MUTANT INTERLEUKIN-2 POLYPEPTIDES
20190322765 ยท 2019-10-24
Assignee
Inventors
- Oliver Ast (Bassersdorf, CH)
- Peter Bruenker (Hittnau, CH)
- ANNE FREIMOSER-GRUNDSCHOBER (ZUERICH, CH)
- Sylvia Herter (Regensdorf, CH)
- Thomas U. Hofer (Zuerich, CH)
- Ralf Hosse (Cham, CH)
- Christian Klein (Bonstetten, CH)
- Ekkehard Moessner (Kreuzlingen, CH)
- Valeria G. Nicolini (Erlenbach/ZH, CH)
- Pablo Umana (Wollerau, CH)
Cpc classification
A61K47/6891
HUMAN NECESSITIES
C07K16/28
CHEMISTRY; METALLURGY
A61K47/6849
HUMAN NECESSITIES
A61K47/6813
HUMAN NECESSITIES
A61P35/00
HUMAN NECESSITIES
International classification
A61K47/68
HUMAN NECESSITIES
A61K39/395
HUMAN NECESSITIES
Abstract
The present invention generally relates to mutant interleukin-2 polypeptides that exhibit reduced affinity to the -subunit of the IL-2 receptor, for use as immunotherapeutic agents. In addition, the invention relates to immunoconjugates comprising said mutant IL-2 polypeptides, polynucleotide molecules encoding the mutant IL-2 polypeptides or immunoconjugates, and vectors and host cells comprising such polynucleotide molecules. The invention further relates to methods for producing the mutant IL-2 polypeptides or immunoconjugates, pharmaceutical compositions comprising the same, and uses thereof.
Claims
1. A mutant interleukin-2 (IL-2) polypeptide comprising at a first amino acid mutation that abolishes or reduces affinity of the mutant IL-2 polypeptide to the high-affinity IL-2 receptor and preserves affinity of the mutant IL-2 polypeptide to the intermediate-affinity IL-2 receptor, each compared to a wild-type IL-2 polypeptide, characterized in that said first amino acid mutation is at a position corresponding to residue 72 of human IL-2.
2. The mutant interleukin-2 polypeptide of claim 1, wherein said first amino acid mutation is an amino acid substitution, selected from the group of L72G, L72A, L72S, L72T, L72Q, L72E, L72N, L72D, L72R, and L72K.
3. The mutant interleukin-2 polypeptide of claim 1 or 2, comprising a second amino acid mutation that abolishes or reduces affinity of the mutant IL-2 polypeptide to the high-affinity IL-2 receptor and preserves affinity of the mutant IL-2 polypeptide to the intermediate-affinity IL-2 receptor, each compared to a wild-type IL-2 polypeptide.
4. The mutant interleukin-2 polypeptide of claim 3, wherein said second amino acid mutation is at a position selected from the positions corresponding to residue 35, 38, 42, 43, and 45 of human IL-2.
5. The mutant interleukin-2 polypeptide of claim 3 or 4, wherein said second amino acid mutation is at a position corresponding to residue 42 of human IL-2.
6. The mutant interleukin-2 polypeptide of claim 5, wherein said second amino acid mutation is an amino acid substitution, selected from the group of F42A, F42G, F42S, F42T, F42Q, F42E, F42N, F42D, F42R, and F42K.
7. The mutant interleukin-2 polypeptide of any one of claims 3 to 6, comprising a third amino acid mutation that abolishes or reduces affinity of the mutant IL-2 polypeptide to the high-affinity IL-2 receptor and preserves affinity of the mutant IL-2 polypeptide to the intermediate-affinity IL-2 receptor, each compared to a wild-type IL-2 polypeptide.
8. The mutant interleukin-2 polypeptide of any one of claims 1 to 7, comprising three amino acid mutations that abolish or reduce affinity of the mutant IL-2 polypeptide to the high-affinity IL-2 receptor and preserve affinity of the mutant IL-2 polypeptide to the intermediate-affinity IL-2 receptor, each compared to a wild-type IL-2 polypeptide, wherein said three amino acid mutations are at positions corresponding to residue 42, 45, and 72 of human IL-2.
9. The mutant interleukin-2 polypeptide of claim 8, wherein said three amino acid mutations are amino acid substitutions selected from the group of F42A, F42G, F42S, F42T, F42Q, F42E, F42N, F42D, F42R, F42K, Y45A, Y45G, Y45S, Y45T, Y45Q, Y45E, Y45N, Y45D, Y45R, Y45K, L72G, L72A, L72S, L72T, L72Q, L72E, L72N, L72D, L72R, and L72K.
10. The mutant interleukin-2 polypeptide of any one of claims 1 to 9, further comprising an amino acid mutation which eliminates the O-glycosylation site of IL-2 at a position corresponding to residue 3 of human IL-2.
11. The mutant interleukin-2 polypeptide of any one of claims 1 to 10, wherein said mutant IL-2 polypeptide is linked to a non-IL-2 moiety.
12. The mutant interleukin-2 polypeptide of any one of claims 1 to 11, wherein said mutant IL-2 polypeptide is linked to a first and a second non-IL-2 moiety.
13. The mutant interleukin-2 polypeptide of claim 12, wherein said mutant IL-2 polypeptide shares a carboxy-terminal peptide bond with said first non-IL-2 moiety and an amino-terminal peptide bond with said second non-IL-2 moiety.
14. The mutant interleukin-2 polypeptide of any one of claims 11 to 13, wherein said non-IL-2 moiety is an antigen binding moiety.
15. An immunoconjugate comprising a mutant IL-2 polypeptide according to any one of claims 1 to 10 and an antigen binding moiety.
16. The immunoconjugate of claim 15, wherein said mutant IL-2 polypeptide shares an amino- or carboxy-terminal peptide bond with said antigen binding moiety.
17. The immunoconjugate of claim 15 or 16, wherein said immunoconjugate comprises as first and a second antigen binding moiety.
18. The immunoconjugate of claim 17, wherein said mutant IL-2 polypeptide shares an amino- or carboxy-terminal peptide bond with said first antigen binding moiety and said second antigen binding moiety shares an amino- or carboxy-terminal peptide bond with either i) said mutant IL-2 polypeptide or ii) said first antigen binding moiety.
19. The mutant interleukin-2 polypeptide of claim 14 or the immunoconjugate of any one of claims 15 to 18, wherein said antigen binding moiety is an antibody or an antibody fragment.
20. The mutant interleukin-2 polypeptide of claim 14 or the immunoconjugate of any one of claims 15 to 18, wherein said antigen binding moiety is selected from a Fab molecule and a scFv molecule.
21. The mutant interleukin-2 polypeptide of claim 14 or the immunoconjugate of any one of claims 15 to 18, wherein said antigen binding moiety is an immunoglobulin molecule, particularly an IgG molecule.
22. The mutant interleukin-2 polypeptide of claim 14 or the immunoconjugate of any one of claims 15 to 21, wherein said antigen binding moiety is directed to an antigen presented on a tumor cell or in a tumor cell environment.
23. The mutant interleukin-2 polypeptide or immunoconjugate of claim 22, wherein said antigen is selected from the group of Fibroblast Activation Protein (FAP), the A1 domain of Tenascin-C (TNC A1), the A2 domain of Tenascin-C (TNC A2), the Extra Domain B of Fibronectin (EDB), Carcinoembryonic Antigen (CEA) and the Melanoma-associated Chondroitin Sulfate Proteoglycan (MCSP).
24. An isolated polynucleotide encoding the mutant IL-2 polypeptide or immunoconjugate of any one of claims 1 to 23.
25. An expression vector comprising the polynucleotide of claim 24.
26. A host cell comprising the polynucleotide of claim 24 or the expression vector of claim 25.
27. A method of producing a mutant IL-2 polypeptide or an immunoconjugate thereof, comprising culturing the host cell of claim 26 under conditions suitable for the expression of the mutant IL-2 polypeptide or the immunoconjugate.
28. A mutant IL-2 polypeptide or immunoconjugate produced by the method of claim 27.
29. A pharmaceutical composition comprising the mutant IL-2 polypeptide or immunoconjugate of any one of claims 1 to 23 or 28 and a pharmaceutically acceptable carrier.
30. The mutant IL-2 polypeptide or immunoconjugate of any one of claims 1 to 23 or 28 for use in the treatment of a disease in an individual in need thereof.
31. The mutant IL-2 polypeptide or immunoconjugate of claim 30, wherein said disease is cancer.
32. Use of the mutant IL-2 polypeptide or immunoconjugate of any one of claims 1 to 23 or 28 for manufacture of a medicament for treating a disease in an individual in need thereof.
33. A method of treating disease in an individual, comprising administering to said individual a therapeutically effective amount of a composition comprising the mutant IL-2 polypeptide or immunoconjugate of any one of claims 1 to 23 or 28 in a pharmaceutically acceptable form.
34. The method of claim 33, wherein said disease is cancer.
35. A method of stimulating the immune system of an individual, comprising administering to said individual a effective amount of a composition comprising the mutant IL-2 polypeptide or immunoconjugate of any one of claims 1 to 23 or 28 in a pharmaceutically acceptable form.
36. The invention as described hereinbefore.
Description
SHORT DESCRIPTION OF THE FIGURES
[0234]
[0235]
[0236]
[0237]
[0238]
[0239]
[0240]
[0241]
[0242]
[0243]
[0244]
[0245]
[0246]
[0247]
[0248]
[0249]
[0250]
[0251]
[0252]
[0253]
[0254]
[0255]
[0256]
[0257]
[0258]
[0259]
[0260]
[0261]
[0262]
[0263]
[0264]
[0265]
[0266]
[0267]
[0268]
[0269]
[0270]
[0271]
[0272]
[0273]
[0274]
[0275]
[0276]
[0277]
[0278]
[0279]
[0280]
[0281]
[0282]
[0283]
EXAMPLES
[0284] The following are examples of methods and compositions of the invention. It is understood that various other embodiments may be practiced, given the general description provided above.
Example 1
[0285] General Methods
[0286] Recombinant DNA Techniques
[0287] Standard methods were used to manipulate DNA as described in Sambrook et al., Molecular cloning: A laboratory manual; Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y., 1989. The molecular biological reagents were used according to the manufacturer's instructions. General information regarding the nucleotide sequences of human immunoglobulins light and heavy chains is given in: Kabat, E. A. et al., (1991) Sequences of Proteins of Immunological Interest, Fifth Ed., NIH Publication No 91-3242.
[0288] DNA Sequencing
[0289] DNA sequences were determined by double strand sequencing.
[0290] Gene Synthesis
[0291] Desired gene segments where required were either generated by PCR using appropriate templates or were synthesized by Geneart AG (Regensburg, Germany) from synthetic oligonucleotides and PCR products by automated gene synthesis. In cases where no exact gene sequence was available, oligonucleotide primers were designed based on sequences from closest homologues and the genes were isolated by RT-PCR from RNA originating from the appropriate tissue. The gene segments flanked by singular restriction endonuclease cleavage sites were cloned into standard cloning/sequencing vectors. The plasmid DNA was purified from transformed bacteria and concentration determined by UV spectroscopy. The DNA sequence of the subcloned gene fragments was confirmed by DNA sequencing. Gene segments were designed with suitable restriction sites to allow sub-cloning into the respective expression vectors. All constructs were designed with a 5-end DNA sequence coding for a leader peptide which targets proteins for secretion in eukaryotic cells. SEQ ID NOs 263-273 give exemplary leader peptides and polynucleotide sequences encoding them.
[0292] Preparation of IL-2R Subunit-Fc Fusions and IL-2R Subunit Fc Fusion
[0293] To study IL-2 receptor binding affinity, a tool was generated that allowed for the expression of a heterodimeric IL-2 receptor, the -subunit of the IL-2 receptor was fused to an Fc molecule that was engineered to heterodimerize (Fc(hole)) (see SEQ ID NOs 274 and 275) using the knobs-into-holes technology (Merchant et al., Nat Biotech. 16, 677-681 (1998)). The -subunit of the IL-2 receptor was then fused to the Fc(knob) variant (see SEQ ID NOs 276 and 277), which heterodimerized with Fc(hole). This heterodimeric Fc-fusion protein was then used as a substrate for analyzing the IL2/IL-2 receptor interaction. The IL-2R -subunit was expressed as monomeric chain with an AcTev cleavage site and an Avi His tag (SEQ ID NOs 278 and 279). The respective IL-2R subunits were transiently expressed in HEK EBNA 293 with serum for the IL-2R subunit construct and without serum for the -subunit construct. The IL-2R subunit construct was purified on protein A (GE Healthcare), followed by size exclusion chromatography (GE Healthcare, Superdex 200). The IL-2R -subunit was purified via His tag on a NiNTA column (Qiagen) followed by size exclusion chromatography (GE Healthcare, Superdex 75).
[0294] Preparation of Immunoconjugates
[0295] Details about the preparation and purification of Fab-IL-2-Fab immunoconjugates, including generation and affinity maturation of antigen binding moieties can be found in the Examples appended to PCT publication no. WO 2011/020783, which is incorporated herein by reference in its entirety. As described therein, various antigen binding domains directed to FAP have been generated by phage display, including the ones designated 4G8, 3F2, 28H1, 29B11, 14B3, and 4B9 used in the following examples. Clone 28H1 is an affinity matured antibody based on parental clone 4G8, while clones 29B11, 14B3 and 4B9 are affinity matured antibodies based on parental clone 3F2. The antigen binding domain designated MHLG1 KV9 used herein is directed to MCSP.
[0296] The sequences of immunoconjugates comprising wild-type IL-2 that were used in the following examples can also be found in PCT publication no. WO 2011/020783. The sequences corresponding to the immunoconjugates comprising quadruple mutant IL-2 that were used in the following examples are: 4G8: SEQ ID NOs 211 and 233; 3F2: SEQ ID NOs 209 and 231; 28H1: SEQ ID NOs 219 and 233; 29B11: SEQ ID NOs 221 and 231; 14B3: SEQ ID NOs 229 and 231; 4B9: SEQ ID NOs 227 and 231; MHLG1-KV9: SEQ ID NOs 253 and 255. The DNA sequences were generated by gene synthesis and/or classical molecular biology techniques and subcloned into mammalian expression vectors (one for the light chain and one for the heavy chain/IL-2 fusion protein) under the control of an MPSV promoter and upstream of a synthetic polyA site, each vector carrying an EBV OriP sequence. Immunoconjugates as applied in the examples below were produced by co-transfecting exponentially growing HEK293-EBNA cells with the mammalian expression vectors using calcium phosphate-transfection. Alternatively, HEK293 cells growing in suspension were transfected by polyethylenimine (PEI) with the respective expression vectors. Alternatively, stably transfected CHO cell pools or CHO cell clones were used for production in serum-free media. While 4G8-based FAP-targeted Fab-IL-2-Fab constructs comprising wild-type or (quadruple) mutant IL-2 can be purified by affinity chromatography using a protein A matrix, affinity matured 28H1-based FAP-targeted Fab-IL-2-Fab constructs were purified by affinity chromatography on a protein G matrix in small scale.
[0297] Briefly, FAP-targeted 28H1 Fab-IL-2-Fab, comprising wild-type or (quadruple) mutant IL-2, was purified from cell supernatants by one affinity step (protein G) followed by size exclusion chromatography (Superdex 200, GE Healthcare). The protein G column was equilibrated in 20 mM sodium phosphate, 20 mM sodium citrate pH 7.5, supernatant was loaded, and the column was washed with 20 mM sodium phosphate, 20 mM sodium citrate pH 7.5. Fab-IL-2-Fab was eluted with 8.8 mM formic acid pH 3. The eluted fractions were pooled and polished by size exclusion chromatography in the final formulation buffer: 25 mM potassium phosphate, 125 mM sodium chloride, 100 mM glycine pH 6.7. Exemplary results from purification and analytics are given below.
[0298] FAP-targeted 3F2 Fab-IL-2-Fab or 4G8 Fab-IL-2-Fab, comprising wild-type or (quadruple) mutant IL-2, were purified by a similar method composed of one affinity step using protein A followed by size exclusion chromatography (Superdex 200, GE Healthcare). The protein A column was equilibrated in 20 mM sodium phosphate, 20 mM sodium citrate pH 7.5, supernatant was loaded, and the column was washed with 20 mM sodium phosphate, 20 mM sodium citrate, 500 mM sodium chloride pH 7.5, followed by a wash with 13.3 mM sodium phosphate, 20 mM sodium citrate, 500 mM sodium chloride pH 5.45. A third wash with 10 mM MES, 50 mM sodium chloride pH 5 was optionally performed. Fab-IL-2-Fab was eluted with 20 mM sodium citrate, 100 mM sodium chloride, 100 mM glycine, pH 3. The eluted fractions were pooled and polished by size exclusion chromatography in the final formulation buffer 25 mM potassium phosphate, 125 mM sodium chloride, 100 mM glycine pH 6.7. Exemplary detailed purification procedures and results are given for selected constructs below.
[0299] FAP-targeted IgG-IL-2 qm fusion proteins were generated based on the FAP-antibodies 4G8, 4B9 and 28H1, wherein one single IL-2 quadruple mutant (qm) was fused to the C-terminus of one heterodimeric heavy chain as shown in
[0300] Heterodimerization resulting in the presence of a single IL-2 quadruple mutant is achieved by application of the knob-into-hole technology. In order to minimize the generation of homodimeric IgG-cytokine fusions the cytokine was fused to the C-terminus (with deletion of the C-terminal Lys residue) of the knob-containing IgG heavy chain via a G.sub.4-(SG.sub.4).sub.2- or (G.sub.4S).sub.3-linker. The antibody-cytokine fusion has IgG-like properties. To reduce FcR binding/effector function and prevent FcR co-activation, P329G L234A L235A (LALA) mutations were introduced in the Fc domain. The sequences of these immunoconjugates are given in SEQ ID NOs 297, 299 and 233 (28H1), SEQ ID NOs 301, 303 and 231 (4B9), and SEQ ID NOs 315, 317 and 233 (4G8)). In addition, a CEA-targeted IgG-IL-2 qm fusion protein and a control DP47GS non-targeted IgG-IL-2 qm fusion protein wherein the IgG does not bind to a specified target was generated. The sequences of these immunoconjugates are given in SEQ ID NOs 305, 307 and 309 (DP47GS), and SEQ ID NOs 319, 321 and 323 (CH1A1A).
[0301] The IgG-IL-2 constructs were generated by transient expression in HEK293 EBNA cells and purified essentially as described above for the Fab-IL-2-Fab constructs. Briefly, IgG-IL-2 fusion proteins were purified by one affinity step with protein A (HiTrap ProtA, GE Healthcare) equilibrated in 20 mM sodium phosphate, 20 mM sodium citrate pH 7.5. After loading of the supernatant, the column was first washed with 20 mM sodium phosphate, 20 mM sodium citrate, pH 7.5 and subsequently washed with 13.3 mM sodium phosphate, 20 mM sodium citrate, 500 mM sodium chloride, pH 5.45. The IgG-cytokine fusion protein was eluted with 20 mM sodium citrate, 100 mM sodium chloride, 100 mM glycine, pH 3. Fractions were neutralized and pooled and purified by size exclusion chromatography (HiLoad 16/60 Superdex 200, GE Healthcare) in final formulation buffer: 25 mM potassium phosphate, 125 mM sodium chloride, 100 mM glycine pH 6.7. Exemplary detailed purification procedures and results are given for selected constructs below. The protein concentration of purified protein samples was determined by measuring the optical density (OD) at 280 nm, using the molar extinction coefficient calculated on the basis of the amino acid sequence. Purity and molecular weight of immunoconjugates were analyzed by SDS-PAGE in the presence and absence of a reducing agent (5 mM 1,4-dithiothreitol) and stained with Coomassie blue (SimpleBlue SafeStain, Invitrogen). The NuPAGE Pre-Cast gel system (Invitrogen) was used according to the manufacturer's instructions (4-20% Tris-glycine-gels or 3-12% Bis-Tris). The aggregate content of immunoconjugate samples was analyzed using a Superdex 200 10/300GL analytical size-exclusion column (GE Healthcare) in 2 mM MOPS, 150 mM NaCl, 0.02% NaN.sub.3, pH 7.3 running buffer at 25 C.
[0302] FAP Binding Affinity
[0303] The FAP binding activity of the cleaved Fab fragments used in these examples as antigen binding, moieties was determined by surface plasmon resonance (SPR) on a Biacore machine. Briefly, an anti-His antibody (Penta-His, Qiagen 34660) was immobilized on CM5 chips to capture 10 nM human, murine or cynomolgus FAP-His (20 s). Temperature was 25 C. and HBS-EP was used as buffer. Fab analyte concentration was 100 nM down to 0.41 nM (duplicates) at a flow rate of 50 l/min (association: 300 s, dissociation: 600 s (4B9, 14B3, 29B11, 3F2) or 1200 s (28H1, 4G8), regeneration: 60 s 10 mM glycine pH 2). Fitting was performed based on a 1:1 binding model, RI=0, Rmax=local (because of capture format). Table 2 gives the monovalent affinities as determined by SPR.
TABLE-US-00002 TABLE 2 Affinity (K.sub.D) of FAP-targeted Fab fragments to FAP as determined by SPR. K.sub.D in nM Human FAP Cynomolgus FAP Murine FAP 4B9 Fab 0.3 0.23 5 0.31 0.24 5.2 14B3 Fab 0.47 0.61 4.7 0.47 0.59 4.7 29B11 Fab 0.19 0.21 1.3 0.19 0.2 1.2 3F2 Fab 6 4.7 8.9 6 5.3 9.5 28H1 Fab 2.6 3.7 0.13 2.6 3.7 0.18 4G8 Fab 53 (48 steady state) 33 (33 steady state) 0.07 51 (48 steady state) 35 (34 steady state) 0.07
[0304] Biological Activity Assays with Targeted IL-2 Immunoconjugates
[0305] The biological activity of FAP- or MCSP-targeted Fab-IL-2-Fab immunoconjugates and of FAP-targeted IgG-IL-2 immunoconjugates, comprising wild-type or (quadruple) mutant IL-2, was investigated in several cellular assays in comparison to commercially available IL-2 (Proleukin, Novartis, Chiron).
[0306] IFN- Release by NK Cells (in Solution)
[0307] IL-2 starved NK92 cells (100000 cells/well in 96-U-well plate) were incubated with different concentrations of IL-2 immunoconjugates, comprising wild-type or (quadruple) mutant IL-2, for 24 h in NK medium (MEM alpha from Invitrogen (#22561-021) supplemented with 10% FCS, 10% horse serum, 0.1 mM 2-mercaptoethanol, 0.2 mM inositol and 0.02 mM folic acid). Supernatants were harvested and the IFN- release was analysed using the anti-human IFN- ELISA Kit II from Becton Dickinson (#550612). Proleukin (Novartis) served as positive control for IL-2-mediated activation of the cells.
[0308] NK Cell Proliferation
[0309] Blood from healthy volunteers was taken in heparin-containing syringes and PBMCs were isolated. Untouched human NK cells were isolated from the PBMCs using the Human NK Cell Isolation Kit II from Miltenyi Biotec (#130-091-152). The CD25 expression of the cells was checked by flow cytometry. For proliferation assays, 20000 isolated human NK cells were incubated for 2 days in a humidified incubator at 37 C., 5% CO.sub.2 in the presence of different IL-2 immunoconjugates, comprising wild-type or (quadruple) mutant IL-2. Proleukin (Novartis) served as control. After 2 days, the ATP content of the cell lysates was measured using the CellTiter-Glo Luminescent Cell Viability Assay from Promega (#G7571/2/3). The percentage of growth was calculated setting the highest Proleukin concentration to 100% proliferation and untreated cells without IL-2 stimulus to 0% proliferation.
[0310] STAT5 Phosphorylation Assay
[0311] Blood from healthy volunteers was taken in heparin-containing syringes and PBMCs were isolated. PBMCs were treated with IL-2 immunoconjugates, comprising wild-type or (quadruple) mutant IL-2, at the indicated concentrations or with Proleukin (Novartis) as control. After 20 min incubation at 37 C., PBMCs were fixed with pre-warmed Cytofix buffer (Becton Dickinson #554655) for 10 min at 37 C., followed by permeabilization with Phosflow Perm Buffer III (Becton Dickinson #558050) for 30 min at 4 C. Cells were washed twice with PBS containing 0.1% BSA before FACS staining was performed using mixtures of flow cytometry antibodies for detection of different cell populations and phosphorylation of STAT5. Samples were analysed using a FACSCantoII with HTS from Becton Dickinson.
[0312] NK cells were defined as CD3.sup.CD56.sup.+, CD8 positive T cells were defined as CD3.sup.+CD8.sup.+, CD4 positive T cells were defined as CD4.sup.+CD25.sup.CD127.sup.+ and T.sub.reg cells were defined as CD4.sup.+CD25.sup.+FoxP3.sup.+.
[0313] Proliferation and AICD of T Cells
[0314] Blood from healthy volunteers was taken in heparin-containing syringes and PBMCs were isolated. Untouched T cells were isolated using the Pan T Cell Isolation Kit II from Miltenyi Biotec (#130-091-156). T cells were pre-stimulated with 1 g/ml PHA-M (Sigma Aldrich #L8902) for 16 h before adding Proleukin or Fab-IL-2-Fab immunoconjugates, comprising wild-type or (quadruple) mutant IL-2, to the washed cells for another 5 days. After 5 days, the ATP content of the cell lysates was measured using the CellTiter-Glo Luminescent Cell Viability Assay from Promega (#G7571/2/3). The relative proliferation was calculated setting the highest Proleukin concentration to 100% proliferation.
[0315] Phosphatidylserine (PS) exposure and cell death of T cells were assayed by flow cytometric analysis (FACSCantoII, BD Biosciences) of annexin V (Annexin-V-FLUOS Staining Kit, Roche Applied Science) and propidium iodide (PI)-stained cells. To induce activation-induced cell death (AICD); the T cells were treated with an apoptosis-inducing anti-Fas antibody (Millipore clone Ch11) for 16 h after the 16 h PHA-M and 5 days treatment with Fab-IL-2-Fab immunoconjugates. Annexin V staining was performed according to the manufacturer's instructions. Briefly, cells were washed with Ann-V Binding Buffer (Ix stock: 0.01 M Hepes/NaOH pH7.4, 0.14 M NaCl, 2.5 mM CaCl.sub.2) and stained for 15 min at RT in the dark with Annexin V FITC (Roche); Cells were washed again in Ann-V-Binding buffer before addition of 200 l/well Ann-V-Binding Buffer containing PI (0.3 g/ml). The cells were analysed immediately by flow cytometry.
[0316] Binding to FAP Expressing Cells
[0317] Binding of FAP-targeted IgG-IL-2 qm and Fab-IL-2 qm-Fab immunoconjugates to human FAP expressed on stably transfected HEK293 cells was measured by FACS. Briefly, 250 000 cells per well were incubated with the indicated concentration of the immunoconjugates in a round-bottom 96-well plate, incubated for 30 min at 4 C., and washed once with PBS/0.1% BSA. Bound immunoconjugates were detected after incubation for 30 min at 4 C. with FITC-conjugated AffiniPure F(ab)2 Fragment goat anti-human F(ab)2 Specific (Jackson Immuno Research Lab #109-096-097, working solution: 1:20 diluted in PBS/0.1% BSA, freshly prepared) using a FACS CantoII (Software FACS Diva).
[0318] Analysis of FAP Internalization Upon Binding by FACS
[0319] For several FAP antibodies known in the art it is described that they induce FAP internalization upon binding (described e.g. in Baum et al., J Drug Target 15, 399-406 (2007); Bauer at al., Journal of Clinical Oncology, 2010 ASCO Annual Meeting Proceedings (Post-Meeting Edition), vol. 28 (May 20 Supplement), abstract no. 13062 (2010); Ostermann et al., Clin Cancer Res 14, 4584-4592 (2008)). Thus, we analyzed the internalization properties of our Fab-IL-2-Fab immunoconjugates. Briefly, GM05389 cells (human lung fibroblasts,) cultured in EMEM medium with 15% FCS, were detached, washed, counted, checked for viability and seeded at a density of 210.sup.5 cells/well in 12-well plates. The next day, FAP-targeted Fab-IL-2-Fab immunoconjugates were diluted in cold medium and allowed to bind to cell surface for 30 min on ice. The excess of unbound antibody was washed away using cold PBS and cells were further incubated in 0.5 ml complete pre-warmed medium at 37 C. for the indicated time periods. When the different time points were reached, cells were transferred on ice, washed once with cold PBS and incubated with the secondary antibody (FITC-conjugated AffiniPure F(ab)2 Fragment goat anti-human F(ab)2 specific, Jackson Immuno Research Lab #109-096-097, 1:20 dilution) for 30 min at 4 C. Cells were then washed twice with PBS/0.1% BSA, transferred to a 96-well plate, centrifuged for 4 min at 4 C., 400g and cell pellets were resuspended by vortexing. Cells were fixed using 100 l 2% PFA. For FACS measurement, cells were re-suspended in 200 l/sample PBS/0.1% BSA and measured with the plate protocol in FACS CantoII (Software FACS Diva).
Example 2
[0320] We designed mutated versions of IL-2 that comprised one or more of the following mutations (compared to the wild-type IL-2 sequence shown in SEQ ID NO: 1): [0321] 1. T3Aknockout of predicted O-glycosylation site [0322] 2. F42Aknockout of IL-2/IL-2R interaction [0323] 3. Y45Aknockout of IL-2/IL-2R interaction [0324] 4. L72Gknockout of IL-2/IL-2R interaction [0325] 5. C125Apreviously described mutation to avoid disulfide-bridged IL-2 dimers
[0326] A mutant IL-2 polypeptide comprising all of mutations 1-4 is denoted herein as IL-2 quadruple mutant (qm). It may further comprise mutation 5 (see SEQ ID NO: 19).
[0327] In addition to the three mutations F42A, Y45A and L72G designed to interfere with the binding to CD25, the T3A mutation was chosen to eliminate the O-glycosylation site and obtain a protein product with higher homogeneity and purity when the IL-2 qm polypeptide or immunoconjugate is expressed in eukaryotic cells such as CHO or HEK293 cells.
[0328] For purification purposes a His6 tag was introduced at the C-terminus linked via a VD sequence. For comparison a non-mutated analogous version of IL-2 was generated that only contained the C145A mutation to avoid undesired inter-molecular disulfide bridges (SEQ ID NO: 3). The respective molecular weights without signal sequence were 16423 D for naked IL-2 and 16169 D for the naked IL-2 qm. The wild-type and quadruple mutant IL-2 with His tag were transfected in HEK EBNA cells in serum-free medium (F17 medium) The filtered supernatant was buffer exchanged over a cross-flow, before loading it on a NiNTA Superflow Cartridge (5 ml, Qiagen). The column was washed with wash buffer 20 mM sodium phosphate, 0.5 M sodium chloride pH 7.4 and eluted with elution buffer. 20 mM sodium phosphate, 0.5 M sodium chloride 0.5 M imidazole pH 7.4. After loading the column was washed with 8 column volumes (CV) wash buffer, 10 CV 5% elution buffer (corresponds to 25 mM imidazole), then eluted with a gradient to 0.5 M imidazole. The pooled eluate was polished by size exclusion chromatography on a HiLoad 16/60 Superdex75 (GE Healthcare) column in 2 mM MOPS, 150 mM sodium chloride, 0.02% sodium azide pH 7.3.
[0329] A second batch of His-tagged quadruple mutant IL-2 was purified as described above by immobilized metal ion affinity chromatography (IMAC) and followed by size exclusion chromatography-(SEC). The buffers used for IMAC were 50 mM Tris, 20 mM imidazole, 0.5M NaCl pH 8 for column equilibration and washing, and 50 mM Tris, 0.5 M imidazole, 0.5 M NaCl pH 8 for elution. The buffer used for SEC and final formulation buffer was 20 mM histidine, 140 mM NaCl pH 6.
[0330] Subsequently, affinity for the IL-2R heterodimer and the IL-2R -subunit were determined by surface plasmon resonance (SPR). Briefly, the ligandeither human IL-2R -subunit (Fc2) or human IL2-R knob hole heterodimer (Fc3)was immobilized on a CM5 chip. Subsequently, naked wild-type (pool 1 and 2) or quadruple mutant IL-2, and Proleukin (Novartis/Chiron) were applied to the chip as analytes at 25 C. in HBS-EP buffer in concentrations ranging from 300 nM down to 1.2 nM (1:3 dil.). Flow rate was 30 l/min and the following conditions were applied for association: 180s, dissociation: 300 s, and regeneration: 230 s 3M MgCl.sub.2 for IL2-R knob hole heterodimer, 10 s 50 mM NaOH for IL-2R -subunit. 1:1 binding was applied for fitting (1:1 binding RI0, Rmax=local for IL-2R , apparent K.sub.D, 1:1 binding RI=0, Rmax=local for IL-2R ). Table 3 shows the respective K.sub.D values for binding of human wild-type and quadruple mutant IL-2 as well as of Proleukin to IL-2R and IL-2R -subunit.
TABLE-US-00003 TABLE 3 Affinity of mutant IL-2 polypeptides to the intermediate affinity IL-2R and the IL-2R -subunit. K.sub.D in nM Hu IL-2R Hu IL-2R Hu IL-2R T = 25 C. (kinetic) (kinetic) (steady state) Naked IL-2 wt, pool 1 5.6 17.4 30.3 5 16.6 23.9 Naked IL-2 wt, pool 2 2.8 10.6 19.7 1.8 10 17.6 Naked IL-2 qm 2.7 no binding no binding 2 Proleukin 2.4 7.5 19 2.8 12.5 17.8
[0331] The data show that the naked IL-2 quadruple mutant shows the desired behaviour and has lost binding for the IL-2R -subunit whereas binding to IL-2R is retained and comparable to the respective wild type IL-2 construct and Proleukin. Differences between pools 1 and 2 of the wild-type IL-2 can probably be attributed to differences in O-glycosylation. This variability and heterogeneity has been overcome in the IL-2 quadruple mutant by introduction of the T23A mutation.
Example 3
[0332] The three mutations F42A, Y45A and L72G and the mutation T3A were introduced in the Fab-IL-2-Fab format (
TABLE-US-00004 TABLE 4 Affinity of FAP-targeted immunoconjugates comprising mutant IL-2 polypeptides to the intermediate affinity IL-2R and the IL-2R -subunit (K.sub.D). Construct Hu Fab-IL-2-Fab Hu IL-2R IL-2R Mu IL-2R Mu IL-2R 4G8 IL-2 wt 3.8 nM 4.5 nM 45.6 nM 29 nM 4G8 IL-2 T3A 1.6 nM 4.9 nM 15.6 nM 15 nM 4G8 IL-2 F42A 4.7 nM 149 nM 57 nM 363 nM 4G8 IL-2 Y45A 3.9 nM 22.5 nM 41.8 nM 369 nM 4G8 IL-2 L72G ND 45.3 nM ND ND 4G8 IL-2 triple 5.6 nM no binding 68.8 nM ND mutant F42A/ Y45A/L72G 4G8 IL-2 5.2 nM no binding 56.2 nM no binding quadruple mutant T3A/F42A/ Y45A/L72G
[0333] Simultaneous binding to the IL 2R heterodimer and FAP was shown by SPR. Briefly, the human IL 2R knob-into-hole construct was immobilized on a CM5 chip chemically and 10 nM Fab-IL-2-Fib constructs were captured for 90 s. Human FAP served as analyte at concentrations of 200 nM down to 0.2 nM. Conditions were: temperature: 25 C., buffer: HBS-EP, flow: 30 l/min, association: 90 s, dissociation: 120 s. Regeneration was done for 60 s with 10 mM glycine pH 2. Fitting was performed with a model for 1:1 binding, RI0, Rmax-global. The SPR bridging assay showed that the Fab-IL-2-Fab constructs, both as wild-type and as quadruple mutant, as well as based on the affinity matured FAP binder 28H1 or the parental 3F2 or 4G8 antibodies, was-able to bind at a concentration of 10 nM simultaneously to the IL 2R heterodimer immobilized on the chip as well as to human FAP used as analyte (
TABLE-US-00005 TABLE 5 Affinity of FAP-targeted immunoconjugates, comprising mutant IL-2 polypeptides and bound to the intermediate affinity IL-2R, to FAP (K.sub.D). Construct Fab-IL-2-Fab K.sub.D 4G8 Fab-IL-2 wt-Fab 5.0 nM 4G8 Fab-IL-2 qm-Fab 5.6 nM 29B11 Fab-IL-2 wt-Fab 0.32 nM 29B11 Fab-IL-2 qm-Fab 0.89 nM 3F2 Fab-IL-2 wt-Fab 1.2 nM
[0334] Taken together the SPR data showed that i) the T3A mutation does not influence binding to CD25, ii) the three mutations F42A, Y45A and L72G do not influence the affinity for the IL 2R heterodimer while they reduce the affinity for CD25 in this order: wt=T3A>Y45A (ca. 5 lower)>L72G (ca. 10 lower)>F42A (ca. 33 lower); iii) the combination of the three mutations F42A, Y45A and L72G with or without the O-glycosylation site mutant T3A results in a complete loss of CD25 binding as determined under SPR conditions, iv) although affinity of human IL-2 for murine IL-2R heterodimer and IL-2R -subunit is reduced approximately by a factor of 10 compared to human IL-2 receptors the selected mutations do not influence affinity for the murine IL-2R heterodimer, but abolish binding to murine IL-2R -subunit accordingly. This indicates that the mouse represents a valid model for the study of pharmacological and toxicological effects of IL-2 mutants, although overall IL-2 exhibits less toxicity in rodents than in humans.
[0335] Apart from the loss of O-glycosylation one additional advantage of the combination of the four mutations T3A, F42A, Y45A, L72G is a lower surface hydrophobicity of the IL-2 quadruple mutant due to the exchange of surface exposed hydrophobic residues such as phenylalanine, tyrosine or leucine by alanine. An analysis of the aggregation temperature by dynamic light scattering showed that the aggregation temperature for the FAP-targeted Fab-IL-2-Fab immunoconjugates comprising wild-type or quadruple mutant IL-2 were in the same range: ca. 57-58 C. for the 3F2 parental Fab-IL-2-Fab and for the affinity matured 29B11 3F2-derivative; and in the range of 62-63 C. for the 4G8 parental Fab-IL-2-Fab and the affinity matured 28H1, 4B9 and 14B3 4G8-derivatives, indicating that the combination of the four mutations had no negative impact on protein stability. In support of the favorable properties of the selected IL-2 quadruple mutant, transient expression yields indicated that the quadruple mutant in the Fab-IL-2 qm-Fab format may even result in higher expression yields than those observed for the respective Fab-IL-2 wt-Fab constructs. Finally, pharmacokinetic analysis shows that both 4G8-based Fab-IL-2 qm-Fab and Fab-IL-2 wt-Fab have comparable PK properties (see example 9 below). Based on these data and the cellular data described in example 4 below the quadruple mutant T3A, F42A, Y45A, L72G was selected as ideal combination of mutations to abolish CD25 binding of IL-2 in the targeted Fab-IL-2-Fab immunoconjugate.
Example 4
[0336] The 4G8-based FAP-targeted Fab-IL 2-Fab immunoconjugates, comprising wild-type IL-2 or the single mutants 4G8 IL-2 T3A, 4G8 IL-2 F42A, 4G8 IL-2 Y45A, 4G8 IL-2 L72G or the respective triple (F42A/Y45A/L72G) or quadruple mutant (T3A/F42A/Y45A/L72G) IL-2, were subsequently tested in cellular assays in comparison to Proleukin as described above.
[0337] IL-2 induced IFN- release was measured following incubation of the NK cell line NK92 with the constructs (
TABLE-US-00006 TABLE 6 Induction of IFN- release from NK cells by Fab-IL-2-Fab immunoconjugates comprising mutant IL-2 polypeptides. Construct EC.sub.50 [nM] Proleukin 4.1 4G8 Fab-IL 2 wt-Fab 23.0 4G8Fab-IL-2 (T3A)-Fab 16.2 4G8 Fab-IL-2 (F42A)-Fab 15.4 4G8 Fab-IL-2(Y45A)-Fab 20.9 4G8 Fab-IL-2 (L72G)-Fab 16.3 4G8 Fab-IL-2 (triple mutant 42/45/72)-Fab 24.4
[0338] Subsequently, induction of proliferation of isolated human NK cells by Fab-IL-2-Fab immunoconjugates was assessed in a proliferation assay (Cell Titer Glo, Promega) (
Example 5
[0339] Based on the data described in examples 2 and 3 affinity matured FAP-targeted Fab-IL-2 qm-Fab immunoconjugates based on clones 28H1 or 29B11 were generated and purified as described above in the general methods section. In more detail, the FAP-targeted 28H1 targeted Fab-IL-2 qm-Fab was purified by one affinity step (protein G) followed by size exclusion chromatography (Superdex 200). Column equilibration was performed in PBS and supernatant from a stable CHO pool (CDCHO medium) was loaded onto a protein G column (GE Healthcare), the column was washed with PBS and samples were subsequently eluted with 2.5 mM HCl and fractions were immediately neutralized with 10PBS. Size exclusion chromatography was performed in the final formulation buffer 25 mM sodium phosphate, 125 mM sodium chloride, 100 mM glycine pH 6.7 on a Superdex 200 column.
[0340] FAP-targeted 408, 3F2 and 29B11 Fab-IL-2 qm-Fab and MCSP-targeted MHLG1 KV9 Fab-IL-2 qm-Fab immunoconjugates were purified by one affinity step (protein A) followed by size exclusion chromatography (Superdex 200). Column equilibration was performed in 20 mM sodium phosphate, 20 mM sodium citrate pH 7.5 and supernatant was loaded onto the protein A column. A first wash was performed in 20 mM sodium phosphate, 20 mM sodium citrate, pH 7.5 followed by a second wash: 13.3 mM sodium phosphate, 20 mM sodium citrate, 500 mM sodium chloride, pH 5.45. The Fab-IL-2 qm-Fab immunoconjugates were eluted in 20 mM sodium citrate, 100 mM sodium chloride, 100 mM glycine pH 3. Size exclusion chromatography was performed in the final formulation buffer: 25 mM potassium phosphate, 125 mM sodium chloride, 100 mM glycine pH 6.7.
[0341] FAP-targeted IgG-IL-2 qm fusion proteins based on the FAP-antibodies 4G8, 4B9 and 28H1, and a control DP47GS non-targeted IgG-IL-2 qm fusion protein were generated as described above in the general methods section.
[0342] The FAP binding activity of the IgG-IL-2 qm immunoconjugates based on 4G8 and 28H1 anti-FAP antibodies were determined by surface plasmon resonance (SPR) on a Biacore machine in comparison to the corresponding unmodified IgG antibodies. Briefly, an anti-His antibody (Penta-His, Qiagen 34660) was immobilized on CM5 chips to capture 10 nM His-tagged human FAP (20 s). Temperature was 25 C. and HBS-EP was used as buffer. Analyte concentration was 50 nM down to 0.05 nM at a flow rate of 50 l/min (association: 300 s, dissociation: 900 s, regeneration: 60 s, with 10 mM glycine pH 2). Fitting was performed based on a 1:1 binding model, RI=0, Rmax=local (because of capture format). Table 7 gives the estimated apparent bivalent affinities (pM avidity) as determined by SPR fitted with 1:1 binding RI=0, Rmax-local.
TABLE-US-00007 TABLE 7 K.sub.D [pM] Hu FAP 4G8 IgG-IL-2 qm 100 4G8 IgG 50 28H1 IgG-IL-2 qm 175 28H1 IgG 200
[0343] The data show that within the error of the method affinity for human FAP is retained for the 28H1-based immunoconjugate or only slightly decreased for the 4G8-based immunoconjugate as compared to the corresponding unmodified antibodies.
Example 6
[0344] The affinity of the FAP-targeted, affinity matured 28H1 and 29B11-based Fab-IL-2-Fab immunoconjugates, each comprising wild-type or quadruple mutant IL-2, and of the 3F2-based Fab-IL-2 wt-Fab were determined by surface plasmon resonance (SPR) for the human, murine and cynomolgus IL-2R heterodimer using recombinant IL-2R heterodimer under the following conditions: ligand: human, murine and cynomolgus IL-2R knob hole heterodimer immobilized on CM5 chip, analyte: 28H1 or 29B11 Fab-IL-2-Fab (comprising wild-type or quadruple mutant IL-2), 3F2 Fab-IL-2-Fab (comprising wild-type IL-2), temperature: 25 C. or 37 C., buffer HBS-EP, analyte concentration: 200 nM down to 2.5 nM, flow: 30 l/min, association: 300 s, dissociation: 300 s, regeneration: 60 s 3M MgCl.sub.2, fitting: 1:1 binding. RI0, Rmax=global. The affinity of the FAP-targeted affinity matured 28H1 and 29B11-based Fab-IL-2-Fab immunoconjugate, each containing wildtype or quadruple mutant IL-2, and of the 3F2-based Fab-IL-2 wt-Fab were determined by surface plasmon resonance (SPR) for the human, murine and cynomolgus IL-2R -subunit using recombinant monomeric IL-2R -subunit under the following conditions: ligand: human, murine and cynomolgus IL-2R -subunit immobilized on a CM5 chip, analyte: 28H1 or 29B11 Fab-IL-2-Fab (comprising wild-type or mutant IL-2), 3F2 Fab-IL-2-Fab (comprising wild-type IL-2), temperature: 25 C. or 37 C., buffer: HBS-EP, analyte concentration 25 nM down to 0.3 nM, flow: 30 l/min, association: 120 s, dissociation: 600 s, regeneration: none, fitting: 1:1 binding, RI=0, Rmax=global.
[0345] Results of the kinetic analysis with the IL-2R heterodimer are given in Table 8.
TABLE-US-00008 TABLE 8 Binding of Fab-IL-2-Fab immunoconjugates comprising affinity matured Fab and mutant IL-2 to IL-2R heterodimers. Hu IL-2R Hu IL-2R Cyno IL-2R Cyno IL-2R Mu IL-2R Mu IL-2R K.sub.D in nM (25 C.) (37 C.) (25 C.) (37 C.) (25 C.) (37 C.) 28H1 Fab-IL- 9.7 19 11.5 29.2 112 186 2 wt-Fab 9 22 11.6 30.4 79 219 28H1 Fab-IL- 7.5 14.3 8.9 21.3 66 142 2 qm-Fab 6.9 14.7 8.4 21.2 54 106 29B11 Fab- 6.5 9.5 6.9 14 93 71 IL-2 wt-Fab 5.7 12.4 6.7 19 74 74 29B11 Fab-IL- 12 13.1 7.8 16.7 60 44 2 qm-Fab 7.4 13 8.4 18.1 63 42 3F2 Fab-IL- 5 ND 6.4 ND 40 ND 2 wt-Fab 4.8 6.1 40
[0346] Whereas the affinity of human IL-2 to the human IL 2R heterodimer is described to be around 1 nM, the Fab-IL-2-Fab immunoconjugates (comprising wild-type or quadruple mutant IL-2) both have a reduced affinity between 6 and 10 nM, and as shown for the naked IL-2 above the affinity to the murine IL-2R is around 10 times weaker than for the human and cynomolgus IL-2R.
[0347] Results of the kinetic analysis with the IL-2R -subunit are given in Table 9. Under the chosen conditions there is no binding detectable of the immunoconjugates comprising the IL-2 quadruple mutant to the human, murine or cyno IL-2R -subunit.
TABLE-US-00009 TABLE 9 Binding of Fab-IL-2-Fab immunoconjugates comprising affinity matured Fab and mutant IL-2 to IL-2R -subunits. Hu IL-2R Hu IL-2R Cyno IL-2R Cyno IL-2R Mu IL-2R Mu IL-2R K.sub.D in nM (25 C.) (37 C.) (25 C.) (37 C.) (25 C.) (37 C.) 28H1 Fab-IL- 16 28.8 16 36.5 43.3 67.5 2 wt-Fab 16.2 28.2 16.2 35.6 44 61.1 28H1 Fab-IL- no no no no no no 2 qm-Fab binding binding binding binding binding binding 29B11 Fab- 5 7.6 4.8 7.3 11.4 13.3 IL-2 wt-Fab 4.6 7.7 4.3 7.4 9.6 13.8 29B11 Fab- no no no no no no IL-2 qm-Fab binding binding binding binding binding binding 3F2 Fab-IL- 5.7 ND 5 ND 12.3 ND 2 wt-Fab 6.1 5.4 12.1
[0348] The affinity of the MCSP-targeted MHLG1-KV9 Fab-IL-2-Fab immunoconjugates, comprising the wild-type or quadruple mutant IL-2, were determined by surface plasmon resonance (SPR) for the human IL-2R heterodimer using recombinant IL-2R heterodimer under the following conditions: human IL-2R knob hole heterodimer was immobilized on a CM5 chip (1600 RU). MHLG1-KV9 Fab-IL-2 wt-Fab and Fab-IL-2 qm-Fab were used as analyte at 25 C. in HBS-P buffer. Analyte concentration was 300 nM down to 0.4 nM (1:3 dil.) for IL-2R at a flow of 30 p/min (association time 180 s, dissociation time 300 s). Regeneration was done for 230 s with 3M MgCl.sub.2 for IL-2R . Data were fitted using a 1:1 binding, RI0, Rmax=local for IL-2R .
[0349] The affinity of the MCSP-targeted MHLG1-KV9 Fab-IL-2-Fab immunoconjugates, comprising the wild-type or quadruple mutant IL-2, were determined by surface plasmon resonance (SPR) for the human IL-2R -subunit using recombinant monomeric IL-2R -subunit under the following conditions: human IL-2R -subunit was immobilized on a CM5 chip (190 RU). MHLG1-KV9 Fab-IL-2 wt-Fab and Fab-IL-2 qm-Fab were used as analyte at 25 C. in HBS-P buffer. Analyte concentration was 33.3 nM down to 0.4 nM (1:3 dil.) for IL-2R at a flow of 30 l/min (association time 180 s, dissociation time 300 s). Regeneration was done for 10 s with 50 mM NaOH for IL-2R . Data were fitted using a 1:1 binding, RI=0, Rmax-global for IL-2R .
[0350] Results of the kinetic analysis with the IL-2R heterodimer are given in Table 10.
TABLE-US-00010 TABLE 10 K.sub.D in nM Hu IL 2R Hu IL 2R Hu IL 2R T = 25 C. (kinetic) (kinetic) (steady state) MHLG1-KV9 Fab-IL-2 8.6 8.8 6.8 wt-Fab 9.8 10.1 10.9 MHLG1-KV9 Fab-IL 2 7.3 No binding No binding qm-Fab 10.7
[0351] The data confirm that the MCSP-targeted MHLG1-KV9 Fab-IL-2 qm-Fab immunoconjugate has retained affinity for the IL-2R receptor, whereas binding affinity to CD25 is abolished compared to the immunoconjugate comprising wild-type IL-2.
[0352] Subsequently, the affinity of the 4G8- and 28H1-based IgG-IL-2 qm immunoconjugates to the IL-2R heterodimer and the IL-2R -subunit were determined by surface plasmon resonance (SPR) in direct comparison to the Fab-IL-2 qm-Fab immunoconjugate format. Briefly, the ligandseither the human IL-2R -subunit or the human IL-2R heterodimerwere immobilized on a CM5 chip. Subsequently, the 4G8- and 28H1-based IgG-IL-2 qm immunoconjugates or the 4G8- and 28H1-based Fab-IL-2 qm-Fab immunoconjugates were applied to the chip as analytes at 25 C. in HBS-EP buffer in concentrations ranging from 300 nM down to 1.2 nM (1:3 dil.). Flow rate was 30 l/min and the following conditions were applied for association: 180s, dissociation: 300 s, and regeneration: 230 s with 3 M MgCl.sub.2 for IL-2R heterodimer, 10 s with 50 mM NaOH for IL-2R -subunit. 1:1 binding was applied for fitting (1:1 binding RI0, Rmax=local for IL-2R , apparent K.sub.D, 1:1 binding RI=0, Rmax-local for IL-2R ). The respective K.sub.D values are given in Table 11.
TABLE-US-00011 TABLE 11 Apparent K.sub.D [nM] Hu IL-2R Hu IL-2R 4G8 IgG-IL-2 qm 5.9 No binding 4G8 Fab-IL-2 qm-Fab 10.4 No binding 28H1 IgG-IL-2 qm 6.2 No binding 28H1 Fab-IL-2 qm-Fab 11.4 No binding
[0353] The data show that the 4G8- and 28H1-based IgG-IL-2 qm immunoconjugates bind with at least as good affinity as the Fab-IL-2 qm-Fab immunoconjugates to the IL-2 heterodimer, whereas they do not bind to the IL-2R -subunit due to the introduction of the mutations interfering with CD25 binding. Compared to the corresponding Fab-IL-2 qm-Fab immunoconjugates the affinity of the IgG-IL-2 qm fusion proteins appears to be slightly enhanced within the error of the method.
Example 7
[0354] In a first set of experiments we confirmed that the FAP-targeted Fab-IL-2-Fab immunoconjugates comprising either wild-type or mutant IL-2 were able to bind to human FAP-expressing HEK.293-FAP cells by FACS (
TABLE-US-00012 TABLE 12 Binding of Fab-IL-2-Fab immunoconjugates to FAP-expressing HEK cells. EC.sub.50 values nM 28H1 Fab-IL-2-Fab 0.64 28H1 Fab-IL-2 qm-Fab 0.70 29B11 Fab-IL-2-Fab 0.66 29B11 Fab-IL-2 qm-Fab 0.85 4G8 Fab-IL-2-Fab 0.65
[0355] In particular, these binding experiments showed that the affinity matured FAP binders 28H1, 29B11, 14B3 and 4B9 as Fab-IL-2 qm-Fab showed superior absolute binding to the HEK 293-FAP target cells compared to the Fab-IL-2-Fab immunoconjugates based on the parental FAP binders 3F2 (29B11, 14B3, 4B9) and 4G8 (28H1) (
[0356] In a further experiment, binding of FAP-targeted 4G8-based IgG-IL-2 qm and Fab-IL-2 qm-Fab immunoconjugates to human FAP expressed on stably transfected HEK293 cells was measured by FACS. The results are shown in
[0357] The affinity matured anti-FAP Fab-IL-2-Fab immunoconjugates comprising wildtype IL-2 or the quadruple mutant were subsequently tested in cellular assays in comparison to Proleukin as described in the examples above.
[0358] IL-2 induced IFN- release was measured in the supernatant by ELISA following incubation of the NK-cell line NK92 with these immunoconjugates (
[0359] In addition, the MCSP-targeted MHLG1-KV9-based Fab-IL-2 qm-Fab immunoconjugate was compared to the 28H1 and 29B11 based Fab-IL-2 qm-Fab immunoconjugates in the IFN- release assay on NK92 cells.
[0360] Subsequently, induction of proliferation of NK92 cells by IL-2 over a period of 3 days was assessed in a proliferation assay by ATP measurement using CellTiter Glo (Promega) (
[0361] In a further experiment we studied the effects of the 28H1 affinity matured FAP-directed Fab-IL-2 qm-Fab immunoconjugate on induction of STAT5 phosphorylation compared to 28H1 Fab-IL-2 wt-Fab and Proleukin on human NK cells, CD4.sup.+ T cells, CD8.sup.+ T cells and T.sub.reg cells from human PBMCs. (
TABLE-US-00013 TABLE 13 Induction of IFN- release from NK cells by 28H1 FAP-targeted Fab-IL-2-Fab immunoconjugates comprising mutant IL-2 polypeptides. NK CD8.sup.+ CD4.sup.+ T.sub.reg EC.sub.50 [pM] cells T cells T cells cells Proleukin 222 1071 92 1 28H1 Fab-IL-2 wt-Fab 3319 14458 3626 15 28H1 Fab-IL-2 qm-Fab 3474 20583 70712 19719
[0362] In another set of experiments, the biological activity of FAP-targeted 4G8-based IgG-IL-2 qm and Fab-IL-2 qm-Fab immunoconjugates was investigated in several cellular assays.
[0363] FAP-targeted 4G8-based IgG-IL-2 qm and 28H1-based Fab-IL-2 qm-Fab immunoconjugates were studied for the induction of IFN- release by NK92 cells as induced by activation of IL-2R signaling.
[0364] We also studied the effects of the FAP-targeted 4G8-based IgG-IL-2 qm immunoconjugate on the induction of STAT5 phosphorylation compared to the 28H1 based Fab-IL-2 wt-Fab and Fab-IL-2 qm-Fab immunoconjugates as well as Proleukin on human NK cells, CD4.sup.+ T cells, CD8.sup.+ T cells and T.sub.reg cells from human PBMCs. The results of these experiments are shown in
[0365] Taken together the IL-2 quadruple mutant described here is able to activate IL-2R signaling through the IL-2R heterodimer similar to wild-type IL-2, but does not result in a preferential stimulation of T.sub.reg cells over other effector cells.
Example 8
[0366] The anti-tumoral effects of FAP-targeted Fab-IL-2 qm-Fab immunoconjugates were evaluated in vivo in comparison to FAP-targeted Fab-IL-2 wt-Fab immunoconjugates in ACHN xenograft and LLC1 syngeneic models. All FAP-targeted Fab-IL-2-Fab immunoconjugates (comprising wild-type or quadruple mutant IL-2) recognize murine FAP as well as the murine IL-2R. While the ACHN xenograft model in SCID-human FcRIII transgenic mice is strongly positive for FAP in IHC, it is an immunocompromised model and can only reflect immune effector mechanisms mediated by NK cells and/or macrophages/monocytes, but lacks T cell mediated immunity and thus cannot reflect AICD or effects mediated through T.sub.reg cells. The syngeneic LLC1 model in contrast in fully immunocompetent mice can reflect adaptive T cell mediated immune effector mechanisms as well, but shows fairly low expression of FAP in the murine stroma. Each of these models thus partially reflects the situation as encountered in human tumors.
[0367] ACHN Renal Cell Carcinoma Xenograft Model
[0368] The FAP-targeted 4G8 Fab-IL-2 wt-Fab and 4G8 Fab-IL-2 qm-Fab immunoconjugates were tested using the human renal cell adenocarcinoma cell line ACHN, intra-renally injected into SCID-human FcRIII transgenic mice. ACHN cells were originally obtained from ATCC (American Type Culture Collection) and after expansion deposited in the Glycart internal cell bank. ACHN cells were cultured in DMEM containing 10% FCS, at 37 C. in a water-saturated atmosphere at 5% CO.sub.2. In vitro passage 18 was used for intrarenal injection, at a viability of 98.4%. A small incision (2 cm) was made at the right flank and peritoneal wall of anesthetized SCID mice. Fifty l cell suspension (110.sup.6 ACHN cells in AimV medium) was injected 2 mm subcapsularly in the kidney. Skin wounds and peritoneal wall were closed using clamps. Female SCID-FcRIII mice (GLYCART-RCC), aged 8-9 weeks at the beginning of the experiment (bred at RCC, Switzerland) were maintained under specific-pathogen-free conditions with daily cycles of 12 h light/12 h darkness according to committed guidelines (GV-Solas; Felasa; TierschG). The experimental study protocol was reviewed and approved by local government (P 2008016). After arrival, animals were maintained for one week to get accustomed to new environment and for observation. Continuous health monitoring was carried out on a regular basis. Mice were injected intrarenally on study day 0 with 110.sup.6 ACHN cells, randomized and weighed. One week after the tumor cell injection, mice were injected i.v. with 4G8 Fab-IL-2 wt-Fab and 4G8 Fab-IL-2 qm-Fab three times a week for three weeks. All mice were injected i.v. with 200 l of the appropriate solution. The mice in the vehicle group were injected with PBS and the treatment groups with 4G8 Fab-IL-2 wt-Fab or 4G8 Fab-IL-2 qm-Fab immunoconjugate. To obtain the proper amount of immunoconjugate per 200 l, the stock solutions were diluted with PBS when necessary.
TABLE-US-00014 TABLE 14-A Concen- tration Compound Dose Formulation buffer (mg/mL) 4G8 Fab-IL-2- 20 g 25 mM potassium phosphate, 1.45 Fab wild type = 125 mM NaCl, FAP 4G8 wt 100 mM glycine, pH 6.7 4G8 Fab-IL-2- 20 g 25 mM potassium phosphate, 4.25 Fab quadruple 125 mM NaCl, mutant = FAP 100 mM glycine, pH 6.7 4G8 qm
[0369] LLC1 Lewis Lung Carcinoma Syngeneic Model
[0370] The FAP-targeted 4G8 Fab-IL-2 qm-Fab and 28H1 Fab-IL-2 qm-Fab immunoconjugates were tested using the mouse Lewis lung carcinoma cell line LLC1, i.v. injected into Black 6 mice. The LLC1 Lewis lung carcinoma cells were originally obtained from ATCC and after expansion deposited in the Glycart internal cell bank. The tumor cell line was routinely cultured in DMEM containing 10% FCS (Gibco) at 37 C. in a water-saturated atmosphere at 5% CO.sub.2. Passage 10 was used for transplantation, at a viability of 97.9%. 210.sup.5 cells per animal were injected i.v. into the tail vein in 200 l of Aim V cell culture medium (Gibco). Black 6 mice (Charles River, Germany), aged 8-9 weeks at the start of the experiment, were maintained under specific-pathogen-free conditions with daily cycles of 12 h light/12 h darkness according to committed guidelines (GV-Solas; Felasa; TierschG). The experimental study protocol was reviewed and approved by local government (P 2008016). After arrival, animals were maintained for one week to get accustomed to the new environment and for observation. Continuous health monitoring was carried out on a regular basis. Mice were injected i.v. on study day 0 with 210.sup.5 of LLC1 cells, randomized and weighed. One week after the tumor cell injection, mice were injected i.v. with 4G8 Fab-IL-2qm-Fab or 28H1 Fab-IL-2 qm-Fab, three times a week for three weeks. All mice were injected i.v. with 200 l of the appropriate solution. The mice in the vehicle group were injected with PBS and the treatment group with the 4G8 Fab-IL-2 qm-Fab or 28H1 Fab-IL-2 qm-Fab constructs. To obtain the proper amount of immunoconjugate per 200 l, the stock solutions were diluted with PBS when necessary.
TABLE-US-00015 TABLE 14-B Concen- tration Compound Dose Formulation buffer (mg/mL) 28H1 Fab-IL-2- 30 g 25 mM potassium phosphate, 2.74 Fab quadruple 125 mM NaCl, mutant = FAP 100 mM glycine, pH 6.7 28H1 qm 4G8 Fab-IL-2- 30 g 25 mM potassium phosphate, 4.25 Fab quadruple 125 mM NaCl, mutant = FAP 100 mM glycine, pH 6.7 4G8 qm
[0371] In another experiment, the FAP-targeted 28H1 Fab-IL-2 wt-Fab and 28H1 Fab-IL-2 qm-Fab immunoconjugates were tested in the same mouse Lewis lung carcinoma cell line LLC1, i.v. injected into Black 6 mice. Passage 9 was used for transplantation, at a viability of 94.5%. 210.sup.5 cells per animal were injected i.v. into the tail vein in 200 l of Aim V cell culture medium (Gibco). Mice were injected i.v. on study day 0 with 210.sup.5 of LLC1 cells, randomized and weighed. One week after the tumor cell injection, mice were injected i.v. with 28H1 Fab-IL-2 wt-Fab or 28H1 Fab-IL-2 qm-Fab, three times a week for three weeks. All mice were injected i.v. with 200 l of the appropriate solution. The mice in the vehicle group were injected with PBS and the treatment group with the 28H1 Fab-IL-2 wt-Fab or 28H1 Fab-IL-2 qm-Fab constructs. To obtain the proper amount of immunoconjugate per 200 l, the stock solutions were diluted with PBS when necessary.
TABLE-US-00016 TABLE 14-C Concen- tration Compound Dose Formulation buffer (mg/mL) 28H1 Fab-IL-2- 45 g 25 mM potassium phosphate, 2.74 Fab quadruple 125 mM NaCl, mutant = FAP 100 mM glycine, pH 6.7 28H1 qm 28H1 Fab-IL-2- 45 g 25 mM potassium phosphate, 1.66 Fab wild-type = 125 mM NaCl, FAP 28H1 wt 100 mM glycine, pH 6.7
Example 9
[0372] The 4G8 based FAP-targeted Fab-IL-2 qm-Fab was subsequently compared to the 4G8 based FAP-targeted Fab-IL-2 wt-Fab immunoconjugate in a seven-day intravenous toxicity and toxicokinetic study in Black 6 mice. Table 15 shows the study design of the toxicity and toxicokinetic studies.
TABLE-US-00017 TABLE 15 Study design. Dose Group Type [g/g] Purpose 1 DPBS 0 Control 2 4G8 Fab-IL-2 wt-Fab 4.5 Toxicity titration 3 9.0 4 4G8 Fab-IL-2 qm-Fab 4.5 5 9.0 6 4G8 Fab-IL-2 wt-Fab 4.5 Toxicokinetic 7 9.0 study 8 4G8 Fab-IL-2 qm-Fab 4.5 9 9.0
[0373] The purpose of this study was to characterize and compare the toxicity and toxicokinetic profiles of FAP-targeted 4G8 Fab-IL2-Fab wild type (wt) interleukin-2 (IL-2) and FAP-targeted G48 Fab-IL-2-Fab quadruple mutant IL-2 (qm) after once daily intravenous administration to non-tumor-bearing male mice for 7 days. For this study, 5 groups of 5 male mice/group were administered intravenously 0 (vehicle control), 4.5 or 9 g/g/day wt IL-2, or 4.5 or 9 g/g/day qm IL-2. An additional 4 groups of 6 male mice/group were administered 4.5 or 9 g/g/day wt IL-2, or 4.5 or 9 g/day qm IL-2 in order to assess toxicokinetics. The study duration was changed from 7 days to 5 days due to clinical signs observed in animals given 4.5 and 9 g/g/day wt IL-2. Assessment of toxicity was based upon mortality, in-life observations, body weight, and clinical and anatomic pathology. Blood was collected at various time points from animals in the toxicokinetic groups for toxicokinetic analysis. The toxicokinetic data showed that the mice treated with wt IL-2 or qm IL-2 had measurable plasma levels up to the last bleeding time, indicating that the mice were exposed to the respective compounds throughout the duration of treatment. Day 1 AUC0-inf values suggest comparable exposure of wt IL-2 and qm IL-2 at both dose levels. Sparse samples were taken on Day 5 and showed equivalent plasma concentrations to Day 1, suggesting no accumulation occurred after 5 days of dosing either compound. In more details the following findings were observed.
[0374] Toxicokinetics
[0375] Table 16 summarizes the mean plasma toxicokinetic parameters for the FAP-targeted 4G8 Fab-IL-2 qm-Fab and the FAP-targeted 4G8 Fab-IL-2 wt-Fab as determined by WinNonLin Version 5.2.1 and a commercial kappa-specific ELISA (Human Kappa ELISA Quantitation Set, Bethyl Laboratories).
TABLE-US-00018 TABLE 16 Group 6 Group 7 Group 8 Group 9 4G8-FAP-Wild 4G8-FAP-Wild 4G8-FAP- 4G8-FAP- Parameter Units Type IL-2 Type IL-2 Mutant IL-2 Mutant IL-2 Cmax ng/ml 47198 97986 60639 146415 Cmax/Dose (ng/ml)/(ug/g) 0.011 0.011 0.0135 0.016 AUC ng*h/ml 331747 747449 355030 926683 AUC/Dose (ng*h/ml)/(ug/g) 0.074 0.083 0.079 0.103 Tz h 3.6 3.11 4.3 3.12 Original Dose ug/g 4.5 9 4.5 9 Route IV IV IV IV *TK Parameters were calculated in WinNonlin Version 5.2.1 using noncompartmental analysis
[0376] The individual serum concentrations are given in the following:
TABLE-US-00019 Serum Mean Bleed Time conc. conc Group (dose) Day (h) Animal (ng/ml) (ng/ml) Group 6 1 1 26 64241 47198 (4.5 g/g) 27 30155 4G8 Fab-II2- 1 5.5 28 14693 15784 Fab WT 29 16875 1 24 30 318 419 31 520 5 5.5 29 13061 13335 30 13620 31 13325 Group 7 1 1 32 101208 97986 (9 g/g) 33 94764 4G8 Fab-II2- 1 5.5 34 35766 34062 Fab WT 35 32359 1 24 36 573 580 37 588 5 5.5 32 31779 37473 33 51143 35 53409 36 13562 Group 8 1 1 38 73326 60639 (4.5 g/g) 39 47953 4G8 Fab-II2- 1 5.5 40 12168 13269 Fab Mutant 41 14371 1 24 42 494 490 43 487 5 5.5 40 6561 10957 41 15352 5 24 38 608 721 39 543 42 1298 43 437 Group 9 1 1 44 162970 146416 (9 g/g) 45 129862 4G8 Fab-II2- 1 5.5 46 20475 24800 Fab Mutant 47 29125 1 24 48 478 493 49 509 5 5.5 48 20504 48031 47 75557 5 24 44 634 703 45 796 48 661 49 719
[0377] These data show that both, the 4G8 Fab-IL-2 qm-Fab and the 4G8 Fab-IL-2 wt-Fab show comparable pharmacokinetic properties with slightly higher exposure for the 4G8 Fab-IL-2 qm-Fab.
[0378] Mortality
[0379] In the 9 g/g FAP-targeted 4G8 Fab-IL-2 wt-Fab group, treatment-related mortality occurred in one animal prior to necropsy on Day 5. Hypoactivity, cold skin, and hunched posture were noted prior to death. This animal likely died due to a combination of cellular infiltration in the lung that was accompanied with edema and hemorrhage and marked bone marrow necrosis. Mortality is summarized in Table 17.
TABLE-US-00020 TABLE 17 Mortality day 5. Severe Dose Found toxicity Group Type [g/g] dead Sacrifice** Total 1 DPBS 0 0/5 0/5 0/5 2 4G8 Fab-IL-2 wt-Fab 4.5 0/5 5/5 5/5 3 9 1/5* 4/5 4/5 4 4G8 Fab-IL-2 qm-Fab 4.5 0/5 0/5 0/5 5 9 0/5 0/5 0/5 6 4G8 Fab-IL-2 wt-Fab 4.5 1/6 5/6 6/6 7 9 2/6 4/6 6/6 8 4G8 Fab-IL-2 qm-Fab 4.5 0/6 0/6 0/6 9 9 0/6 0/6 0/6 *in route to necropsy **study was planned for seven days but all mice treated with the wild-type IL-2 immunoconjugate were markedly affected by Day 5 and were sacrificed as they were not expected to survive.
[0380] Clinical Observations
[0381] Observations of hypoactivity, cold skin, and hunched posture were noted in animals given 4.5 and 9 g/g/day wt IL-2. Clinical observations are summarized in Table 18.
TABLE-US-00021 TABLE 18 Clinical observations day 5. Dose Hunched Hypo- Cool to Group Type [g/g] posture active touch 1 DPBS 0 0/5 0/5 0/5 2 4G8 Fab-IL-2 wt-Fab 4.5 4/5 4/5 5/5 3 9 5/5 5/5 5/5 4 4G8 Fab-IL-2 qm-Fab 4.5 0/5 0/5 0/5 5 9 0/5 0/5 0/5 6 4G8 Fab-IL-2 wt-Fab 4.5 6/6 2/6 2/6 7 9 6/6 5/6 6/6 8 4G8 Fab-IL-2 qm-Fab 4.5 0/6 0/6 0/6 9 9 0/6 0/6 0/6
[0382] Body Weight
[0383] A moderate decrease in body weight was observed after 5 days of treatment in animals given 4.5 and 9 (9% and 11%, respectively) g/g/day wt IL-2. A slight decrease in body weight was observed after 5 days of treatment in animals given 4.5 or 9 (2% and 1%, respectively) g/g/day qm IL-2. A moderate (9%) decrease in body weight was also observed in vehicle controls after 5 days of treatment. However, the percent decrease would have been 5% if a potential outlier (Animal #3) was excluded. The body weight loss in the vehicle group may have been attributed to stress.
[0384] Hematology
[0385] A reduced platelet count was observed in animals given 4.5 (4.5 fold) and 9 g/g/day (11 fold) 4G8 Fab-IL-2 wt-Fab, which correlated with reduced megakaryocytes in the bone marrow as well as systemic consumptive effects (fibrin) in spleen and lung of these animals (see Histopathology section below) These findings indicated that reduced platelets were likely due to combined effects of consumption and decrease in production/bone marrow crowding due to increase in lymphocyte/myeloid cell production as a direct or indirect effect of IG-2.
[0386] Hematologic findings of uncertain relationship to compound administration consisted of absolute lymphocyte count decreases with 4G8 Fab-IL-2 wt-Fab at 4.5 (5-fold) and 9 g/g (3-fold) compared to the mean value of the vehicle control group. These findings lacked clear dose-dependency, but could be considered secondary to effects associated with stress noted in in-life observations or exaggerated pharmacology of the compound (lymphocytes migrating into tissues). There were no treatment-related hematological changes attributed to the administration of 4G8-Fab-IL-2 qm-Fab. A few isolated hematologic findings were statistically different from their respective controls. However, these findings were of insufficient magnitude to suggest pathological relevance.
[0387] Gross Pathology and Histopathology
[0388] Treatment-related gross findings included enlarged spleen found in 5/5 and 4/5 mice of 4.5 and 9 g/g 4G8 Fab-IL-2 wt-Fab groups, respectively, and in 1/5 in both 4.5 and 9 g/g 4G8 Fab-IL-2 qm-Fab treatment groups.
[0389] Treatment-related histopathology findings were present in groups given 4.5 and 9 g/g 408 Fab-IL-2 wt-Fab and 4.5 and 9 g/g 4G8 Fab-IL-2 qm-Fab in lung, bone marrow, liver, spleen, and thymus, with differences in incidence, severity grading or nature of the changes, as reported below.
[0390] Treatment-related histopathology findings in the lung consisted of mononuclear infiltration found mild to marked in 5/5 mice of the 4.5 and 9 g/g 408 Fab-IL-2 wt-Fab groups and marginally in 5/5 mice of the 4.5 and 9 g/g 4G8 Fab-IL-2 qm-Fab groups. Mononuclear infiltration consisted of lymphocytes (some of which were noted as having cytoplasmic granules) as well as reactive macrophages. These cells were most often noted to have vasocentric patterns, often with margination noted within the vessels in the lung. These cells were also noted surrounding the vessels, but in more severe cases, the pattern was more diffuse. Hemorrhage was seen marginal to mild in 5/5 mice of the 4.5 and 9 g/g 4G8 Fab-IL-2 wt-Fab groups and marginally in 2/5 mice in the 9 g/g 4G8 Fab-IL-2.qm-Fab group. Though the hemorrhage was most often noted perivascularly, in more severe cases, it was noted in alveolar spaces. Edema was noted mild to moderate in 5/5 mice in the 4.5 and 9 g/g 4G8 Fab-IL-2 wt-Fab groups and marginally in 5/5 mice in the 9 g/g 4G8 Fab-IL-2 qm-Fab group. Though the edema was frequently seen perivascularly, in more severe cases, it was noted in alveolar spaces as well. Marginal cellular degeneration and karyorrhexis was noted in 2/5 and 5/5 mice in the 4.5 and 9 g/g 4G8 Fab-IL-2 wt-Fab groups, respectively and consisted of degeneration of infiltrative or reactive leukocytes. Selected animals with MSB stains were positive for fibrin found within the lungs of animals in both 4.5 and 9 g/g 4G8 Fab-IL-2 wt-Fab groups which correlates in part with the reduced platelets noted in these animals.
[0391] Treatment-related changes in the bone marrow included marginal to mild increased overall marrow cellularity in 5/5 mice and 2/5 mice of both 4.5 and 5/5 mice and 2/5 mice of both 9 g/g 4G8 Fab-IL-2 wt-Fab and 4G8 Fab-IL-2 qm-Fab groups, respectively. This was characterized by increased marginal to moderate lymphocyte-myelocyte hyperplasia in these groups that was supported, in part, by increased numbers of CD3 positive T cells within the marrow and sinuses (specifically T-lymphocytes, confirmed by immunohistochemistry with the pan-T-cell marker CD3 done on selected animals). CD3 positive T cell increase was moderate in both 4G8 Fab-IL-2 wt-Fab groups and marginal to mild in both 4G8 Fab-IL-2 qm-Fab groups Marginal to mild decreases in megakaryocytes were observed in 2/5 mice in the 4.5 and 5/5 mice in the 9 g/g 4G8 Fab-IL-2 wt-Fab groups and marginal to moderate decreases in erythroid precursors were noted in 3/5 mice in the 4.5 and 5/5 mice in the 9 g/g 4G8 Fab-IL-2 wt-Fab groups. Bone marrow necrosis was noted in 1/5 mice in 4.5 (minimal) and 5/5 mice in 9 (mild to marked) g/g 4G8 Fab-IL-2 wt-Fab groups. The reduced number of megakaryocytes in the bone marrow correlated with decreased platelets which could be due to direct crowding of the bone marrow by increased lymphocytes/myeloid precursors and/or the bone marrow necrosis, and/or consumption of platelets due to inflammation in various tissues (see spleen and lung). The decreased erythroid precursors noted in the bone marrow, did not correlate with the peripheral blood hematology findings likely due to temporal effects (seen in bone marrow before peripheral blood) and the longer half-life of peripheral erythrocytes (compared to platelets). The mechanism of bone marrow necrosis in the bone marrow may be secondary due to overt overcrowding of the marrow cavity (due to production and growth of lymphocytes/myeloid cells), systemic or local release of cytokines from the proliferating cell types, possibly related to local affects of hypoxia or other pharmacologic effects of the compound.
[0392] Treatment-related findings in the liver consisted of mild to moderate primarily vasocentric mononuclear cell infiltrate and marginal to mild single cell necrosis in 5/5 mice of the 4.5 and 9 g/g 4G8 Fab-IL-2 wt-Fab groups. Marginal single cell necrosis was seen in 2/5 and 4/5 mice in the 4.5 and 9 g/g 4G8 Fab-IL-2 qm-Fab groups, respectively. The mononuclear infiltrate consisted primarily of lymphocytes (specifically T-lymphocytes, confirmed by immunohistochemistry with the pan-T cell marker CD3 done on selected animals) that were most often noted vasocentrically as well as marginating within the central and portal vessels. Selected animals for immunohistochemistry staining for F4/80 showed increased numbers and size (activated) of macrophages/Kupffer cells throughout the hepatic sinusoids in 9 g/g 4G8 Fab-IL-2 wt-Fab and 4G8 Fab-IL-2 qm-Fab groups.
[0393] Treatment-related findings in the spleen consisted of moderate to marked lymphoid hyperplasia/infiltration and mild to moderate macrophage hyperplasia/infiltration in 5/5 mice in 4.5 and 9 g/g 408 Fab-IL-2 wt-Fab groups and mild to moderate lymphoid hyperplasia/infiltration with marginal to mild macrophage hyperplasia/infiltration in 5/5 mice in 4.5 and 9 g/g 4G8 Fab-IL-2 qm-Fab groups. Immunohistochemistry for 9 g/g 4G8 Fab-IL-2 wt-Fab and 4G8 Fab-IL-2 qm-Fab showed different patterns using the pan-T cell marker CD3, as well as the macrophage marker F4/80. For 9 g/g 4G8 Fab-IL-2 wt-Fab, the pattern of T-cell and macrophage immunoreactivity remained primarily within the red pulp areas, as the architecture of the primary follicles had been altered by lymphocytolysis and necrosis (described below). For 9 g/g 4G8 Fab-IL-2 qm-Fab, special stains showed a pattern similar to that of the vehicle control, but with periarteriolar lymphoid sheath (PALS) white pulp expansion, by a T-cell population and a larger, expanded red pulp area. T-cell and macrophage positivity was also evident within the red pulp, with a similar pattern to the vehicle control group, but expanded. These findings correlate with the gross findings of enlarged spleen. Necrosis was noted marginally in 3/5 mice and marginally to mildly in 5/5 mice in 4.5 and 9 g/g 4G8 Fab-IL-2 wt-Fab groups, respectively. Necrosis was usually located around the area of the primary follicles and selected animals using MSB stain were positive for fibrin in both 4.5 and 9 g/g 4G8 Fab-IG-2 wt-Fab groups which correlates in part with the reduced platelets noted in these animals. Lymphocytolysis was seen in the 4.5 g/g (minimal to mild) and 9 g/g (moderate to marked) 4G8 Fab-IL-2 wt-Fab groups.
[0394] Treatment-related findings in the thymus included minimal to mild increases in lymphocytes in both 4.5 and 9 g/g 4G8 Fab-IL-2 wt-Fab and in 4.5 ug/g 4G8 Fab-IL-2 qm-Fab groups. The cortex and medulla were not individually evident, in 4G8 Fab-IL-2 wt-Fab groups, but immunohistochemistry for the pan T cell marker (CD3) on selected animals in 9 g/g 4G8 Fab-IL-2 wt-Fab and 9 g/g 4G8 Fab-IL-2 qm-Fab groups showed strong positivity for the majority of the cells within the thymus. Increased lymphocytes in the thymus was considered to be a direct pharmacologic effect of both compounds where IL-2 induced proliferation of lymphocytes migrating to the thymus (T cells) from the bone marrow for further differentiation and clonal expansion. This occurred in all groups except 9 g/g 4G8 Fab-IL-2 qm-Fab, which is likely a temporal effect. Lymphocytolysis was mild in 4.5 g/g 4G8 Fab-IL-2 wt-Fab group, and was moderate to marked in the 9 g/g 4G8 Fab-IL-2 wt-Fab group. Moderate lymphoid depletion was noted in both 4.5 and 9 g/g 4G8 Fab-IL-2 wt-Fab groups. While these findings appear more robust in the 4.5 and 9 g/g 4G8 Fab-IL-2 wt-Fab groups, these animals were described as moribund on Day 5, and the mild to marked lymphocytolysis as well as moderate lymphoid depletion may be related to this in-life observation (stress-related effects due to poor physical condition).
[0395] Histopathology findings of uncertain relationship to compound administration in the liver consisted of a marginal mixed cell (lymphocytes and macrophages) infiltrate/activation noted as small foci/microgranulomas scattered randomly throughout the liver in 5/5 mice in both 4.5 and 9 g/g 4G8 Fab-IL-2 qm-Fab groups. This marginal change was also seen in the vehicle control group but with fewer incidence and severity. Stomach glandular dilation and atrophy was seen marginally to mildly in 5/5 mice and ileal villous atrophy was seen marginally in 3/5 mice in the 9 g/g 4G8 Fab-IL-2 wt-Fab group. This finding is most likely attributed to poor physical condition seen in these mice such as reduced body weight, especially in the 9 g/g 4G8 Fab-IL-2 wt-Fab group noted in the in-life observations.
[0396] Injection site findings included mixed cell infiltrate, perivascular edema, and myodegeneration that was noted equally in vehicle control, 9 g/g 4G8 Fab-IL-2 wt-Fab and 9 g/g 4G8 Fab-IL-2 qm-Fab groups. One animal had epidermal necrosis. These findings were not attributed to the treatment(s) itself, but to the daily i.v. injection and handling of the tail. Another animal had macrophage infiltration of the skeletal muscle (noted on the lung tissue histology section) associated with myodegeneration and myodegeneration likely due to a chronic lesion and was not attributed to the treatment. Marginal lymphoid depletion was noted in 3/5 and 4/5 mice in the 4.5 and 9 g/g 4G8 Fab-IL-2 qm-Fab groups, respectively and was most likely attributed to normal physiologic changes seen in the thymus as mice get older (also seen in similar incidence, 4/5 mice, and severity in vehicle control animals).
[0397] In conclusion, the daily intravenous administration of 4G8 Fab-IL-2 wt-Fab or 4G8 Fab-IL-2 qm-Fab at doses of 4.5 or 9 g/g/day for up to 5 days in male mice resulted in similar treatment-related histologic findings with both compounds. However, the findings were generally more prevalent and more severe with FAP-targeted 4G8 Fab-IL-2 wt-Fab in the lung (
[0398] Based on the clinical and anatomic pathologic findings, as well as clinical observations, and the comparable systemic exposure of both compounds, the qm IL-2 under conditions of this study exhibited markedly less systemic toxicity following 5 doses than wt IL-2.
Example 10
[0399] Induction of NK Cell IFN- Secretion by Wild Type and Quadruple Mutant IL-2
[0400] NK-92 cells were starved for 2 h before seeding 100000 cells/well into a 96 well-F-bottom plate. IL-2 constructs were titrated onto the seeded NK-92 cells. After 24 h or 48 h, plates were centrifuged before collecting the supernatants to determine the amount of human IFN- using a commercial IFN- ELISA (BD #550612).
[0401] Two different in-house preparations of wild type IL-2 (probably differing slightly in their 0-glycosylation profiles, see Example 2), a commercially available wild-type IL-2 (Proleukin) and in-house prepared quadruple mutant IL-2 (first batch) were tested.
[0402]
Example 11
[0403] Induction of NK Cell Proliferation by Wild Type Ad Quadruple Mutant IL-2
[0404] NK-92 cells were starved for 2 h before seeding 10000 cells/well into 96-well-black-F-clear bottom plates. IL-2 constructs were titrated onto the seeded NK-92 cells. After 48 h the ATP content was measured to determine the number of viable cells using the CellTiter-Glo Luminescent Cell Viability Assay Kit from Promega according to the manufacturer's instructions.
[0405] The same IL-2 preparations as in Example 10 were tested.
[0406]
[0407] In a second experiment, the following IL-2 preparations were tested: wild-type IL2 (pool 2), quadruple mutant IL-2 (first and second batch).
[0408]
Example 12
[0409] Induction of human PBMC proliferation by Immunoconjugates comprising wild type or quadruple mutant IL-2
[0410] Peripheral blood mononuclear cells (PBMC) were prepared using Histopaque-1077 (Sigma Diagnostics Inc., St. Louis, Mo., USA). In brief, venous blood from healthy volunteers was drawn into heparinized syringes. The blood was diluted 2:1 with calcium- and magnesium-free PBS, and layered on Histopaque-1077. The gradient was centrifuged at 450g for 30 min at room temperature (RT) without breaks. The interphase containing the PBMCs was collected and washed three times with PBS (350g followed by 300g for 10 min at RT).
[0411] Subsequently, PBMCs were labeled with 40 nM CFSE (carboxyfluorescein succinimidyl ester) for 15 min at 37 C. Cells were washed with 20 ml medium before recovering the labeled PBMCs for 30 min at 37 C. The cells were washed, counted, and 100000 cells were seeded into 96-well-U-bottom plates. Pre-diluted Proleukin (commercially available wild-type IL-2) or IL2-immunoconjugates were titrated onto the seeded cells which were incubated for the indicated time points. After 4-6 days, cells were washed, stained for appropriate cell surface markers, and analyzed by FACS using a BD FACSCantoII. NK cells were defined as CD3/CD56.sup.+, CD4.sup.+ T cells as CD3.sup.+/CD8.sup. and CD8 T cells as CD3.sup.+/CD8.sup.+.
[0412]
[0413]
[0414]
[0415]
Example 13
[0416] Effector Cell Activation by Wild-Type and Quadruple Mutant IL-2 (pSTAT5 Assay)
[0417] PBMCs were prepared as described above. 500000 PBMCs/well were seeded into 96-well-U-bottom plates and rested 45 min at 37 C. in RPMI medium containing 10% FCS and 1% Glutamax (Gibco). Afterwards, PBMCs were incubated with Proleukin, in-house produced wild-type IL-2 or quadruple mutant IL-2 at the indicated concentrations for 20 min at 37 C. to induce phosphorylation of STAT5. Subsequently, cells were immediately fixed (BD Cytofix Buffer) for 10 min at 37 C. and washed once, followed by a permeabilization step (BD Phosflow Perm Buffer III) for 30 min at 4 C. Afterwards, cells were washed with PBS/0.1% BSA and stained with mixtures of FACS antibodies for detection of NK cells (CD3.sup./CD56.sup.+), CD8.sup.+ T cells (CD3.sup.+/CD8.sup.+), CD4.sup.+ T cells (CD3.sup.+/CD4.sup.+/CD25.sup./CD127.sup.+) or T.sub.reg cells (CD4.sup.+/CD25.sup.+/CD127.sup./FoxP3.sup.+), as well as pSTAT5 for 30 min at RT in the dark. Cells were washed twice with PBS/0.1% BSA and resuspended in 2% PFA before flow cytometric analysis (BD FACSCantoII).
Example 14
[0418] Effector Cell Activation by Wild-Type and Quadruple Mutant IgG-IL-2 (pSTAT5 Assay)
[0419] Experimental conditions were as described above (see Example 13).
[0420]
Example 15
[0421] Maximum Tolerated Dose (MTD) of FAP-Targeted Fab-IL2 wt-Fab and Fab-IL2 qm-Fab Immunoconjugates
[0422] Escalating doses of FAP-targeted Fab-IL2-Fab immunoconjugates, comprising either wild type (wt) or quadruple mutant (qm) IL-2, were tested in tumor free immunocompetent Black 6 mice. Female Black 6 mice (Charles River, Germany), aged 8-9 weeks at the start of the experiment, were maintained under specific-pathogen-free conditions with daily cycles of 12 h light/12 h darkness according to committed guidelines (GV-Solas; Felasa; TierschG). The experimental study protocol was reviewed and approved by local government (P 2008016). After arrival, animals were maintained for one week to get accustomed to the new environment and for observation. Continuous health monitoring was carried out on a regular basis.
[0423] Mice were injected i.v. once a day for 7 days with 4G8 Fab-IL2 wt-Fab at doses of 60, 80 and 100 g/mouse or 4G8 Fab-IL2 qm-Fab at doses of 100, 200, 400, 600 and 1000 g/mouse. All mice were injected i.v. with 200 l of the appropriate solution. To obtain the proper amount of immunoconjugate per 200 l, the stock solutions were diluted with PBS as necessary.
[0424]
TABLE-US-00022 TABLE 19 Concen- tration Compound Dose Formulation buffer (mg/mL) 4G8 Fab- 60, 80, 25 mM potassium phosphate, 3.32 IL2 wt-Fab 100 g 125 mM NaCl, (=stock 100 mM glycine, pH 6.7 solution) 4G8 Fab- 100, 200, 25 mM potassium phosphate, 4.25 IL2 qm-Fab 400, 600, 125 mM NaCl, (=stock 1000 g 100 mM glycine, pH 6.7 solution)
Example 16
[0425] Pharmacokinetics of a Single Dose of FAP-Targeted and Untargeted IgG-IL2 wt and Qm
[0426] A single dose pharmacokinetics (PK) study was performed in tumor-free immunocompetent 129 mice for FAP-targeted-IgG-IL2 immunoconjugates comprising either wild type or quadruple mutant IL-2, and untargeted IgG-IL2 immunoconjugates comprising either wild type or quadruple mutant IL-2.
[0427] Female 129 mice (Harlan, United Kingdom), aged 8-9 weeks at the start of the experiment, were maintained under specific-pathogen-free conditions with daily cycles of 12 h light/12 h darkness according to committed guidelines (GV-Solas; Felasa; TierschG). The experimental study protocol was reviewed and approved by local government (P 2008016). After arrival, animals were maintained for one week to get accustomed to the new environment and for observation. Continuous health monitoring was carried out on a regular basis.
[0428] Mice were injected i.v. once with FAP-targeted 28H1 IgG-IL2 wt (2.5 mg/kg) or 28H1 IgG-IL2 qm (5 mg/kg), or untargeted DP47GS IgG-IL2 wt (5 mg/kg) or DP47GS IgG-IL2 qm (5 mg/kg). All mice were injected i.v. with 200 l of the appropriate solution. To obtain the proper amount of immunoconjugate per 200 l, the stock solutions were diluted with PBS as necessary. Mice were bled at 1, 8, 24, 48, 72, 96 h; and every 2 days thereafter for 3 weeks. Sera were extracted and stored at 20 C. until ELISA analysis. Immunoconjugate concentrations in serum were determined using an ELISA for quantification of the IL2-immunoconjugate antibody (Roche-Penzberg). Absorption was measured using a measuring wavelength of 405 nm and a reference wavelength of 492 nm (VersaMax tunable microplate reader, Molecular Devices).
[0429]
TABLE-US-00023 TABLE 20 Concen- tration Compound Dose Formulation buffer (mg/mL) 28H1-IgG- 2.5 mg/kg 20 mM Histidine, 3.84 IL2 wt 140 mM NaCl, pH 6.0 (=stock solution) 28H1-IgG- 5 mg/kg 20 mM Histidine, 2.42 IL2 qm 140 mM NaCl, pH 6.0 (=stock solution) DP47GS- 5 mg/kg 20 mM Histidine, 3.74 IgG-IL2wt 140 mM NaCl, pH 6.0 (=stock solution) DP47GS- 5 mg/kg 20 mM Histidine, 5.87 IgG- 140 mM NaCl, pH 6.0 (=stock IL2QM solution)
Example 17
[0430] Pharmacokinetics of a Single Dose of Untargeted Fab-IL2 wt-Fab and Fab-IL2 qm-Fab
[0431] A single dose pharmacokinetics (PK) study was performed in tumor-free immunocompetent 129 mice for untargeted Fab-IL2-Fab immunoconjugates comprising either wild type or quadruple mutant IL-2.
[0432] Female 129 mice (Harlan, United Kingdom), aged 8-9 weeks at the start of the experiment, were maintained under specific-pathogen-free conditions with daily cycles of 12 h light/12 h darkness according to committed guidelines (GV-Solas; Felasa; TierschG). The experimental study protocol was reviewed and approved by local government (P 2008016). After arrival, animals were maintained for one week to get accustomed to the new environment and for observation. Continuous health monitoring was carried out on a regular basis.
[0433] Mice were injected i.v. once with DP47GS Fab-IL2 wt-Fab at a dose of 65 nmol/kg or DP47GS Fab-IL2 qm-Fab at a dose of 65 nM/kg. All mice were injected i.v. with 200 l of the appropriate solution. To obtain the proper amount of immunoconjugate per 200 l, the stock solutions were diluted with PBS as necessary.
[0434] Mice were bled at 0.5, 1, 3, 8, 24, 48, 72, 96 hours and thereafter every 2 days for 3 weeks. Sera were extracted and stored at 20 C. until ELISA analysis. Immunoconjugate concentrations in serum were determined using an ELISA for quantification of IL2-immunoconjugate antibody (Roche-Penzberg). Absorption was measured using a measuring wavelength of 405 nm and a reference wavelength of 492 nm (VersaMax tunable microplate reader, Molecular Devices).
[0435]
TABLE-US-00024 TABLE 21 Concen- tration Compound Dose Formulation buffer (mg/mL) DP47GS 65 nM/kg 100 mM glycine, 3.84 Fab-IL2 wt- 125 mM NaCl, (=stock Fab 25 mM KH.sub.2PO.sub.4, pH 6.7 solution) DP47GS 65 nM/kg 100 mM glycine, 2.42 Fab-IL2 125 mMNaCl, (=stock qm-Fab 25 mM KH.sub.2PO.sub.4, pH 6.7 solution)
Example 18
[0436] Activation Induced Cell Death of IL-2 Activated PBMCs
[0437] Freshly isolated PBMCs from healthy donors were pre-activated overnight with PHA-M at 1 g/ml in RPMI1640 with 10% FCS and 1% Glutamine. After pre-activation PBMCs were harvested, labeled with 40 nM CFSE in PBS, and seeded in 96-well plates at 100 000 cells/well. Pre-activated PBMCs were stimulated with different concentrations of IL-2 immunoconjugates (4B9 IgG-IL-2 wt, 4B9 IgG-IL-2 qm, 4B9 Fab-IL-2 wt-Fab, and 4B9 Fab-IL-2 qm-Fab). After six days of IL-2 treatment PBMCs were treated with 0.5 g/ml activating anti-Fas antibody overnight. Proliferation of CD4 (CD3.sup.+CD8.sup.) and CD8 (CD3.sup.+CD8.sup.+) T cells was analyzed after six days by CFSE dilution. The percentage of living T cells after anti-Fas treatment was determined by gating on CD3.sup.+ Annexin V negative living cells.
[0438] As shown in
[0439] As shown in
[0440] Although the foregoing invention has been described in some detail by way of illustration and example for purposes of clarity of understanding, the descriptions and examples should not be construed as limiting the scope of the invention. The disclosures of all patent and scientific literature cited herein are expressly incorporated in their entirety by reference.