. The process as claimed in claim 29, wherein the single domain specific binding molecule is humanised.

31. The process of claim 29, wherein the fusion is chemical synthetic or biosynthetic.

32. The process of claim 31, wherein the biosynthetic fusion is enzymatic.

33. A pharmaceutical composition comprising a fusion protein as claimed in claim 1.

34. The pharmaceutical composition as claimed in claim 33, further comprising at least one of a pharmaceutically acceptable carrier, a pharmaceutically acceptable diluent, a pharmaceutically acceptable adjuvant, and a pharmaceutically acceptable buffer solution.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIGS. 1A to 1C show the amino acid sequences of the natural specific binding domains of the invention: a) an alignment of the anti-HSA binding domains isolated after round 2 and 3 of selection (SEQ ID NOs: 33-50); b) (i) the amino acid of the specific binding domain E06 (SEQ ID NO: 46); (ii) a comparison of sequence E06 (SEQ ID NO: 46) and BB11 (SEQ ID NO: 85); and (iii) sequence E06 with a 3×Ala-6×HIS tag (AAA-6×His tag is in italic) (SEQ ID NO: 51); c) sequence alignment of anti-HSA binding domains isolated from the pre-selected phage display library (SEQ ID NOs: 52-64). (Shading (with solid grey) indicates residues that differ from the consensus in FIG. 1(a) and 1(c)).

(2) FIGS. 2A to 2C: (a) shows structural sequence alignment of VNAR E06 (SEQ ID NO: 46) and its humanized variants of the invention (SEQ ID NOs: 65-74) including human germline V-kappa light chain DPK9/JK1 (SEQ ID NO: 65), shark VNAR 5A7 (SEQ ID NO: 66), and a humanized 5A7 variant, 5A7-IVabc (SEQ ID NO: 67). The residue numbering refers to E06 sequence; (b) individual amino acid sequences of the humanized VNAR domains of the invention: 5A7 (5A7 IVabc) (SEQ ID NO: 67), humanized versions of the anti-HSA domain E06 versions 1.1 through to 1.10 and H08 version 1.1 (SEQ ID NOs: 68-78); (c) alignment of (SEQ ID NO: 46), humanized E06v1.10 (SEQ ID NO: 74), and improved versions of the humanized E06 v1.10 of the invention (SEQ ID NOs: 79-89).

(3) FIGS. 3A and 3B show: (a) a ribbon model representation of the crystal structure of E06 in complex with HSA; (b) sequence of E06 (SEQ ID NO: 46) highlighting-, CDR1, CDR3, HV2 and HV4 and framework regions. Residues identical to human V kappa framework DPK9 are shown in bold. Residues within 5A of HSA in crystal structures, and involved in various modes of interactions with HSA, are marked by arrows. Humanized sequences v1.1 (SEQ ID NO: 68) and v1.10 (SEQ ID NO: 74) are also aligned so illustrate the re-introduction of contact residues in v1.10 post crystal dataset.

(4) FIGS. 4A to 4E show ability of E06 to extend the serum half-life of an unrelated domain across three species of PK model; (a) in vivo PK analyses of E06 and H08 in fusion with 2V as measured via iodination, 2V alone and PEGylated 2V as a PK control; (b) (i) LC-MS measurements of half-life of 2V-E06 and (ii) 2V-E06-2V in a murine model of PK both intra-venous and subcutaneous administration; (c) E06-2V and 2V-E06 in a rat model of PK by intra-venous administration; (d) E06-2V and 2V-E06 both intra-venous and subcutaneous administrations in a cynomologus monkey model of PK with E06-2V also as subcutaneous administration; (e) allometric scaling of 2V-E06 based on data attained and compared to half-life of albumin in each species.

(5) FIGS. 5A and 5B show the retention of the therapeutic domains A1 and CC3 to still bind target and retain functionality when in fusion to E06; (a) cell neutralization assays showing the ability of anti-mICOSL domains A1 and CC3 to still prevent ligand-receptor binding; (b) T-cell proliferation assays showing the ability of A1 and CC3 to retain the ability to inhibit T-cell proliferation when in complex with E06 (i) (hFc controls shown in (ii).

(6) The invention will also be further described by way of reference to the following Examples which are present for the purposes of illustration only and are not to be construed as being limiting on the invention.

(7) The abbreviations used are:

(8) HSA, BSA, MSA, RSA—human, bovine, mouse and rat serum albumin, respectively; HBS, HEPES-buffered saline; HEL, hen egg lysozyme; CDR, complementarity determining region; CM, conditioned medium; FW, framework region; HV, hypervariable region; HRP, horseradish peroxidase; MME, monomethyl ether; PBS, phosphate-buffered saline; PEG, polyethylene glycol; RMSD, root mean squared deviation; LC-MS, liquid chromatography mass spectrometry.

EXAMPLE 1

(9) Isolation of human serum albumin-binding shark VNARs

(10) As a prerequisite to this work, a sequence database of approximately 1600 VNAR sequences from the Spiny dogfish (Squalus acanthias) were compiled, aligned and analysed to facilitate the design of primer pairs to capture the immune repertoire. This ensured the subsequent immune phage libraries constructed were as representative as possible and lacked bias toward certain isotypes. Spiny dogfish were immunized with human serum albumin and VNAR sequences from a seropositive animal were isolated and made into a phage display library. The library was rescued and a solid phase sequential bio-panning strategy against 50 □g/ml, 50 g/ml and 0.5 □g/ml HSA was employed. After three rounds several high affinity, highly similar clones against albumin were isolated (FIG. 1(a)). Binding at both pH 7.0 and pH 6.0 (to facilitate lyzosomal recycling) was incorporated as part of the screening strategy. One of the best binders, which reacted to HSA, MSA and RSA at pH 7.0 and pH 6.0, was called E06 (FIG. 1(b)).

(11) VNAR clone E06 is 103 residues in length and has low sequence similarity to human variable domain sequences (<30% identity), with the closest human germline sequences being from VL6 and VH4 families. E06 sequence is a typical shark VNAR lacking the CDR2 region of mammalian antibody V domains and carrying instead a HV2 stretch in FW2 and also HV4 loop as part of FW3 sequence. Similar to other Ig molecules, there is a 6-amino acid CDR1 sequence and relatively short 9-amino acid CDR3 sequence (FIG. 1(b)). E06 belongs to a structural type IV of shark VNARs, which is distinct from better-characterized type I (e.g. 5A7 (Dooley et al, Mol Immunol 40(1) 25-33 (2003)) and type II (e.g. PBLA8; a phage-display library clone from HEL-immunized nurse shark) (Dooley et al, PNAS 103(6) 1846-51 (2006))). Type IV VNARs have only 2 canonical Ig domain cysteine residues (positions 22 and 83 in E06), compared to 6 cysteines in type I and 4 cysteines in type II.

(12) The immunized library was also screened prior to panning which resulted in the isolation of several more diverse anti-HSA VNAR domains (FIG. 1(c)). The response was sufficiently robust that selection was not necessary as the unselected library showed 16% positive binding to HSA.

EXAMPLE 2

(13) Humanization Strategy for Type I and Type IV VNARs

(14) As an initial strategy the well characterized type I VNAR, 5A7 was humanized. To this end, we selected as a template the human Ig variable light domain (V01) germline sequence, DPK9, which was considered structurally the closest to 5A7 of human Ig variable domain sequences. 5A7 was humanized by resurfacing, whereby multiple solvent-exposed as well as core framework residues of 5A7 were replaced by human DPK9/J01 residues (FIG. 2(a)). Specifically, the following structural elements were replaced in 5A7: FW1 (residues 6-21), FW2/part of HV2 (residues 38-47), FW3b (residues 67-82) and FW4 (residues 106-113). All six cysteine residues of type I VNAR scaffold were retained. The resulting molecule, which we call 5A7-IVabc, has 60 out of 86 (69.8%) of non-CDR residues, and 60 out of 112 (53.6%) of all residues identical to DPK9

(15) To further validate the humanization-by-resurfacing approach taken with 5A7, we created a series of humanized variants of E06 using 5A7-IVabc molecule as a guide as well as using the crystal structure to ensure critical interface residues remained within the humanized variants to retain function of the original domain (FIG. 2(a)). To make humanized E06 variant 1.1 (huE06 v1.1), 30 residues out of 103 total in E06 were replaced with DPK9 residues. Specifically, the majority of framework residues: FW1 (residues 6-21), FW2 (residues 38-40), FW3b (residues 66-82), and FW4 (residues 99-103) of E06 were made identical to DPK9. The majority of these changes parallel those used to make 5A7-IVabc. The regions left intact (shark) were first 4 amino-terminal residues, CDR1 (residues 28-33) and CDR3 (residues 86-94), HV2 (residues 43-52), FW3a and HV4 (residues 53-65). In huE06 v1.1 molecule, 54 out of 85 (63.5%) of non-CDR residues are identical to human DPK9.

(16) To modify E06 further, mutations toward 5A7-IVabc sequence were introduced into HV4 region of v1.1 (K61S and T63S) to make huE06 v1.2. Further DPK9-like changes were made in HV2 region (.sup.43SSNKE.sup.47.fwdarw..sup.43KAPK.sup.46) (SEQ ID NOs: 99 & 100) to produce huE06 v1.7. These two sets of changes (in HV2 and HV4) were combined to make huE06 v1.3. A derivative of v1.3, which had its N-terminus changed toward DPK9 (.sup.1TRVD.sup.4 to .sup.1DIQMT.sup.5) (SEQ ID NOs: 101 & 102), was made and named huE06 v1.4. We also attempted to redesign E06 by shortening the FW3a/HV4 region; to do that, 5 shark residues were deleted and 3 DPK9 residues introduces in this area of huE06 v1.3 molecule; in addition, Y55F change was introduced. The resulting molecule was named huE06 v1.5. Finally, huE06 v1.10 was derived from v1.1 by restoring .sup.38RKN.sup.40 shark sequence in FW2 from DPK9-like .sup.38QQK.sup.40.

(17) During this step wise process, a total of 10 humanized derivatives of E06 were designed, expressed and characterized—these sequences are listed in FIG. 2(b). Additional variants of v1.10 with improved biophysical characteristics were constructed using random mutagenesis, a phage display library built from these variants and screened on the basis of retention to bind HSA. Briefly, hE06v1.10 sequence was cloned into a phagemid vector and was mutated by error-prone PCR aiming at up to 9 substitutions/VNAR sequence using a GeneMorph II random mutagenesis kit (Agilent Technologies, Santa Clara, Calif.) according to the manufacturer's instructions. The resultant mutational phage display libraries were rescued and selected twice using Nunc Maxisorp immunotubes. Following each pan, two 96-well plates of individual colonies were picked with a QPix2 XT (Genetix, San Jose, Calif., USA). Binding as monoclonal phage and periprep was evaluated by ELISA. All samples were processed with a Perkin Elmer MiniTrak robotic liquid handling system (Waltham, Mass., USA). Unique clones showing OD450 by periprep ELISA at least 25% higher than the readings obtained from parental hE06v1.10 were selected for transfer to a eukaryotic expression vector. Following preliminary screen, clones exhibiting the best binding properties were selected for expression transiently in 1 litre cultures of HEK293 cells and monomeric 6×His tagged VNAR proteins purified by IMAC and cation IEX chromatography. The clones with the best biophysical propertied based on HSA binding and no evidence of aggregation are listed in FIG. 2(c).

EXAMPLE 3

(18) Structure of E06 in Complex with HSA

(19) E06 was expressed as a monomeric 6×His-tagged protein and crystallized in complex with HSA and the structure was determined to 3.0 Å (FIG. 3(a)).

(20) Protocols: Crystallization

(21) E06:HSA crystals were grown by hanging drop vapor diffusion at 18° C. in drops containing 1.0 μl protein stock solution (11.0 mg/ml protein complex, 25 mM Tris pH 7.4, 150 mM NaCl) mixed with 1.0 μl well solution (16% PEG 2000 MME, 100 mM sodium acetate pH 4.6) and equilibrated against 0.5 ml of well solution. Chunky crystals grew in approximately one week, measuring ˜50 μm across.

(22) Data Collection and Processing

(23) E06 complex crystals belong to the space group P3221 with unit cell parameters 127.98×127.98×151.76 Å, and contain two molecules of E06 and two molecules of HSA in the asymmetric unit, implying a solvent content of 54.2%. Crystals were drawn through a solution of 20% DMSO and 80% well solution, and cooled rapidly in liquid nitrogen. Diffraction data were recorded at APS beamline 22-ID on a MAR-300 detector. Intensities were integrated and scaled using the program Xia2.

(24) Phasing, Model Building, and Refinement

(25) The structure of E06 in complex with HSA was determined by molecular replacement with PHASER using the crystal structure of apo HSA (PDB ID: 1AO6) as a starting search model. A few rounds of refinement with Phenix were performed, after which clear density for the p-sheet regions of EO6 was obtained. After subsequent placement of a poly-alanine model of EO6 and several iterative cycles of model rebuilding with Coot and refinement with autoBuster, the final Rwork and Rfree values of 23.73% and 26.63% were obtained. In contrast to the classical antigen-antibody recognition mode, it was found that most extensive interactions with HSA originate from the CDR3 residues and framework residues on E06 (FIG. 3(b)). Antigen binding results in a large buried surface area of 705 Å, which is 12.5% of total surface area of E06. The interaction elucidated is also unusual for a VNAR due to the planar nature of the interface and inclusion and contribution of framework residues in the antigen-antibody complex.

EXAMPLE 4

(26) Functional properties of humanized E06 variants

(27) 5A7-IVabc protein showed excellent expression profile in mammalian cells with very little aggregation in either monomeric or dimeric (with human Fc) format and full retention of binding activity to HEL. The binding constant for monomeric VNARs to HEL, as determined by Biacore, was 13.6 nM for parental 5A7 and 14.8 nM for humanized 5A7-IVabc (Table 1). The results show the retention of binding antigen by humanized VNAR domains of the invention as measured by BIAcore against target; (Table 1) shows retention of binding of humanised 5A7 to hen egg lysozyme (HEL); (Table 2 and 3) shows retention of albumin binding by humanized E06 v1.1 and v1.10 to human, mouse and rat albumin at pH 7.4 and pH 6.0.

(28) TABLE-US-00001 TABLE 1 ka kd KD KD Rmax Chi.sup.2 % surface Sample (1/Ms) (1/s) (M) (nM) (RU) tc (RU.sup.2) active 5A7-AAA-6xHis 4.48E+06 6.09E−02 1.36E−08 13.6 61.76 5.47E+07 0.31 65.0% 5A7IVabc-AAA-6xHis 1.17E+07 1.73E−01 1.48E−08 14.8 69.67 6.26E+07 0.272 74.1%

(29) E06 and its humanized variants were first expressed as human Fc fusions. The expression levels of the humanized variants, such as 1.1, 1.2, 1.3, 1.5 and 1.7, were not dramatically different from the parental E06 molecule and were in 10-40 μg/ml range in transient COS-1 system.

(30) To assess the kinetic parameters of binding of E06, huE06 v1.1 and huE06 v1.10 to serum albumins, the monomeric (6×His-tagged) V-NARs were tested in BIAcore experiments. As shown in Table 2 (at pH 7.4) and Table 3 (at pH 6.0), the humanized versions retained the ability to bind mammalian albumin species with nanomolar affinities. Increased affinity of the humanized versions was seen at the lower pH values compared of wild type E06.

(31) TABLE-US-00002 TABLE 2 ka kd KD Fold change Rmax % Surface Chi.sup.2 IgNAR Albumin (1/Ms) (1/s) (nM) from E06 (RU) Active (RU.sup.2) E06 human 3.48E+06 6.55E−04 0.19 n/a 86.18 85.3% 0.81 huE06v1.1 human 8.72E+05 1.40E−02 16.0 84.9 78.6 77.8% 0.234 huE06v1.10 human 5.59E+05 3.57E−03 6.4 33.9 80.62 79.8% 0.181 E06 mouse 2.18E+06 1.80E−03 0.83 n/a 109.9 85.2% 0.871 huE06v1.1 mouse 8.41E+05 4.14E−02 49.2 59.4 100.9 78.2% 0.354 huE06v1.10 mouse 4.12E+05 9.59E−03 23.2 28.1 100.9 78.2% 0.26 E06 rat 2.21E+06 3.20E−03 1.45 n/a 93.71 88.4% 0.687 huE06v1.1 rat 4.36E+05 3.29E−02 75.3 52.0 77.32 72.9% 0.616 huE06v1.10 rat 4.49E+05 1.65E−02 36.8 25.4 84.03 79.3% 0.311

(32) TABLE-US-00003 TABLE 3 ka kd KD Fold change Rmax % Surface Chi.sup.2 IgNAR Albumin (1/Ms) (1/s) (nM) from E06 (RU) Active (RU.sup.2) E06 human 1.12E+07 1.59E−03 0.14 n/a 50.1 49.6% 0.254 huE06 v1.1 human 3.62E+06 1.84E−02 5.07 35.6 29.3 29.0% 0.196 huE06 v1.10 human 3.52E+06 6.37E−03 1.81 12.7 39.7 39.3% 0.334 E06 mouse 9.85E+06 3.23E−03 0.33 n/a 44.1 34.2% 0.259 huE06 v1.1 mouse 3.43E+06 3.18E−02 9.28 28.3 21.0 16.3% 0.143 huE06 v1.10 mouse 3.14E+06 1.25E−02 3.98 12.2 31.7 24.5% 0.182 E06 rat 1.26E+07 4.33E−03 0.34 n/a 41.8 39.5% 0.244 huE06 v1.1 rat 3.12E+06 3.10E−02 9.94 29.0 18.5 17.4% 0.118 huE06 v1.10 rat 3.65E+06 1.52E−02 4.17 12.2 29.1 27.4% 0.16

EXAMPLE 5

(33) E06 Extends Plasma Half-Life of Unrelated Fusion Proteins In Vivo

(34) The unrelated naïve VNAR domain, 2V, was identified as a type IV during the initial database acquisition of sequences. It has no known binding partner and was therefore chosen as a suitable “dummy” protein partner to study the PK and PD of E06 in several animal models. Molecular fusions of E06 with N-terminal, C-terminal and dual terminal constructs were created using G4S linker sequences bridging the VNAR domains, and C-terminal AAA-6×HIS tags for purification purposes. Dimers and trimers were expressed in HEK293 cells and purified using standard Ni-NTA methods and SEC as a final polishing step. Proteins were assessed for rodent viruses and endotoxin levels prior to use in animal models. BIAcore analysis was conducted to determine the affinities of each fusion variant compared to wild-type E06. Table 4 shows the on, off rates and KD affinity values for wild-type, 2V-E06, E06-2V and 2V-E06-2V fusions with 2V control and the anti-HEL 5A7 as an additional control. In particular Table 4 shows the BIAcore analyses of E06 alone, as an N-terminal, C-terminal and dual fusion construct with 2V, 2V alone and 5A7 as control binding to HAS.

(35) TABLE-US-00004 TABLE 4 mean ka ± SE mean kd ± SE mean KD ± SE Sample Ligand (×10.sup.6 M.sup.−1 s.sup.−1) (×10.sup.−4 s.sup.−1) (nM) n E06 HSA 3.092 ± 0.034  5.825 ± 0.103 0.189 ± 0.005 4 CSA 2.675 ± 0.023  5.845 ± 0.185 0.219 ± 0.005 2 MSA 2.316 ± 0.010 17.130 ± 0.360 0.740 ± 0.012 2 RSA 2.240 ± 0.058 30.110 ± 0.110 1.345 ± 0.030 2 HEL — — — 2 E06-2V HSA 7.464 ± 0.384  5.307 ± 0.147 0.071 ± 0.002 4 CSA 8.996 ± 0.841  5.915 ± 0.370 0.066 ± 0.002 2 MSA 5.299 ± 0.463 14.695 ± 0.415 0.279 ± 0.017 2 RSA 5.553 ± 0.452 27.510 ± 0.310 0.498 ± 0.035 2 HEL — — — 2 2V-E06 HSA 0.953 ± 0.097  6.412 ± 0.116 0.721 ± 0.095 6 CSA 0.820 ± 0.138  6.821 ± 0.163 0.910 ± 0.157 4 MSA 0.862 ± 0.018 21.420 ± 0.150 2.486 ± 0.036 2 RSA 0.623 ± 0.091 34.418 ± 0.550 5.949 ± 0.984 4 HEL — — — 2 2V-E06-2V HSA 0.689 ± 0.064  5.762 ± 0.068 0.896 ± 0.091 8 CSA 0.592 ± 0.069  5.904 ± 0.186 1.052 ± 0.155 4 MSA 0.507 ± 0.047 19.783 ± 0.473 4.032 ± 0.472 4 RSA 0.448 ± 0.041 31.353 ± 0.919 7.245 ± 0.934 4 HEL — — — 4 2V HSA — — — 4 CSA — — — 2 MSA — — — 2 RSA — — — 2 HEL — — — 2 5A7 HSA — — — 2 CSA N.A. N.A. N.A. — MSA — — — 2 RSA N.A. N.A. N.A. — HEL 1.543 ± 0.025 468.450 ± 11.350 30.365 ± 0.255  2

(36) The data shows that E06 tolerates bath N-terminal and C-terminal fusions creating dimers proteins in addition to both N and C terminal fusion constructs creating trimers. All constructs retain high affinity binding to target (HSA) and across other albumin species (mouse, rat and monkey). Note that in Table 4, “-” indicates no binding and N.A. that the experiment was not conducted. All three constructs were tested in animal model of PK to measure the ability of E06 to extend plasma half-life of an unrelated partner protein. As an initial study, EOG (SEQ ID NO: 46) and the related protein H08 (sequence listed in FIG. 1(a) as P2_H08, SEQ ID NO: 36, was isolated from same panning strategy as E06) were fused to 2V and compared to the clearance of 2V alone. Purified E06-2V and H08-2V were studied in a murine single dose, PK model to determine the serum half-life of 2V as an independent domain or in complex with the anti-HSA VNAR domains. To ensure iodination did not affect HSA binding, E06-2V and H08-2V were “cold” iodinated showing that this did not interfere with E06 binding HSA.

(37) Male C57BL/6 mice were injected i.v. with the following doses of protein: 1 mg/kg for 2V and 2V-PEG (2×NOF) and 0.3 mg/kg for the tandem VNARs (E06-2V and H08-2V). The concentrations for tandem VNARs were scaled for 1 mg/kg dosage and the dose for 2V-PEG was based on protein only. Radioactive equivalent concentrations were determined by gamma-counting. Individual concentration values <LOQ (defined as 3*background cpm) were treated as zero for calculations of the mean and SD; N=6 per time point. Plasma concentration are illustrated in FIG. 4(a) and PK parameters summarized in Table 5 which shows PK and PD measurements showing increase in half-life of E06 fusion with 2V compared to 2V alone.

(38) TABLE-US-00005 TABLE 5 AUC.sub.0-∞ AUC.sub.0-∞/Dose Compound Dose C.sub.max.sup.a (μg (μg eq. hr/mL)/ CL Vd.sub.ss t.sub.1/2 (Protocol) (mg/kg) (μg eq./mL) eq. hr/mL) (mg/kg) (mL/hr/kg) (mL/kg) (hr) E06 (anti-HSA)-2V 0.3 4.53 47.5 158 6.32 203 47.8 tandem (09_2691) H08 (anti-HSA)-2V 0.3 4.81 47.9 160 6.26 179 39.7 tandem (09_2992) 2V -PEG (09_2313) 1 17.8 533 533 1.89 95.8 53.5 2V (08_3952) 1 8.44 2.33 2.33 430 869 0.146 (~9 min)

(39) The resultant measured half-life is shown in FIG. 5(b) clearly showing that E06 and H08 extended the plasma half-life of 2V from approximately 9 minutes to 47.8 h and 39.7 h respectively.

(40) As a more precise means of measuring protein clearance and ensuring that fusion partners are intact, an LC-MS method was carried out measuring peptides specific to E06 and 2V. For the mouse model, 2V-E06 and 2V-E06-2V were injected at a dose of 4 and 2 mg/kg, respectively, both i.v. and s.c. into groups of 12 CD1 mice. Two blood samples plus terminal bleeds were taken from each animal at intervals to provide duplicate samples to cover time points from 1 h-168 h (FIG. 5(b). For the rat model, E06-2V and 2V-E06 were injected i.v. at 1 mg/kg into groups of 3 wistar rats and blood samples taken from 0.25-168 h. A PK study in cynomolgus monkeys was carried out with E06-2V and 2V-E06 dosed at 1 mg/kg i.v. to groups of two animals. Blood sampling was carried out at from 0.25 h-28 days. After 28 days, the 2V-E06 group were injected s.c. with 0.5 mg/kg protein and sampling carried out over 14 days.

(41) Table 6 summarizes the pharmacokinetic data measured from these experiments showing extended half-life of the E06 containing fusions, good correlation between intra-venous and subcutaneous administration and rapid distribution where pharmacokinetic parameters were measured across all three PK models.

(42) TABLE-US-00006 TABLE 6 Dose Cl Vdss t½ F Species Construct (mg/kg) Route (ml/hr/kg) (ml/kg) (h) (%) Rat E06-2V 1 iv 3.3 105  22 — 2V-E06 1 iv 2.7 94 25 — Mouse 2V-E06 4 iv 2   96 33 — 2V-E06 4 sc — — 25 48 2V-E06- 2 iv 3.8 112  21 — 2V 2V-E06- 2 sc — — 26 46 2V NHP 2V-E06 1 iv  0.25 75 210 — E06-2V 1 iv  0.29 69 164 — E06-2V 0.5 sc — — * >75  * = Not sufficient data to determine accurate half-life for E06-2V delivered via the subcutaneous route, although data shown graphically to approximate to intravenous dosing half-life.

(43) The half-life of 2V-E06 fusions across all three species in comparison to that of albumin was plotted. Overall these data demonstrate that the half-life values for 2V-E06 lie within 2 fold of those of albumin in all 3 species investigated. However, as all the half-life values obtained for 2V-E06 are similar to those of albumin it is believed that the addition of 2V-E06 has improved the pharmacokinetic of 2V because high affinity binding of the molecule to albumin has led to the 2V-E06 albumin complex taking on the pharmacokinetic and clearance properties of albumin.

(44) Assuming that the 2V-E06 albumin complex takes on the pharmacokinetic properties of albumin, the prediction of the likely pharmacokinetic properties of this molecule in human becomes simply a case of understanding the half-life of albumin in human. Literature data is available for the half-life of albumin in human (19 days). It is therefore anticipated that the half-life of 2V-E06 in human will approximate to 19 days and that it volume of distribution will approximate to 0.1 I/kg (assume volume of distribution is conserved across species).

(45) For each animal model, VNAR concentrations in plasma were analyzed by quantitative LC-MS as described. Briefly, plasma samples were treated as follows: 50 μl plasma was added to 50 μl 6 M guanidine containing the peptide internal standard and reduced with 20 μl of 32 mM Tris (2-carboxy-ethyl) phosphine-hydrochloride (TCEP) at 56° C. for 45 minutes. Samples were alkylated by addition of 10 μl of 128 mM iodoacetamide at 37° C. for 60 minutes. Samples were diluted by the addition of 150 μl 100 mM phosphate, pH8, 0.1% CHAPS. Using a Kingfisher magnetic bead processor, magnetic Ni-beads (25 μl/sample) were washed in 100 mM phosphate, pH8, 0.1% CHAPS before being transferred to plasma sample plate and incubated for 1 h. Three washes were carried out: 1st and 2nd wash: transfer beads to plate containing 100 μl phosphate, pH8, 0.1% CHAPS; 3rd wash: transfer beads to plate containing 100 μl phosphate, pH8, 0.1% CHAPS+20 mM imidazole. Bound VNAR was then eluted by transferring the beads to a plate containing 100 μl phosphate, pH8, +250 mM imidazole. Beads were removed and 100 μl 100 mM TRIS, pH8 containing 20 μg/ml Trypsin was added and incubated for 4 h at 37° C. Following this, 20 μl 100 mM TRIS, pH8 containing 100 μg/ml Trypsin was added and incubated overnight at 37° C. Samples were then loaded into a CTC PAL auto-sampler and analysed using LC-MS. Signature peptides were separated on an Agilent 1100 HPLC system using an Onyx monolithic RP C18 guard trapping cartridge and a Waters XBridge BEH130 C18 Column, 3.5 μM, 2.1×100 mm analytical column. Peptides were eluted with a gradient of 5% to 45% acetonitrile in water with 0.1% formic acid. Signature peptides within each partner were analysed independently: E06 signature peptide—EQISISGR and 2V signature peptide—AQSLAISTR. The analytes were detected by atmospheric pressure electrospray ionisation MS/MS using an AB Sciex AP15500 QTRAP triple quadrupole mass spectrometer. The ion chromatograms were quantified by reference to standards spiked into fresh control plasma and analysed over the range 0.04 to 50 μg/ml. The ion chromatograms were integrated and quantified by interpolation of the standard curve with a 1/y weighting using AB Sciex Analyst 1.5.1 software.

EXAMPLE 6

(46) Therapeutically Relevant Domains Retain Function when Fused to E06

(47) As 2V had no inherent target antigen or function, it was a good partner for in vivo PK work however to determine the ability of E06 not to inhibit nor impair the function of a fused partner protein, anti-ICOSL VNAR domains (A1 and CC3) in fusion with E06 were studied. Both A1 and CC3 have exhibited efficacy in vitro and in vivo and were good candidates to validate E06's utility as a domain capable of extending the half-life of therapeutically relevant domains.

(48) N, C and N/C terminal fusion of anti-murine ICOSL VNAR (A1 and CC3) and E06 were constructed, expressed, purified and assessed for dual binding against both HSA and mICOSL by BIAcore analyses (Table 7). All constructs were immobilized with HSA on the chip, with flowing over mICOSL. Both dimers and trimers of E06 with A1 and CC3 retained the ability to bind both HSA and ICOSL (A1-Fc average affinity: KD=6.27×10.sup.−7 M; CC3-Fc average affinity: KD=4.96×10.sup.−8 M. 2V did not bind ICOSL).

(49) TABLE-US-00007 TABLE 7 Domain KD (M) A1-E06 2.6 × 10.sup.−7 ± 2.5 × 10.sup.−8 E06-A1  2.1 × 10.sup.−7 ± 3.43 × 10.sup.−8 A1-E06-A1 4.7 × 10.sup.−7 ± 9.0 × 10.sup.−8 CC3-E06 7.7 × 10.sup.−8 ± 1.6 × 10.sup.−8 E06-CC3 1.0 × 10.sup.−7 ± 1.7 10.sup.−8   CC3-E06-CC3 1.0 × 10.sup.−7 ± 1.3 × 10.sup.−8 2V-E06-2V —

(50) To assess the ability of A1 and CC3 to retain the ability to block ligand binding to receptor in cell based neutralization assays, E06 fusions were incorporated (FIG. 6(b)) and the IC.sub.50 values measured. The assay was carried out follows: CHO cells expressing murine ICOSL receptor were grown to confluency in DMEM/F12+5% FBS media in 96-well cell culture plates (Greiner, Bio-One). mICOSL-hFc (20 μl at 450 ng/ml) was pre-incubated for 1 h with 40 μl of anti-mICOSL-NAR fused to EO6 in DMEM/F12+2% FBS and then added to the cells. Following 1 h incubation at 16° C. cells were gently washed 3 times with DMEM/F12+2% FBS and incubated for another 40 min at 16° C. with goat anti-human Fc-HRP (SIGMA) diluted 1:10 000 in the same media. Afterwards the cells were washed again 3 times with DMEM/F12+2% FBS media and ones with PBS and developed with TMB substrate. The results show that single digit nM efficacy was achieved showing the retention of neutralization by both A1 and CC3 when in complex with E06. Results are shown in Table 8 of BIAcore analyses of fusion domains immobilized with HSA and binding mICOSL.

(51) TABLE-US-00008 TABLE 8 Concentration IC.sub.50 IC.sub.50 Name mg/ml uM M.sub.r ng/ml nm A1-E06 0.056 2.16 25.9 182.6 7.05 E06-A1 0.078 3.01 25.9 125.3 4.84 A1-E06-A1 0.077 1.95 39.5 53.81 1.36 CC3-E06 0.039 1.51 25.9 61.18 2.36 E06-CC3 0.051 1.97 25.9 244 9.42 CC3-E06-CC3 0.043 1.09 39.5 24.45 0.62

(52) As a secondary measurement of functionality, T-cell proliferation assays were carried out as follows: antibodies were titrated in 96 well TC flat bottom plate in 100 ul assay media (use the media listed above, but leave out the Rat T stim, IL-2 and IL-lalpha). Tosyl activated magnetic Dynal beads are coated per product insert instructions with hu or mu ICOSL, anti-mu CD3e and hIgG1 filler (1 ug ICOSL/0.5 ug anti-CD3/3.5 ug hIgG1 per 1×10{circumflex over ( )}7 beads). Prior to assay set up, titer beads to determine optimal concentration that gives around 8000-40,000 CPM. Generally for hu ICOSL this will be around 40,000 beads/well and for mu ICOSL, around 20,000 beads/well. Add 50 μl/well of the appropriate beads to the titered antibodies. D10.G4.1 cells are washed 4× with assay media and resuspended to 8×10{circumflex over ( )}5 cells/ml and add 50 μl/well=40,000 cells/well. All wells are brought up to a final volume of 200 μl and incubated for 48 hours. 1 μci/well 3H-thymidine is added and incubated for 5-7 hours. Harvest and count CPM. Both A1 and CC3 still retained the ability to inhibit T cell proliferation when in fusion with E06 as shown in Table 9 and Table 10 with pM IC.sub.50 values.

(53) TABLE-US-00009 TABLE 9 Sample IC.sub.50 nM A1-E06-A1 24.02 CC3-E06-CC3 5.77 2V-E06-2V >128.21 Anti-ICOSL control 0.04

(54) TABLE-US-00010 TABLE 10 Sample IC.sub.50 nM A1-hFc 2.28 CC3-hFC 0.32 Anti-ICOSL control 0.03 Rat IgG2A control —