Human antibodies and binding fragments thereof to tenascin
11332520 · 2022-05-17
Assignee
Inventors
Cpc classification
A61P1/04
HUMAN NECESSITIES
A61P29/00
HUMAN NECESSITIES
A61P1/02
HUMAN NECESSITIES
A61P17/02
HUMAN NECESSITIES
A61P19/08
HUMAN NECESSITIES
A61P9/10
HUMAN NECESSITIES
C07K2317/33
CHEMISTRY; METALLURGY
A61K39/3955
HUMAN NECESSITIES
A61P1/18
HUMAN NECESSITIES
A61P9/14
HUMAN NECESSITIES
C07K2317/92
CHEMISTRY; METALLURGY
A61P1/16
HUMAN NECESSITIES
A61K2039/545
HUMAN NECESSITIES
A61P35/00
HUMAN NECESSITIES
A61K39/00
HUMAN NECESSITIES
Y02A50/30
GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
C12N15/63
CHEMISTRY; METALLURGY
A61P7/04
HUMAN NECESSITIES
C07K2317/76
CHEMISTRY; METALLURGY
A61P25/28
HUMAN NECESSITIES
A61P1/00
HUMAN NECESSITIES
A61P37/06
HUMAN NECESSITIES
A61P15/00
HUMAN NECESSITIES
International classification
C07K16/00
CHEMISTRY; METALLURGY
C12N5/00
CHEMISTRY; METALLURGY
C12N15/00
CHEMISTRY; METALLURGY
A61K39/00
HUMAN NECESSITIES
Abstract
The present disclosure relates to antibodies and binding fragments to a Tenascin, in particular the FBG domain of a Tenascin, which are potentially less immunogenic than the parent antibody. The disclosure also relates to compositions comprising the antibody or binding fragment and use of any one of the same for diagnosis, prognosis and/or treatment of disorders such as those associated with chronic inflammation. The disclosure further provides methods of making the antibodies.
Claims
1. An isolated human antibody or antigen-binding fragment thereof specific to a fibrinogen-like globe (FBG) domain of Tenascin C, wherein the antibody or antigen-binding fragment thereof comprises a VL sequence as shown in SEQ ID NO: 23, and a VH sequence with a CDRH1 of SEQ ID NO: 3, a CDRH2 of SEQ ID NO: 4 and a CDRH3 selected from SEQ ID NO: 5, 12, 14, 16, 18, 20, 24, 26, 28, 30 or 32.
2. The antibody or antigen-binding fragment thereof specific to a FBG domain of Tenascin C of claim 1, wherein the antibody or antigen-binding fragment thereof comprises a VL sequence as shown in SEQ ID NO: 22 and a VH sequence with a CDRH1 of SEQ ID NO: 3, a CDRH2 of SEQ ID NO: 4 and a CDRH3 of SEQ ID NO: 18.
3. The antibody or antigen-binding fragment of claim 1, comprising a VH sequence selected from SEQ ID NO: 6, 13, 15, 17, 19, 21, 25, 27, 29, 31, 33, 35 or a variant thereof wherein up to 5 amino acids in the sequence are changed.
4. The antibody or antigen-binding fragment of claim 3, comprising a VH sequence of SEQ ID NO: 19.
5. The antibody or antigen-binding fragment of claim 1, wherein the antibody or antigen-binding fragment is a Fab or Fab′ fragment.
6. The antibody or antigen-binding fragment of claim 1, wherein the antibody or antigen-binding fragment is a full length antibody.
7. The antibody or antigen-binding fragment of claim 6, wherein the light chain has a sequence as shown in SEQ ID NO: 1.
8. The antibody or antigen-binding fragment of claim 6, wherein the heavy chain has a sequence as shown in SEQ ID NO: 2.
9. The antibody or antigen-binding fragment of claim 1, wherein the heavy chain has a sequence as shown in SEQ ID NO: 34.
10. A pharmaceutical composition comprising an antibody or antigen-binding fragment according to claim 1 and a pharmaceutically acceptable excipient, diluent or carrier.
11. A polynucleotide encoding the antibody or antigen-binding fragment of claim 1.
12. A vector comprising the polynucleotide of claim 11.
13. A host cell comprising the polynucleotide of claim 11.
14. A host cell comprising the vector of claim 12.
Description
BRIEF DESCRIPTION OF FIGURES
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
EXAMPLES
Example 1—Generation of Purified Tenascin-c FBG as Antigen and Assay Reagents
(10) Purified soluble proteins containing the FBG domain of tenascin-C (TNC FBG) were generated for use as antigens in antibody selections and as reagents in subsequent screening and characterisation assays. To enable selection strategies for isolation of antibodies that bind tenascin-C of multiple mammalian species, a range of DNA expression constructs were synthesised, which incorporated the TNC FBG domain of either human, mouse, rat or dog. A human tenascin-R FBG construct was also prepared for identification of antibodies that displayed unwanted binding to this homologue. Constructs were produced as 6His-tagged proteins with either a rat CD4 or human IgG1 Fc tag coupled to either a C- or N-terminal FBG domain as described below.
(11) Protein Expression Constructs
(12) All synthetic DNA constructs for antigen expression were synthesised and sequence confirmed by Genscript (Piscataway, USA). FBG domains were cloned into the mammalian expression vectors pBIOCAM4 or BIOCAM5, which fuse the expressed domains with either a rat Cd4 (domains 3 and 4) tag (Chapple et al, 2006) or a human IgG1 Fc tag (Falk et al, 2012) respectively. The vectors were modified from the pCMV/myc/ER plasmid (Invitrogen) (Falk et al, 2012), which contains an endoplasmic reticulum (ER) signal sequence derived from the mouse VH chain, for secretion of expressed proteins. For all constructs which resulted in an N-terminal FBG (e.g. FBG-Fc-His or FBG-rCd4-His) the digested PCR products were ligated with NcoI/NotI cut pBIOCAM4 or pBIO CAMS vectors. For all constructs which resulted in a C-terminal FBG (e.g. Fc-His-FBG or rCd4-His-FBG), digested PCR products were ligated with BamHI/HindIII cut pBIOCAM4 or pBIOCAM5 vectors. The primers used to amplify the FBG domains are listed in
(13) Protein Expression and Cell Culture
(14) Transfection quality plasmid DNA was prepared using the Machery Nagel Nucleobond Xtra Midi kit (740410.50, Fisher Scientific, UK). HEK293F suspension cells and Freestyle media, for antigen and antibody expression, and RPMI media were from Life Technologies (Paisley, UK). Transfection of HEK293F cells was carried out as described previously (Chapple et al, 2006).
(15) Protein Purification and QC
(16) Protein affinity purification employed either Ni-NTA agarose or immobilised recombinant protein A resin.
(17) For purification of His-tagged proteins, culture supernatants were mixed with Ni-NTA agarose (1018240, Qiagen, Crawley, UK) for 1 h and the resin transferred to Proteus 1-step midi spin columns (Generon, UK) for centrifugation (200×g, 2 min). Unbound proteins were washed out with phosphate buffered saline (PBS) supplemented with 20 mM imidazole (pH 8). Bound proteins were eluted in fractions through addition of 300 mM imidazole in PBS (pH 8) and column centrifugation (200×g, 2 min). Pooled fractions containing eluted protein were placed in Gebaflex Midi dialysis tubes (Generon D010; molecular weight cut-off 3.5 kDa) and dialysed against PBS.
(18) Fc-tagged proteins and antibodies expressed as human IgG4 were purified using protein A sepharose (PC-A25, Generon, Maidenhead, UK). Culture supernatants were clarified by centrifugation (2500×g, 15 min) and mixed with protein A sepharose overnight at 4° C. before transfer of the resin to Proteus 1-step midi spin columns (Generon, UK). Columns were centrifuged (200×g, 2 min) and washed with PBS to remove unbound protein. Fc-tagged or IgG4 proteins were eluted in fractions from the protein A with 0.2 M glycine (pH 2.8) into Tris-HCl (pH 8) by centrifugation (200×g, 2 min). Eluted fractions were pooled and dialysed against PBS in Gebaflex Maxi dialysis tubes (Generon D045; molecular weight cut-off 8 kDa).
(19) Proteins were analysed for purity and concentration by SDS-PAGE (4-12% gel) and spectrophotometry (OD280 using theoretical extinction coefficient). Where purified proteins were used in cell-based assays the endotoxin content was first determined by limulus amoebocyte lysate chromogenic endotoxin assay (Pierce). Proteins were not used if endotoxin levels exceeded 1 endotoxin unit per milligram (i.e. 1 EU/mg).
Example 2—Isolation of Primary Anti-FBG Antibodies
(20) Antibody Phage Display
(21) Antibodies against tenascin-C FBG domain were isolated using the Iontas Ltd proprietary human antibody phage display library, which was constructed using DNA isolated from 43 human lymphocyte donors. Selections, phage rescues and subcloning into pSANG10 (Martin et al, 2006) were all performed as described previously (Schofield et al, 2007) using techniques that are well known in the art.
(22) Two rounds of panning selections were performed on immobilised TNC FBG fused to human IgG1 Fc or rCd4 at either the N terminus of the fusion partner (e.g. FBG-Fc, FBG-rCd4) or at the C terminus (Fc-FBG, rCd4-FBG). Phage antibody libraries containing either kappa (κ) or lambda (λ) variable light chains (V.sub.L) were panned separately to facilitate later sub-cloning to Fab expression vectors containing either constant light (C.sub.L) kappa (κ) or lambda (λ) chains.
(23) Polyclonal phage populations were prepared from the selected populations and were tested in ELISA (polyclonal phage ELISA) using ELISA plates coated with TNC FBG antigen or appropriate fusion partner (Fc or rCd4). After incubation with phage, plates were washed, and bound phage detected using peroxidase-conjugated anti-M13 antibodies. Enrichment of antigen-specific binders between rounds 1 and 2 of selection and a greater proportion of FBG binders compared to anti-Fc or -rCd4 phage in the round 2 output populations, indicating that the selections were successful.
(24) Confirmation of scFv Binding to Antigen and Cross-Reactivity Assay by ELISA
(25) Round 2 selection outputs were expressed as individual scFv clones to confirm antigen recognition in ELISA binding assays. Output populations were sub-cloned into the bacterial expression vector pSANG10 (Martin et al, 2006), transformed into E. coli BL21 (DE3), and individual transformants were induced in 96-well plates as described previously (Schofield et al, 2007). E. coli supernatants were collected and assayed for binding of scFv to TNC FBG using DELFIA-based ELISA, using europium-labelled anti-FLAG detection antibodies.
(26) The most successful selections with the A library were based on panning against the antigens rCd4-FBG and Fc-FBG (selections 147 and 148). For the κ library, the most successful selections were obtained with the antigens FBG-rCd4 (150), rCd4-FBG (152) and Fc-FBG (153). The 79 positive clones from this ELISA screen were selected for further analysis.
(27) Cross-reactivity ELISA showed that 67/79 (85%) of anti-human FBG scFv were cross-reactive to mouse TNC FBG. DNA sequence analysis of the anti-FBG scFv indicated excellent sequence diversity. For example, selections 147 and 148 from the V.sub.L λ library contained 92% unique variable heavy (V.sub.H) complementarity determining region 3 (CDR3) sequences, and selections 150, 152 and 153 from the V.sub.L κ library contained 67%, 91% and 100% unique variable V.sub.H CDR3 sequences, respectively.
(28) A further 1425 clones isolated from the most effective selections were screened by ELISA and this resulted in the identification of an additional 401 scFv with FBG-binding specificity. These clones, together with the 79 scFv identified in initial ELISAs were chosen for further evaluation. The 1425 clones were further tested in a specificity ELISA in which each scFv was tested for binding to human Tenascin R FBG and also to human, mouse, rat and dog TNC FBG. Clones were ranked according to the ELISA signal obtained for binding to Tenascin C divided by the signal for Tenascin R FBG binding. The top 250 clones with a ratio above 50 were taken for subcloning and further analysis.
Example 3—Screening of Primary Anti-FBG Antibodies in a Functional Assay
(29) Anti-FBG scFv were reformatted either as bivalent scFv-Fc or as monomeric Fabs for evaluation of their activity as inhibitors of FBG-evoked signalling in a whole cell assay system.
(30) The top 50 anti-FBG scFv, ranked by primary ELISA signal, for each of the selections 147, 148, 150, 152 and 153 were sub-cloned into the mammalian expression plasmid pBIOCAM5 (Falk et al, 2012) as individual selection populations and expressed by transient transfection in HEK293F cells (Chapple et al, 2006). For Fab expression, pooled λ or κ scFv variable heavy (V.sub.H) and variable light (V.sub.L) inserts were cloned into a dual promoter Fab expression vector (pFab-dual-κ or pFab-dual-λ, depending on the light chain germ-line) using a proprietary Iontas Ltd protocol. Culture supernatants were screened for activity in the THP-1 cell assay and selected scFV-Fc and Fab hits were affinity purified for re-assaying and confirmation of inhibitory activity.
(31) THP1-Blue™ Reporter Cell Assay
(32) Tenascin-C has been shown to elicit the generation of cytokines in inflammatory cells and fibroblasts by interaction of the FBG domain with cellular TLR4 (Midwood et al, 2009). The receptor signalling cascade leading to generation of inflammatory cytokines such as TNFa, IL-8 and IL-6 involves activation of the transcription factor NF-κB. This process can be studied in ‘reporter’ cell lines modified to respond to NF-κB activation with generation of an easily measured protein signal. The THP1-Blue™ reporter cell line (InvivoGen; Toulouse, France) is derived from the human THP-1 monocyte cell line and stably expresses an NF-κB-inducible secreted alkaline phosphatase (SEAP) reporter construct. These cells also constitutively express cell surface TLR4, which enables the signalling activity of TNC FBG fusion proteins to be readily measured using colorimetric or fluorimetric quantitation of SEAP in culture supernatants using medium- to high-throughput assay methods.
(33) Activity at low FBG concentrations is critical to the success of any screening assay; if the concentrations of FBG required to produce a robust increase in the reporter protein are too high then the expression levels and concentrations of scFv, Fc-ScFv or Fab constructs required to fully inhibit any such signal would be unacceptable for a screen. Fc-FBG produces a robust SEAP signal at low nM levels in this cell assay (CD4-FBG did not produce a response in this concentration range).
(34) THP1-Blue™ cells were cultured and passaged in supplemented RPMI media according to supplier's protocols (www.invivogen.com/PDF/THP1_Blue_NF_kB_TDS.pdf), except that cells were grown in ultra-low attachment T75 flasks. For assays, THP1-Blue™ cells were added to 96-well tissue culture plates (100,000 cells/well) containing Fc-FBG (3 or 10 nM) in RPMI medium in a total volume of 170 μl. Culture supernatants containing expressed scFv-Fc or Fab, or affinity purified antibody in PBS, was added in a volume of 30 μl and cells were incubated for 18 h at 37° C. Supernatants were harvested and assayed for either SEAP using the Attophos AP fluorimetric quantitation system (S1000; Promega) or IL-8 content using the DuoSet ELISA development system (DY208; R&D Systems, UK) according to the supplier's instructions. Data were plotted and curves fitted using Prism software (GraphPad).
(35) Screening of anti-FBG antibodies as HEK293F culture supernatants highlighted putative inhibitors of Fc-His-FBG evoked signalling in THP1-Blue™ cells of which 9 were confirmed when re-assayed as purified scFv-Fc or Fab. Fc-His-FBG is key to having the potency assays work. Monomeric FBG does not elicit any cytokine response in THP-1Blue and human cells.
Example 4—Functional Characterisation of Primary Anti-FBG Antibodies
(36) ELISA Cross-Reactivity Assays
(37) The panel of 9 human FBG signalling inhibitors identified in the THP1-Blue™ functional assay was evaluated by ELISA for cross-reactivity to rat, mouse, and dog FBG. Binding to the human tenascin-R FBG homologue was also determined. Assay wells were coated with human, rat, mouse, and dog TNC FBG-rCD4, or human TNR FBG-rCd4 fusion proteins and binding of Fabs was detected using anti-kappa or anti-lambda mAb followed by Europium-conjugated anti-mouse mAb. ELISA results revealed that the C3 antibody showed good cross-reactivity to other mammalian homologues of human TNC FBG, with lower apparent binding to human TNR FBG. These were:
(38) Determination of Binding Affinity by Surface Plasmon Resonance
(39) The affinity and association and dissociation kinetics of selected Fabs for binding to the human, rat and mouse TNC FBG, and human TNR FBG were measured by surface plasmon resonance (SPR) at 25° C. Experiments were performed using a BIAcore T100 instrument with CM5 sensor chip according to the protocol provided with the Human Fab Capture Kit (GE, 28-9583-25). Varying concentrations of rCd4-FBG were injected into a flow-cell with immobilised Fab and a reference flow-cell. After reference signal subtraction, the data was fitted to a global 1:1 fit using the T100 BIAevaluation software.
(40) The calculated kinetic constants are shown in Table 3. The rank order of affinity of Fabs for human TNC FBG was B12 (110 pM)>. All Fabs displayed low nanomolar affinity for rodent TNC FBG, and affinities for human TNR FBG were typically greater than 60-fold lower than human TNR FBG. Inhibitory Potency Assays
(41) The potency of purified Fabs for neutralisation of huFc-His-FBG activity was determined in the THP1-Blue™ assay, using measures of TLR4-mediated secreted alkaline phosphatase and IL-8 cytokine production. Assays were conducted as described in Example 2, except that purified Fabs were added to assay wells at a range of concentrations (0.3-100 nM) to enable calculation of IC.sub.50 values using Prism software (GraphPad).
(42) The C3 antibody of the present disclosure is derived from an antibody referred to as B12.
(43) TABLE-US-00003 TABLE 1 Anti-FBG Fab binding kinetic data determined by surface plasmon resonance (SPR) spectroscopy. Kinetics K.sub.D K.sub.a K.sub.d Steady Fab FBG (nM) (M.sup.−1s.sup.−1) ×10.sup.5 (s.sup.−1) ×10.sup.−4 State B12 Hu TNC 0.111 26.62 3.0 N/A Mu TNC 13 52.15 675.5 18.7 Rat TNC 7.9 94.59 747.9 N/A Hu TNR 33.9 13.96 472.5 36.1 K.sub.D, equilibrium dissociation constant; K.sub.a, association constant; K.sub.d, dissociation constant
Example 5—Generation and Isolation of Optimised Antibodies to huTNC FBG Domain
(44) Affinity Maturation by Targeted CDR Mutagenesis
(45) Anti-FBG antibody B12 was selected for affinity maturation. Targeted CDR mutagenesis was carried out by randomising VH and VL CDR3 residues in blocks of 6 amino acids using Kunkel mutagenesis (Fellouse and Sidhu, 2007; Kunkel et al., 1987; Sidhu and Weiss, 2004). Due to the longer VH CDR3s (10-16 residues) for the given clones randomisation was done in three overlapping blocks and the VL CDR3s (9 residues) were randomised in two overlapping blocks. Randomisations were carried out using NNS (N=A/G/C/T and S=G/C) degenerate primers that could encode any of the 20 amino acids (and only a single amber stop codon) at a given position from 32 codon combinations. The following library was created.
(46) TABLE-US-00004 TABLE 2 Estimated sizes of the CDR3 randomised libraries Library Sub library Size Combined size B12 VH B12 VH 3.1 1.8 × 10.sup.9 6.1 × 10.sup.9 B12 VH 3.2 1.6 × 10.sup.9 B12 VH 3.3 1.7 × 10.sup.9 B12 VL B12 VL 3.1 2.6 × 10.sup.9 7.7 × 10.sup.9 B12 VL 3.2 5.1 × 10.sup.9
(47) High Stringency Phage Display Selections
(48) Phage-antibody selections on streptavidin Dynabeads were performed as described previously (Dyson et al, 2011). Multiple rounds of solution-phase selections were carried out on biotinylated rCd4-His-FBG to enrich for affinity improved clones. The optimum antigen concentrations for each round were determined empirically by selecting against a range of antigen concentrations and comparing the output numbers with a no-antigen control. The stringency of selection was increased by reducing the amount of antigen used in each round. No further rounds of selection were carried out after the selection window (the fold difference between phage titres from selection outputs and no antigen control) dropped below 10. Hence, three rounds of selection were carried out on biotinylated human rCd4-His-FBG for all libraries except B12 which was subjected to a fourth round of selection due to the large selection windows observed at round 3. All libraries were subjected to deselection against streptavidin beads and tenascin-R (100 nM for rounds 1 to 3 and 1 nM for round 4) at each round of selection to avoid unwanted cross reactivity to streptavidin or tenascin-R. In addition, a hybrid selection strategy in which the human and mouse antigens were alternated between rounds of selection was performed for the B12 randomised libraries only. The reason for performing this extra selection on the B12 libraries was the large difference in affinity observed for the B12 parental antibody binding to human and mouse rCd4-his-FBG. Furthermore, an additional round of selection was carried out to select for antibody clones with superior off-rates. In off-rate selections, phage were allowed to bind to the biotinylated antigen (1 nM in this case), and a large excess of non-biotinylated antigen (500 nM) was subsequently added to the reaction and incubated for 20 h or 40 h. The non-biotinylated antigen serves as a competitor and captures the phage antibodies that dissociate from the biotinylated antigen, i.e. only the antibodies with longer off-rates will be recovered at the end of the selection (Hawkins et al., 1992; Zahnd et al., 2010). The output phage titres for each round of selection together with calculated selection windows are shown in Tables 3 to 5 below.
(49) TABLE-US-00005 TABLE 3 Selection output titres. Round 1 selections. Phage output titres were determined as described previously (Schofield et al, 2007). Selection Selection window window CDR3 for for randomised 10 nM 1 nM 0 nM 10 nM 1 nM libraries Selection Selection Selection selection selection B12 VH 6 × 10.sup.7 2.6 × 10.sup.7 1 × 10.sup.5 600 260 B12 VL 6 × 10.sup.7 .sup. 5 × 10.sup.7 2 × 10.sup.5 300 250
(50) TABLE-US-00006 TABLE 4 Selection output titres. Round 2 selections. Phage output titres were determined as described previously (Schofield et al, 2007). Selection Selection window for window for CDR3 randomised 200 pM 50 pM 0 nM 200 pM 50 pM libraries Selection Selection Selection selection selection B12 VH .sup. 1 × 10.sup.8 6.75 × 10.sup.7 2 × 10.sup.4 5000 3375 B12 VL 1.2 × 10.sup.8 8.1 × 10.sup.7 4 × 10.sup.4 3000 2025 B12 VH on mu TNC FBG .sup. 7 × 10.sup.6 2 × 10.sup.4 350 B12 VL on mu TNC FBG 7.5 × 10.sup.6 4 × 10.sup.4 187
(51) TABLE-US-00007 TABLE 5 Selection output titres. Round 3 selections. Phage output titres were determined as described previously (Schofield et al, 2007). Selection Selection CDR3 window window for randomised 5 pM 1 pM 0 nM for 5 pM 1 pM libraries Selection Selection Selection selection selection B12 VH 1.5 × 10.sup.7 .sup. 4 × 10.sup.6 <1 × 10.sup.5 >150 >40 B12 VL 2.7 × 10.sup.7 3.5 × 10.sup.6 <1 × 10.sup.5 >270 >35 Hybrid Selection Selection selections on window window for B12 libraries 20 pM 5 pM 0 pM for 20 pM 5 pM (Hu-mu-hu) Selection Selection Selection selection selection B12 VH .sup. 1 × 10.sup.8 7.7 × 10.sup.6 <1 × 10.sup.5 >1000 >77 B12 VL 1.3 × 10.sup.8 1.8 × 10.sup.7 <1 × 10.sup.5 >1300 >78
(52) ELISA Screen
(53) An anti-FLAG capture ELISA was performed to screen for clones that had an improved affinity for mouse FBG binding compared with the parental antibodies.
(54) E. coli clones harbouring scFv pSANG10 expression plasmids were induced in 96-well plates with auto-induction media as described previously (Schofield et al, 2007). E. coli supernatants were harvested for ELISA assays. ELISA used the DELFIA (dissociation enhanced lanthanide fluorescent immunoassay) system with Europium-labelled anti-FLAG antibody (Sigma, Aldrich, UK). Black immunosorb plates (Nunc) were coated overnight with anti-FLAG M2 antibody (Sigma, F3165, 5 μg/ml in PBS, 50 μl per well), in wells blocked by the addition of 2% milk powder, PBS (PBS-M, 300 μl per well). Plates were washed three times with PBS-T (PBS, 0.1% Tween-20) and three times with PBS followed by the addition of a 1:2 dilution of 96-well auto-induction culture supernatants containing expressed scFv in PBS-M (50 μl per well). The plates were incubated for 1 h, washed as above and biotinylated mouse or human rCd4-His-FBG (5 μg/ml in PBS-M, 50 μl) added to each well. Plates were incubated for a further 1 h, washed and Strepravidin-Eu added (Perkin Elmer, 1 μg/ml, PBS-M, 50 μl), incubated for 30 min, washed and DELFIA enhancement solution added (50 μl) and plates read on a Perkin Elmer Fusion plate reader (excitation=320 nm, emission 620 nm).
(55) In this assay differences in scFv expression level are normalised because the expression levels of scFv in auto-induction cultures saturate the anti-FLAG coated wells. Therefore, the signals obtained in the assay reflect the amount of biotinylated rCd4-His-FBG bound after washing, which will be a function of the off-rate of that clone for mouse or human FBG. ELISA screening of the selection output from the B12 sub-library revealed clones with improved binding to mouse TNC FBG.
(56) HTRF Screen
(57) An HTRF-based competition assay was developed to screen for antibody variants with improved binding to human TNC FBG.
(58) All samples and reagents were prepared in assay buffer (50 mM NaPO.sub.4, 0.1% BSA, 0.4 M KF, pH 7.0) at 4× the stated concentration. 5 μl of each reagent was subsequently added to low volume 384-well assay plates (Greiner, 784075) to give a final reaction volume of 20 μl. IgG antibodies were labelled using the d2 labelling kit (CisBio, 62D2DPEA) as directed by the manufacturer. Streptavidin europium cryptate (CisBio, 610SAKLA, Lot #25C) was used at a final concentration of 1.8 ng active moiety (SA) per 20 μl reaction as recommended by the manufacturer. Biotinylated rCd4-His-FBG was prepared using EZ-link Sulfo-NHS-LC-Biotin reagent (Thermo Scientific, 21327) the extent of biotinylation was quantified using biotinylation fluorescence quantitation kit (Thermo Scientific, 46610). Where appropriate, supernatants containing scFv (prepared as described above for ELISA assays) were added to the 384-well assay plate at a final dilution of 1/20 (i.e. 1/5 dilution in assay buffer followed by addition of 5 μl diluted sample to the 20 μl FRET assay). The concentrations of d2-labelled B12 IgG used for screening were 1.25 nM. Unless otherwise stated, biotinylated rCd4-His-FBG (biotin:protein ratio=1.8:1) was present at either 2.2 nM (in assays using the 2A5 IgG antibody) or 1 nM (in experiments using B12 IgG). Samples were incubated for approximately 1 h at room temperature and the FRET signal was determined using a BMG Pherastar instrument: excitation=320 nm; emission=620 nm and 665 nm; integration start time=60 μs; integration time=500 μs; 100 flashes per well. For competition assays containing culture supernatant, biotinylated rCd4-His-FBG antigen was pre-incubated with streptavidin europium cryptate for 45 min prior to addition of reagents to the assay plate. All FRET signals are presented as ΔR, where R=(E665/E620×104) and ΔR=(Rsample−Rbackground fluorescence).
(59) Culture supernatants containing unlabelled scFv clones from affinity selected mutant libraries were tested for inhibition of the interaction between FBG and the fluorophore-labelled parental IgG antibody. The relative ranking of clones exhibiting FRET signals within the useful range in both assays was broadly unchanged, indicating that they were competing for similar epitopes. Hence, all B12 scFv variants from affinity maturation selections were screened for their ability to inhibit the binding of B12 IgG molecules to human TNC FBG. The parental clones, expressed as scFvs in parallel with the affinity matured clones, were used as benchmarks.
(60) ScFv were sequenced and a panel of clones with unique VH or VL CDR3 sequences was selected for further study in human IgG4 format, based on their binding to mouse and human TNC FBG in the ELISA and HTRF assays, respectively.
(61) TABLE-US-00008 TABLE 6 HTRF screen for clones with improved affinity for human rCD4-FBG. % inhibition Total by parent CDR3 Selection clones % inhibition of FRET signal scFv Library type tested 0-25% 25-50% 51-75% 76-85% 86-90% 91-95% ≥96% 2A5 B12 B12 VH 100 fM 46 6 2 3 8 5 6 16 19 86 B12 VH Hybrid 46 3 3 5 5 3 9 18 19 86 5 pM
(62) Variants of antibody B12 showed ≥4-fold improvement for mouse FBG binding, and ≥91% inhibition of HTRF signal. In total, 31 clones fitting these criteria with unique CDR3 sequences were identified below.
(63) TABLE-US-00009 TABLE 7 Heavy or light chain CDR3 sequences of clones identified with improved binding to mouse and human TNC FBG and chosen for conversion to human IgG format for further study. Clone CDR Library name sequence B12 VH 165_13_B1 VMSSMEDAFDI SEQ ID NO: 12 165_13_B6 GQKGEGDTFDI SEQ ID NO: 14 165_13_D1 GTRGEGDTFDI SEQ ID NO: 16 165_13_C3 SYQSDEDAFDI SEQ ID NO: 18 165_13_D4 GTVGEGDTFDI SEQ ID NO: 24 165_13_A4 DKYPVLDTFDI SEQ ID NO: 26 165_13_B3 ALARGHDTFDI SEQ ID NO: 28 165_13_E1 DISAVMDVPQT SEQ ID NO: 30 180_11_F5 VMRTGLDTFDI SEQ ID NO: 32
(64) These are heavy or light chain sequences of antibody clones that bind to human and mouse TNC FBG and thus have potential utility in the methods, uses, compositions and compounds of the present invention. For example, antibodies that bind TNF FBG having these CDR3 sequences may be useful in identifying, inhibiting the function of, detecting and purifying TNC or TNC FBG.
(65) Conversion to IgG4 Format and Determination of Binding Kinetics
(66) The 31 scFv of interest were sub-cloned into a human IgG4 expression vector for generation of antibodies as human IgG4 with a hinge-stabilising mutation (S241P; Angal et al, 1993). IgG4 antibodies were transiently expressed in HEK-293F cells and culture supernatants were screened using surface plasmon resonance spectroscopy for ranking of their off-rates for binding to human and mouse TNC FBG, and human TNR FBG. Briefly, surface plasmon resonance (SPR) experiments were performed using a BIAcore T100 instrument and followed the protocol according to the Human antibody capture kit protocol (GE, BR-1008-39). For off-rate screening, 10,000 response units (RU) of anti-human Fc IgG (GE, BR-1008-39) was immobilised on flow-cells (FC1 and FC2) of a Series 5 CM5 dextran sensor chip (BR-1005-30) using EDC/NHS cross-linking chemistry according to the amine coupling kit protocol (GE, BR-1000-50). Culture supernatants containing expressed IgG4 were diluted 1:2 with 2×PBS-T and injected into FC2 (flowrate 5 μl/min, 60 s contact time) to enable antibody capture at 25° C. Antibody capture levels ranged from 308 to 1975 RU depending on the expression level of the antibody in the supernatant.
(67) A fixed concentration of antigen (15 nM of human and mouse TNC rCd4-His-FBG and 100 nM of human TNR rCd4-His-FBG) was injected with a flow-path via FC 1 (reference flow cell) and FC 2 (antibody capture flow cell), with a flow rate of 30 μl/min, and the association and dissociation phases measured over 1 and 5 min time periods, respectively. Regeneration of the binding surface employed 3M MgCl.sub.2 with 30 s contact time. Off rates were determined by reference cell subtraction and fitting the sensogram experimental data assuming a 1:1 interaction using BIAevaluation software (GE, BR-1005-97). Results of the off-rate screen are summarised in the Table 8 below.
(68) TABLE-US-00010 TABLE 8 Surface plasmon resonance screen for ranking of human IgG4 anti-FBG off-rates kd (s.sup.−1 × 10.sup.−4) for rCD4-His-FBG Human Mouse Human Clone name TNC FBG TNC FBG TNR FBG 165_13_C3 0.00095 0.033 120 162_02_C3 0.0149 20 6350 B12 parent 1.5 300 1001
(69) Clones were ranked according to low off-rate for human and mouse TNC rCd4-His-FBG, and high-off rate for human TNR rCd4-His-FBG. The 3 highest-ranking antibodies from each library were prioritised for more detailed kinetic analysis as purified IgG4. These clones are shown in Table 9 below.
(70) TABLE-US-00011 TABLE 9 Heavy chain CDR3 amino acid sequences of B12 mutants with improved FBG binding off-rate characteristics (bol-underlined shows the amino acids that were changed in B12 parent). Clone VH CDR3 B12 parent DISAVPDTFDI SEQ ID NO: 5 165_13_B1 VMSSMEDAFDI SEQ ID NO: 12 165_13_D1 GTRGEGDTFDI SEQ ID NO: 16 165_13_C3 SYQSDEDAFDI SEQ ID NO: 18
(71) Detailed kinetic parameters were evaluated for the 9 prioritised IgG4 antibodies. Binding characteristics were determined for interaction with human, rat and dog TNC rCD4-His-FBG, and human TNR rCD4-His-FBG. Kinetic assays followed essentially the same protocols as for the off-rate determinations described above, with some modifications as follows. To improve the accuracy of kinetic parameter determination, anti-human Fc IgG was immobilised at lower levels (2229 RU), resulting in a corresponding reduction in the amount of anti-FBG IgG4 captured.
(72) Purified anti-FBG IgG4 was diluted to a concentration of 3.5 nM in PBS, pH 7.4, 0.05% Tween-20 and injected into FC2 at a flow rate of 10 μl/min, 60 s contact time. This typically resulted in an average of 80 RU of antibody captured (range: 55 RU to 90 RU). Antigens were prepared by doubling dilution in PBS, pH 7.4, 0.05% Tween-20 (highest concentration 100 nM except mouse rCD4-His-FBG which was 7 nM). Assays were performed at 37° C. (30 μl/min, 120 s contact time; mouse rCD4-His-FBGFBG 10 μl/min, 60 s contact time), with both the flow cell and injection chamber equilibrated to this temperature. As before, kinetic parameters were determined by reference cell subtraction and fitting the sensogram experimental data assuming a 1:1 interaction using BIAevaluation software (GE, BR-1005-97).
(73) All nine antibodies displayed improved binding to mouse TNC FBG domain compared to the non-affinity matured parent clones, and antibodies 165_13_B1, 165_13_C3, and 160_01_A4 exhibited sub-nanomolar K.sub.D values for binding to human TNC FBG, with >70-fold lower affinity to the human TNR FBG analogue:
(74) TABLE-US-00012 TABLE 10 Anti-FBG IgG4 binding kinetic data determined by surface plasmon resonance at 37° C. Antibody rCD4-His-FBG K.sub.D Ka K.sub.d IgG4 Parent Species Tenascin (nM) (M.sup.−1s.sup.−1) × 10.sup.4 (s.sup.−1) × 10.sup.−4 B12 B12 Human TNC 0.24 47.1 11.2 Mouse TNC 4.5 30 13.8 165_13_B1 B12 Human TNC 0.26 72.7 18.8 Mouse TNC 0.96 73.3 7.06 Rat TNC 2.20 31.1 68.4 Dog TNC 2.85 65.5 187 Human TNR 94.4 12.2 1149 165_13_C3 B12 Human TNC 0.072 116 8.3 Mouse TNC 0.46 97.2 4.45 Rat TNC 1.22 38.9 47.3 Dog TNC 1.80 59.7 108 Human TNR 35.8 12.0 431
Example 7—Anti-FBG IgG4 Binding to Citrullinated FBG
(75) The binding affinity of antibody B12 to citrullinated FBG was determined by surface plasmon resonance (SPR). B12 was expressed as a human IgG4 with the hinge-stabilising S241P mutation using the QMCF expression technology (Icosagen, Estonia) and purified by protein A affinity chromatography (MabSelect Sure; GE Healthcare).
(76) Citrullination of human TNC FBG
(77) Purified human His-FBG was citrullinated using either peptidylarginine deiminase 2 (PAD2; MQ-16.201-2.5, Modiquest, NL) or peptidylarginine deiminase 4 (PAD4; MQ-16.203-2.5, Modiquest, NL) according to the supplier's instructions. Briefly, His-FBG was diluted to 1 mg/ml in the supplied deimination buffer (0.1 M Tris-HCl pH 7.5, 10 mM CaCl.sub.2, 5 mM dithiothreitol) and 250 μl mixed with 125 mU of either PAD2 or PAD4 enzyme followed by incubation at 37° C. for 2 h. Citrullination was confirmed by amino acid analysis of the enzymatically-treated samples. Aliquots of His-FBG in deimination buffer were incubated for 2 h at 37° C. in the absence of added PAD enzyme, for use as non-citrullinated control protein. Citrullinated and unmodified His-FBG proteins were used in SPR experiments as described below.
(78) Surface Plasmon Resonance
(79) SPR experiments were performed on a BIAcore 3000 instrument Anti-human IgG (GE Healthcare) was covalently coupled to the surface of a CM5 sensor chip using amino coupling chemistry. The amount of the coupled anti-human IgG expressed in RU units varied between 6500-7000 (6.5-7.0 ng/mm.sup.2). B12-hIgG4 (1-13 nM) was attached to the immobilised anti-human IgG in HBS-EP buffer (10 mM Hepes, 0.15 M NaCl, 2.5 mM EDTA and 0.005% Tween-20) at 25° C. Binding of the His-FBG variants to the immobilised B12-hIgG4 was also measured in HBS-EP buffer at 25° C. The flow rate was 5 μl/min in the immobilization experiments and 20 μl/min for kinetic analyses. The sensor chip surface was regenerated using 3 M MgCl.sub.2. Data were analysed using BIAevaluation program 4.1 (GE Healthcare).
(80) Analysis of B12-IgG4 binding to citrullinated His-FBG revealed that the kinetic parameters were essentially unchanged when compared to values obtained for binding to unmodified His-FBG. These results indicate that anti-FBG antibodies of the B12 lineage would be expected to bind both citrullinated and non-citrullinated forms of TNC FBG in therapeutic or diagnostic applications:
(81) TABLE-US-00013 TABLE 11 Kinetic parameters for interaction of B12-hIgG4 with the His-FBG variants. Each kinetic parameter represents the mean ± s.d. of 3 independent determinations. Analyte K.sub.D (M) K.sub.on (M.sup.−1s.sup.−1) K.sub.off (s.sup.−1) His-FBG (1.7 ± 0.3) × 10.sup.−10 (4.1 ± 0.6) × 10.sup.6 (6.8 ± 0.9) × 10.sup.−4 His-FBG + (3.2 ± 0.3) × 10.sup.−10 (3.0 ± 0.4) × 10.sup.6 (9.6 ± 0.8) × 10.sup.−4 PAD2 His-FBG + (3.2 ± 0.7) × 10.sup.−10 (2.6 ± 0.6) × 10.sup.6 (7.8 ± 0.4) × 10.sup.−4 PAD4
Example 8—Detection of TNC FBG in Human RA Tissue Using Immunohistochemistry
(82) Immunohistochemistry studies were performed to determine whether anti-FBG antibodies effectively recognise endogenous forms of the human TNC FBG protein in human tissue. Tenascin-C is expressed at sites of chronic inflammation and its localisation within the inflamed synovium of joints from individuals with rheumatoid arthritis has previously been demonstrated by immunohistochemistry using commercially available antibodies (Goh et al, 2010; Salter D M, 1993).
(83) The B12 antibody was expressed as mouse IgG2a format using the QMCF expression technology (Icosagen, Estonia) and purified by Protein G affinity chromatography followed by Superdex 200 gel filtration. Control mouse IgG1 anti-tenascin-C antibody (Clone 4F10TT; Takara Clontech), which recognises an EGF domain of full-length human tenascin-C was used as a positive control comparator. Mouse IgG1 (Dako X0931) or IgG2a (Dako X0943) against an irrelevant bacterial antigen were used as control primary antibodies to determine the level of non-specific background staining with these isotypes. Frozen sections of human knee joint synovium from donors with confirmed RA diagnosis (Asterand, UK) were equilibrated to room temperature, fixed (10 min) in 1:1 v/v acetone/methanol, and transferred to wash buffer. Immunostaining was performed using a Dako Autostainer with Envision Flex reagents (Dako K8010) according to manufacturer's protocols. Briefly, fixed tissue slides were placed onto the automated stainer and blocked (peroxidase block, 5 min; protein block, 10 min, Dako X0909) before 30 min application of primary antibody (B12 or Clone 4F10TT; 1, 2, or 4 μg/me. In some controls, slides were not exposed to primary antibody. After washing, HRP-labelled goat anti-mouse secondary antibody was applied (20 min) and slides were washed again, followed by 10 min application of DAB+Chromogen. Slides were washed, counterstained with haematoxylin and coverslipped for microscopic visualisation of staining.
(84) In cryosections of RA synovium that were fixed using acetone/methanol, the anti-TNC FBG B12 mouse IgG2a showed a very similar pattern of staining to that obtained with the positive control antibody Clone 4F10TT. Specific immunostaining was observed in the synovium, fibrous capsule, vasculature and within the interstitium. There was no staining within lymphoid aggregates. Some non-specific immunostaining was present in non-immune control treated tissues. These results confirm and extend previous reports of tenascin-C expression within RA synovium, demonstrating that B12 is an effective agent for binding endogenous tenascin-C at sites of inflammation and further indicating that FBG is an accessible target in RA.
Example 9—Antibody Sequences
(85) CDRs in VH and VL sequences indicated by boxes.
(86) TABLE-US-00014 Antibody B12 VH CDR1: DYAMH (SEQ ID NO: 3) VH CDR2: GISGSGGSTYYADSVKG (SEQ ID NO: 4) VH CDR3: DISAVPDTFDI (SEQ ID NO: 5) VH amino acid sequence:
Example 13—Activity of the C3 Antibody In Vitro
(87) In order to confirm that the monoclonal antibody C3 (165_13_C3) acts by disrupting the binding of TNC-FBG to its receptor TLR4, first an in vitro binding assay was developed for TLR4 and Fc-His-FBG then the effect of pre-incubation of Fc-His-FBG with C3 was determined.
(88) Recombinant human TLR4 (R&D systems) (1 ug/ml (14.6 nM)) in PBS (or PBS alone) was bound to a 96-well plate. After blocking (10% BSA) the indicated concentrations of Human Fc-His-FBG was added and detection was carried out by incubation of an anti-human IgG1 MAb (AbD Serotec, clone 2C11) at 1 ug/ml, an anti-mouse HRP conjugated secondary antibody (AbD Serotec, STAR13B) at 1 ug/ml, and TMB substrate. The results are shown in
(89) As shown in
Example 14—Anti-Inflammatory Effect of Antibodies B12 and C3
(90) It was confirmed that the anti-TNC-FBG antibodies B12, and C3 (165_13_C3) have an anti-inflammatory effect in a biological system. To do this, human monocytes were isolated from peripheral blood (London blood bank) by Ficoll gradient and counter-flow centrifugation. The monocytes were then differentiated with 100 ng/ml M-CSF (Peprotec) for 5 days to produce M2 macrophages.
(91) As shown by the results in
(92)
(93) To confirm that the FBG-induced cytokine release was induced by the FBG rather than the Fc portion of the protein, a protein where the Fc portion is mutated to be inactive (Fc-Mut-FBG) was used, Anti-TNC-FBG antibodies, B12, C3 (165_13_C3) and A4 (160_01_A4) were also tested for activity against this molecule. Fc-Mut-FBG (1 uM) and C3, A4 or B12 (1 uM) were pre-incubated for 30 min at RT before being added to human M2 macrophage cultures. After 24 h supernatants were taken and subjected to cytokine ELISA. n=3, mean and SEM shown. Results are shown in
(94)
(95)
(96) To take the C3 antibody into animal studies, IgG4 B12 165-13-C3 product was cloned, expressed and purified at a leading contract manufacturing organisation using a commercial GS-CHO expression. cDNAs for the heavy and light chain variable regions were optimised for CHO expression and synthesised (with commercial signal sequences) by Life Technologies prior to cloning into the expression vectors. CHO cells were transfected as pools and the highest expressing pool was taken forward into large-scale shake flask production (22 L—11×2 L in SL shake flasks). Proprietary feeds were administered on day 4 and 8 prior to harvesting the culture on day 12. Material was centrifuged prior to depth filtration and filter sterilisation. Approximately a 5.5 fold concentration of material was performed using tangential flow filtration (30 kDa molecular weight cut off) and the resulting concentrate was filter sterilised again prior to MabSelect SuRe purification. The product was eluted and product was neutralised and then concentrated/diafiltered to approximately 11 mg/mL in 20 mM NaOAc, pH 5.5, 150 mM NaCl. Reduced and non-reduced SDS-PAGE analysis together with size exclusion—HPLC showed material that was highly pure and greater than 98% monomer. Endotoxin was less than 0.1Eu per mg.
(97) In this experiment the potency of the larger scale antibody batch was compared to the current smaller scale batch. Recombinant Human tenascin-C FBG (1 uM) was pre-incubated for 30 min at RT with MAb C3 (1, 0.2 and 0.04 uM) or isotype control MAb (1 uM) before being added in triplicate to Human M2 macrophage cultures. After 24 h supernatants were taken and subjected to cytokine ELISA. n=1, Ico=laboratory scale Lon=larger scale material. This experiment shows that both batches of antibodies show equal potency in the reduction of FBG-induced cytokine synthesis, i.e. the results are consistent irrespective of production.
Example 15—Monoclonal Antibody C3 (165_13_C3) Reduces the Production of Pro-Inflammatory Cytokines by RA Synovial Fibroblasts Stimulated with Human TNC-FBG
(98) It has been reported that synovial fibroblasts could be an important source of pro-inflammatory cytokine release in RA (R Bucala et al. (1991) Constitutive Production of Mitogenic and Inflammatory Cytokines by Rheumatoid Synovial Fibroblasts. J. Exp. Med. 173:569-574), it was therefore tested whether the C3 antibody also showed similar effects on FBG-induced cytokine release as in the macrophages.
(99) Human RA fibroblasts were grown out of donor RA synovial tissue by digestion of the tissue in RPMI (Lonza) containing 0.5 mg/ml Liberase (Roche) and 0.2 mg/ml DNase (Roche) and incubation at 37° C. for 1-1.5 h. The resulting tissue was pipetted through a 200 μm nylon mesh; the material that did not pass through the mesh was put into a petri-dish containing RPMI with 10% FBS (Life technologies) and 1% pen/strep (Life technologies) and incubated at 37° C. for 5 days. After 5 days synovial fibroblasts grow out of the tissue and the remaining tissue was removed from the RA synovial fibroblast (RASF) culture which was subsequently maintained in DMEM (Lonza) containing 10% FBS and 1% pen/strep. For this experiment RASF were plated out at 1×10.sup.4 cells/well. Recombinant Human TNC-FBG (1 uM) was pre-incubated for 30 min at RT with MAb C3 (1, 0.2 and 0.04 uM) or isotype control MAb (1 uM) before being added in triplicate to the synovial fibroblast cultures. After 24 h supernatants were taken and subjected to cytokine ELISA. n=1, mean and SEM shown (see
(100) These results indicate that C3 acts to reduce FBG induced pro-inflammatory cytokine release (both IL-8 and IL-6) in RA synovial fibroblasts, showing that this is a potential mechanism in multiple cell types found in the inflamed RA joint.
Example 16—Levels of Tenascin-C in Rat Model
(101) Expression of tenascin-C in both mouse and rat CIA (collagen-induced arthritis) models was confirmed and disease activity shown to correlate with clinical score.
(102)
Example 17—Evaluation of C3 Antibody in a Rat Model of Collagen-Induced Arthritis
(103) IgG4 C3 (165_13_C3) was tested for therapeutic activity in the standard rat collagen induced arthritis model. Adult male Lewis rats were randomly allocated to experimental groups and allowed to acclimatise for one week. On Day 0, animals were administered with 500 μl of a 1 mg/ml emulsion of type II bovine collagen in incomplete Freund's adjuvant (CII/IFA) by intra-dermal injection in the lower back. On Day 7, animals received a second injection of CII/IFA. Injections were performed under gas (isoflurane) anaesthesia. Treatments were administered according to the Administration Schedule shown below in Table 12.
(104) TABLE-US-00015 TABLE 12 Administration Schedule Disease Group Treatment Dose Route Regimen Induction 1 Vehicle (0.9% NaCl) n/a IV Twice weekly* Day 0, Day 7: 2 Control IgG4 .sup.1 10 mg/kg IV Day 0-End CII/IFA, ID 3 IgG4 165_13_C3 1 mg/kg IV 4 IgG4 165_13_C3 3 mg/kg IV 5 IgG4 165_13_C3 10 mg/kg IV .sup.1 Fully human IgG4 isotype control, preclinical grade, (ET904, Eureka Therapeutics), n/a: not applicable, IV: intra-venous injections, ID: intra-dermal injections, CII/IFA: Type II collagen and Incomplete Freund's Adjuvant emulsion, *Day 0, Day 3, Day 7, Day 10, Day 14, Day 17, Day 21 and Day 24
(105) From Day 7 until the end of the experiment, animals were scored three times per week for clinical signs of arthritis by an experimenter blind to the treatments. On Day 0, Day 14, Day 21 and Day 28, paw volumes were measured using a plethysmometer by an experimenter blind to the treatments.
(106) Results
(107) Non-Specific Clinical Observations
(108) From Day 0 until the end of the experiment, animals were checked daily for non-specific clinical signs to include abnormal posture (hunched), abnormal coat condition (piloerection) and abnormal activity levels (reduced or increased activity). One animal in Group 6 (ID #6.9, antibody 10 mg/kg-treated) did not recover from the isoflurane anaesthesia on Day 21. Animals did not show any non-specific clinical signs such as abnormal posture, abnormal coat condition and abnormal activity levels. One animal in Group 1 (ID #1.10, vehicle-treated) was culled on Day 22, prior to the end of the experiment, due to the severity of the clinical signs of arthritis.
(109) Clinical Scores
(110) From Day 7 until the end of the experiment, animals were scored three times per week for clinical signs of arthritis to include front and hind limb swelling. The experimenter was blind to the treatments. Each limb was scored on a five-point scale: (0) absence of swelling, (1) slight swelling and/or erythema, (2) mild swelling, (3) moderate swelling and (4) severe swelling and/or joint rigidity. A clinical score was calculated for each animal by adding the score of each limb. Data provided in
(111) Paw Volumes
(112) On Day 0, Day 14, Day 21 and Day 28, hind paw volumes were measured using a plethysmometer (water-displacement device). Measurements were performed under gas (isoflurane) anaesthesia. The experimenter was blind to the treatment. Right and left hind paw volumes from each animal on each experimental day were averaged.
(113) Paw volumes measured in the vehicle-treated group increased significantly from Day 14 until the end of the experiment on Day 28 when compared to the paw volumes measured on Day 0 (p<0.01 on Day 14, p<0.0001 on Day 21 and Day 28). The control IgG4 and 1 mg/kg IgG4 C3 dose groups did not induced any difference in hind paw volumes when compared to the vehicle-treated group between Day 0 and Day 28. IgG4 C3 administered at 3 mg/kg induced a significant decrease of the hind paw volumes when compared to the vehicle-treated group on Day 28 (p<0.01). IgG4 C3 administered at 10 mg/kg induced a significant decrease of the hind paw volumes when compared to the vehicle-treated group on Day 21 (p<0.05) and Day 28 (p<0.01).
(114) Conclusions
(115) The test antibody, IgG4 C3 (165_13_C3), when administered at 3 mg/kg or 10 mg/kg, significantly reduced the severity of the clinical signs.
Example 18—Protocol for In Vivo Testing
(116) Adult male Lewis rats randomly allocated to experimental groups and allowed to acclimatise for one week are employed. On Day 0, animals are administered with 500 μl of a 1 mg/ml emulsion of type II bovine collagen in incomplete Freund's adjuvant (CII/IFA) by intra-dermal injections in the lower back. On Day 7, animals receive a second injection of CII/IFA. Injections are performed under gas (isoflurane) anaesthesia. Treatments are administered according to the Administration Schedule below. From Day 0 until the end of the experiment, animals will be weighed three times per week. From Day 7 until the end of the experiment, animals are scored three times per week for clinical signs of arthritis by an experimenter blind to the treatments. On Day 0, Day 14, Day 21 and Day 28, paw volumes are measured using a plethysmometer by an experimenter blind to the treatments.
(117) Treatment Groups and Dosages
(118) Treatment groups and dosages are summarised in Table 13. Vehicle for test compounds was a 0.9% Sodium Chloride solution (Saline). Administration volume for intra-venous injection was 5 ml/kg. All groups are n=10.
(119) TABLE-US-00016 TABLE 13 Treatment groups and dosages Treatment Disease Group Route Regimen Induction 1 Vehicle (Saline) IV Three times per week: Days 0, 2, Day 0: CII/IFA 4, 7, 9, 11, 14, 16, 18, 21, 23, 25 Day 7: CII/IFA 2 Methotrexate 1 mg/kg IP Twice weekly: Day 0-End 3 IgG4.sup.1 165_13_C3 IV Three times per week: Days 0, 2, (NSCT-121), 10 mg/kg 4, 7, 9, 11, 14, 16, 18, 21, 23, 25 4 IgG4.sup.1 165_13_C3 Days 0,1, 4, 7, 10, 14, 17, 21, 24 (NSCT-121), 30 mg/kg 5 IgG4.sup.1 165_13_C3* Three times per week: Days 0, 2, (NSCT-141), 10 mg/kg 4, 7, 9, 11, 14, 16, 18, 21, 23, 25 6 IgG4.sup.1 165_13_C3* Days 0, 1, 4, 7, 10, 14, 17, 21, 24 (NSCT-141), 30 mg/kg n/a: not applicable, IV: intra-venous injection, IP: intra-peritoneal injection, CII: Type II collagen, IFA: incomplete Freund's adjuvant, .sup.1Hinge modified IgG4 (S241P; Angal etal, 1993).
(120) Clinical Scores
(121) From Day 7 until the end of the experiment, animals are scored three times per week for clinical signs of arthritis to include front and hind limb swelling. The experimenter is blind to the treatments. Each limb is scored on a five-point scale: (0) absence of swelling, (1) slight swelling and/or erythema, (2) mild swelling, (3) moderate swelling and (4) severe swelling and/or joint rigidity. A clinical score is calculated for each animal by adding the score of each limb. In the model methotrexate, which is a control, reduces clinical systems. The C3 antibody reduces clinical systems, the C3* antibody shows activity similar or greater than C3 in this model.
Example 19—Biacore Analysis of C3 and C3* Antibodies Human and Mouse TNC FBG and Human TNR FBG
(122) Kinetic Assay: Characterisation of Binding of Human, and Mouse TNC rCd4-his-FBG and Human TNR rCd4-his-FBG to NSCT Antibodies
(123) SPR experiments were performed to test whether the germline changes made in 165_13_C3* have retained the affinity or specificity of 165_13_C3. SPR experiments were performed using a BIAcore T200 instrument (GE Healthcare) according to the protocol of the Human antibody capture kit (GE, BR-1008-39).
(124) For the kinetic assay ˜1,500-1,900 response units (RU) of anti-human Fc IgG (GE, BR-1008-39) was immobilised in all flow-cells (FC1-4) on a Series S CM5 sensor chip (BR-1005-30) using EDC/NHS cross-linking chemistry according to the amine coupling kit protocol (GE, BR-1000-50). Purified NSCT antibody was diluted in HBS-P+ (Hepes pH 7.4, 150 mM NaCl, 0.05% Tween 20) running buffer to a concentration of 3.5 nM and injected into FC2 and FC4 (using a flowrate 10 μL/min and 115 s contact time) at 25° C. Antibody capture levels ranged typically from 70 RU (for TNR interactions) to 90 RU (for TNC interactions).
(125) A 1:1 dilution series of antigen in HBS-P+ (30 nM of human and mouse TNC rCd4-His-FBG and 480 nM of human TNR rCd4-His-FBG) was injected with a flow-path via the reference flow cell (FC1 and FC3, respectively) and the antibody capture flow cell (FC2 and FC4, respectively) with a flow rate of 30 μl/min. The association and dissociation phases were measured over a 3.5 and 30 min time period, respectively, for human and mouse TNC rCd4-His-FBG, and 2 and 8 min time period for human TNR rCd4-His-FBG. For interaction analysis with human and mouse TNC rCD4-His-FBG regeneration of free capture antibody surfaces after each antigen injection was done by injection of 3 M MgCl2 for 60 s and for each sample injection a fresh NSCT antibody was captured. Carry over of antigen was prohibited by an extra needle wash step with 50% DMSO before every injection. Runs were performed at the more physiologically relevant temperature of 37° C. Kinetic parameters were determined by reference cell subtraction and fitting the sensorgram experimental data to a 1:1 interaction model using the BIAcore T200 Evaluation software version 2.0. Conclusion: 165_13_C3 and 165_13_C3* show comparable binding to the TNC FBG and TNR FBG proteins tested.
(126) TABLE-US-00017 TABLE 14 Kinetic binding data for initial protein samples determined by surface plasmon resonance (SPR) spectroscopy at 37° C. Tmax SA ka SE kd SE Chi.sup.2 U- Protein K.sub.D(M) (RU) (%) (1/Ms) (ka) (1/s) (kd) (RU.sup.2) value C3 hTNR 5.27e−8 13 184 1.127e6 3.2e3 0.059 1.7e−4 0.0260 1 hTNC 1.24e−9 17 298 3.398e6 6.4e2 0.0042 6.8e−7 0.214 1 mTNC 4.33e−10 17 303 9.272e6 5.0e3 0.0040 2.3e−6 0.0801 1 C3* hTNR 5.31e−8 15 233 1.108e6 1.0e3 0.058 4.2e−5 0.0656 1 hTNC 4.43e−10 14 297 9.350e6 5.8e3 0.0041 2.7e−6 0.0750 1 mTNC 6.70e−10 14 304 6.275e6 6.8e3 0.0042 2.2e−6 0.644 2 KD, equilibrium constant (M); T Rmax, theoretical maximal binding level of ligand in response units (RU); SA, surface activity in percent (%); ka, association constant (1/Ms); kd, dissociation constant (1/s); Chi.sup.2, value is a statistical measure of the closeness of fit (typically <2), The U-value is an additional indicator of the parameter significance. This is a parameter that represents the uniqueness of the calculated rate constants and R max, determined by testing the dependence of fitting on correlated variations between selected variables. Lower values indicate greater confidence in the results. A high value (above about 10) indicates that the reported kinetic constants contain no useful information.
Example 20 Expression of 165_13_C3 and 165_13_C3* in a CHO-Pool Shake Flask Production
(127) 165_13_C3 and 165_13_C3* were cloned into a GS-CHO expression vector with a hinge modified IgG4 heavy chain constant region. Cell lines were generated and material expressed and purified as described in example 14. Antibody titre data from the cell culture supernatant prior to concentration and affinity purification is provided in table 15. Expression of IgG4 165_13_C3 * was more than three-fold higher than IgG4 165_13_C3.
(128) TABLE-US-00018 TABLE 15 Titre from cell culture supernatant at the end of production Culture Titre Product Lot Number Volume (L) (mg/L) IgG4 165_13_C3* 357-180516-1 10 375 IgG4 165_13_C3 Pooled 237-230115-01 22 111 and 237-260115-01 IgG4 165_13_C3 443-231116-01 20 110