A CD25-BIASED ANTI-IL-2 ANTIBODY

Abstract

The invention provides a human IL-2 (hlL-2)-specific monoclonal antibody, wherein a complex of hlL-2 and the monoclonal induces IL-2 signalling preferentially via CD25 and the trimeric IL-2R. The invention further provides a pharmaceutical composition comprising hlL-2 and said hlL-2-mAb for use treating inflammatory disease.

Claims

1. A human interleukin-2 (hIL-2)-specific monoclonal antibody (mAb), or antigen-binding fragment thereof, wherein the hIL-2-specific mAb interacts with hIL-2 amino acid residues to provide an epitope, and wherein the epitope comprises the hIL-2 residues H16, D20, Q57, E60, E61, L63, K64, E67, E68, and L80, R81, R83, D84, 186, S87, N88, N90, V91, L94, E95, K97, T101, T102, M104; wherein the hIL-2-specific mAb, or antigen-binding fragment thereof comprises a heavy chain variable (V.sub.H) region comprising a V.sub.H complementarity determining region CDR.sub.H1, CDR.sub.H2 and CDR.sub.H3, and a variable light chain (V.sub.L) region comprising a V.sub.L complementarity determining region CDR.sub.L1, CDR.sub.L2 and CDR.sub.L3, and wherein a. CDR.sub.H1 comprises, or is identical to SEQ ID NO 001; and b. CDR.sub.H2 comprises, or is identical to SEQ ID NO 002, and c. CDR.sub.H3, comprises, or is identical to SEQ ID NO 003; and d. CDR.sub.L1 comprises, or is identical to SEQ ID NO 004; and e. CDR.sub.L2 comprises, or is identical to SEQ ID NO 005; and f. CDR.sub.L3 comprises, or is identical to SEQ ID NO 006.

2. A hIL-2-specific mAb, or antigen-binding fragment thereof, according to claim 1, wherein the binding of the hIL-2-specific mAb to hIL-2 is characterized by: a dissociation constant (K.sub.D) equal or smaller than (≤) 4.3×10.sup.−9, particularly a K.sub.D≤5.13×10.sup.−9, an on-rate (K.sub.on) equal or greater than (≥) 4.12×10.sup.5 Ms.sup.−1, particularly a K.sub.on≥4.66×10.sup.5 Ms.sup.−1, and an off-rate (K.sub.off)≤2.20×10.sup.−3 s.sup.−1, particularly a K.sub.off≤2.39×10.sup.−3 s.sup.−1.

3. The hIL-2-specific mAb, or antigen-binding fragment thereof, according to claim 1, wherein a complex of the hIL-2-specific mAb and hIL-2 combined in a ratio between 2:1 to 1:2, particularly combined in a ratio of 1:1 is characterized by: a ratio of binding to the high-affinity hIL-2 receptor compared to the intermediate-affinity hIL-2 receptor of between 20 to 121, particularly a ratio of between 71 and 121, and/or a ratio of binding affinity for CD25 alone, compared to the intermediate-affinity hIL-2 receptor of between 277 to 483, particularly a ratio of between 380 to 483, and/or dissociation of the hIL-2 mAb from hIL-2 on binding of the complex to the high-affinity hIL-2 receptor, and/or activating human CD3.sup.+ CD4.sup.+ CD127.sup.low Foxp3.sup.+ T.sub.reg cells with an EC50≤0.154, and human CD8.sup.+ T cells with an EC50≥442.9.

4. A hIL-2-specific mAb, or antigen-binding fragment thereof, which comprises a heavy chain variable (V.sub.H) region comprising a V.sub.H complementarity determining region CDR.sub.H1, CDR.sub.H2 and CDR.sub.H3, and a variable light chain (V.sub.L) region comprising a V.sub.L complementarity determining region CDR.sub.L1, CDR.sub.L2 and CDR.sub.L3, and wherein a. CDR.sub.H1 comprises, or is identical to SEQ ID NO 001; and b. CDR.sub.H2 comprises, or is identical to SEQ ID NO 002, and c. CDR.sub.H3, comprises, or is identical to SEQ ID NO 003; and d. CDR.sub.L1 comprises, or is identical to SEQ ID NO 004; and e. CDR.sub.L2 comprises, or is identical to SEQ ID NO 005; and f. CDR.sub.L3 comprises, or is identical to SEQ ID NO 006.

5. A hIL-2-specific mAb, or antigen-binding fragment thereof, according to claim 1, wherein a. the CDR.sub.H1, CDR.sub.H2 and CDR.sub.H3 are comprised in a V.sub.H sequence selected from SEQ ID NO 007, SEQ ID NO 008, SEQ ID NO 009, SEQ ID NO 010, SEQ ID NO 011, SEQ ID NO 012, SEQ ID NO 013, and SEQ ID NO 014, particularly wherein the CDR.sub.H are comprised in SEQ ID NO 007, and wherein, b. the CDR.sub.L1, CDR.sub.L2 and CDR.sub.L3 are comprised in a V.sub.L sequence selected from SEQ ID NO 015, and SEQ ID NO 016, particularly wherein the CDR.sub.L are comprised in SEQ ID NO 015.

6. A hIL-2-specific mAb, or antigen-binding fragment thereof, particularly a hIL-2-specific mAb according to claim 1, which comprises: a. a V.sub.H region sequence ≥96% identical to SEQ ID NO 007, and wherein: positions 74, and/or 84 are serine, and/or position 93 is methionine, and/or position 122 is alanine; and, b. a V.sub.L region ≥99% identical to SEQ ID NO 015, and wherein: position 69 is isoleucine.

7. A hIL-2-specific mAb, or antigen-binding fragment thereof, particularly a hIL-2-specific mAb according to claim 1, wherein a. the V.sub.H region comprises a sequence selected from SEQ ID NO 007, SEQ ID NO 008, SEQ ID NO 009, SEQ ID NO 010, SEQ ID NO 011, SEQ ID NO 012, SEQ ID NO 013, and SEQ ID NO 014, or a functionally similar sequence derived from any one of these reference sequences by the substitution rules given below; and b. the V.sub.L region comprises a sequence selected from SEQ ID NO 015, and SEQ ID NO 016, or a functionally similar sequence derived from any one of these reference sequences by the substitution rules given below, and wherein the substitution rules for deriving the functionally similar sequence from their respective reference sequence are: i. glycine (G) and alanine (A) are interchangeable; valine (V), leucine (L), and isoleucine (I) are interchangeable, A and V are interchangeable; ii. tryptophan (W) and phenylalanine (F) are interchangeable, tyrosine (Y) and F are interchangeable; iii. serine (S) and threonine (T) are interchangeable; iv. aspartic acid (D) and glutamic acid (E) are interchangeable v. asparagine (N) and glutamine (Q) are interchangeable; N and S are interchangeable; N and D are interchangeable; E and Q are interchangeable; vi. methionine (M) and Q are interchangeable; vii. cysteine (C), A and S are interchangeable; viii. proline (P), G and A are interchangeable; ix. arginine (R) and lysine (K) are interchangeable; particularly wherein at most two amino acids are exchanged, more particularly wherein at most one amino acid is exchanged by the substitution rules given above.

8. The hIL-2-specific mAb, or antigen-binding fragment thereof, according to claim 1, and further comprising a. a first sequence ≥90% identical, particularly ≥94%, ≥96% or even ≥98% identical to at least one of SEQ ID NO 007, SEQ ID NO 008, SEQ ID NO 009, SEQ ID NO 010, SEQ ID NO 011, SEQ ID NO 012, SEQ ID NO 013, SEQ ID NO 014, and SEQ ID NO 017; and b. a second sequence ≥90% identical, particularly ≥94%, ≥96% or even ≥98% identical to at least one of SEQ ID NO 015, SEQ ID NO 016, and SEQ ID NO 018.

9. The hIL-2-specific mAb according to claim 1, wherein the hIL-2-specific mAb comprises: a. a heavy chain, said heavy chain comprising or consisting of SEQ ID NO 017; and b. a light chain, said light chain comprising or consisting of SEQ ID NO 018.

10. A nucleic acid molecule encoding the hIL-2-specific mAb, or antigen-binding fragment thereof, according to claim 1.

11. A method of treatment for comprising: administrating to a patient in need an effective amount of a pharmaceutical composition a. the hIL-2-specific mAb, or antigen-binding fragment thereof, according to claim 1, and b. hIL-2  wherein the immune-mediated disease is amenable to IL-2 immunotherapy, more particularly immune-mediated diseases selected from allograft-related disorders, chronic inflammation, allergy, autoimmunity, and metabolic disease,  thereby treating the immune mediate diseases.

12. The method of claim 11, wherein the IL-2 and the hIL-2-specific mAb are covalently associated.

13. The method of claim 11, wherein the autoimmune disease is selected from systemic lupus erythematosus, rheumatoid arthritis, ankylosing spondylitis, autoimmune hepatitis, amyotrophic lateral sclerosis, type-1 diabetes mellitus, type-2 diabetes mellitus, atherosclerosis, multiple sclerosis, inflammatory and autoimmune myopathies, alopecia areata, psoriasis or inflammatory bowel disease.

14. The method of claim 11, wherein the allograft related disorder is diagnosed in a patient receiving a solid organ transplant procedure.

Description

DESCRIPTION OF THE FIGURES

[0287] FIG. 1 (A) Screening design to identify IL-2 receptor (IL-2R)-biasing anti-hIL-2 mAbs with in vivo agonistic properties. Cell-based IL-2R binding assay and flow cytometry plots of cells expressing CD25, CD122+CD132 or CD25+CD122+CD132, gated based on their fluorescent “barcode”. Emission spectra of CD25-CyPet, CD122-YPet and CD132-RFP were detected in the AF488, BV421 and APC channels respectively. IL-2 complex (IL-2cx) binding was recorded in the BV605 channel using a rat anti-mouse IgG-BV605. Histograms (right) show binding of IL-2/anti-IL-2 mAb.sub.CD25 (middle), IL-2/anti-IL-2 mAb.sub.CD122 (bottom) and negative control (top). (B) Quantification of IL-2cx-binding as in (A). Readout is the geometric mean fluorescence intensity (gMFI) of BV605 quantified on cells expressing CD25 (white bars) or CD122+CD132 (gray bars) and plotted individually for each anti-IL-2 mAb. Pooled data from 3-4 experiments±SEM. gMFI background was subtracted. Unpaired t-test, two-tailed. (C) Quantification of flow cytometry of IL-2.sub.Rhod (PE channel) and anti-IL-2 mAb (rat anti-mouse IgG-BV605) binding to HEK293T cells expressing CD25-CyPet, CD122-YPet+CD132-RFP or CD25-CyPet+CD122-YPet+CD132-RFP, gated as in (A). Control cells were incubated without IL-2.sub.Rhod. Bar graphs represent frequencies±SEM of IL-2.sub.Rhod positive (unfilled) or IL-2.sub.Rhod/anti-IL-2 mAb double positive (filled) fractions. Pooled data from 3-4 experiments. (D) Matrix of IL-2.sub.Rhod/anti-IL-2 mAb cx clustered based on their binding to CD25 (Y-axis) or CD122+CD132 (X-axes) (left plot), and CD25 (Y-axis) or CD25+CD122+CD132 (X-axis) (right plot). Mean percentages of IL-2.sub.Rhod/anti-IL-2 mAb cx positive population obtained in (C). (E) Comparison of complexed and free IL-2.sub.Rhod upon incubation with cells expressing the trimeric high affinity IL-2R. Mean percentages of IL-2.sub.Rhod versus IL-2.sub.Rhod/anti-IL-2 mAb complex positive populations obtained in (C). (F) Matrix of “CD25 bias” and “IL-2 delivery” of indicated anti-IL-2 mAb clones. Cluster shows IL-2.sub.Rhod/anti-IL-2 mAb cx binding to CD25-binding (Y-axis) versus free IL-2.sub.Rhod binding to CD25+CD122+CD132 (X-axis). Mean percentages of IL-2.sub.Rhod/anti-IL-2 mAb cx positive and IL-2.sub.Rhod positive populations obtained in (C) are displayed. (G) Difference or Ratio between IL-2.sub.Rhod/anti-IL-2 cx-binding to CD25 and CD25+CD122+CD132, indicating anti-IL-2 mAb release. (H) Summary of tested anti-IL-2 mAb clones.

[0288] FIG. 2 (A) C57BLJ6 mice injected with IL-2 (1.5 μg or 30 μg) and IL-2/anti-IL-2 complexes (1.5 μg/15 μg) on days 0, 1, and 2, euthanized on day 4, and frequencies of CD4.sup.+ CD25.sup.+ Foxp3.sup.+ T cells and CD8.sup.+ CD44.sup.hi CD122.sup.+ T cells were analysed in lymph nodes (LN) and spleens using flow cytometry. (A) Frequency and counts of CD4.sup.+ CD25.sup.+ Foxp3.sup.+ T cells in spleens and LNs quantified on day 4. Shown are mean values±SEM from 2-3 experiments, n=2 for IL-2 (30 μg) and UFKA-50; n=5 for IL-2 (1.5 μg), UFKA-10, UFKA-30, UFKA-40, NARA1; n=6 for PBS and UFKA-20. Unpaired t-test, two-tailed. (B) Mean cell counts of CD4.sup.+ CD25.sup.+ Foxp3.sup.+ T cells, CD8.sup.+ CD44.sup.hi CD122.sup.+ T cells, and CD3.sup.− NK1.1.sup.+ CD122.sup.+ from spleens are shown as fold changes to PBS-treated mice. (C) Ratios of splenic CD4.sup.+ CD25.sup.+ Foxp3.sup.+ T and CD8.sup.+ CD44.sup.hi CD122.sup.+ T cell counts are plotted. Mean values±SEM. Unpaired t-test, two-tailed. Red dashed lines represent mean values, obtained from IL-2 (1.5 μg)-treated animals.

[0289] FIG. 3 Wild-type (WT) C57BL/6 mice were injected with IL-2 (1.5 μg or 30 μg) and IL-2/anti-IL-2 complexes (1.5 μg/15 μg) on days 0, 1, and 2, and euthanized on day 4 to determine cell subset frequencies in lymph nodes and spleens by flow cytometry. (A) Frequencies of CD8.sup.+ CD44.sup.hi CD122.sup.+ T cells in lymph nodes and spleens on day 4. (B) Frequencies of splenic CD3.sup.− NK1.1.sup.+ CD122.sup.+ NK cells. Mean values±SEM from 2-3 experiments, n=2 for IL-2 (30 μg) and UFKA-50; n=5 for IL-2 (1.5 μg), UFKA-10, UFKA-30, UFKA-40, and NARA1; n=6 for PBS and UFKA-20. Red dashed lines represent mean values obtained in IL-2 (1.5 μg)-treated animals.

[0290] FIG. 4A single dose of IL-2 (1 μg or 30 μg) or IL-2/UFKA-20 complexes (1 μg /10 μg) was injected into C57BL/6. (A) Mice were euthanized and the frequency of phosphorylated STAT5.sup.+ (pSTAT5) cells among splenic CD4.sup.+ CD25+, CD8.sup.+ T and NK cells was measured by flow cytometry after 2 hours (2 hr), on day 1 (d1), day 2 (d2), day 4 (d4), and day 8 (d8) after injection with PBS (gray), IL-2 (1 μg, white), IL-2 (30 μg, black) or IL-2/UFKA-20 complex (dark grey). Data are represented as mean±SEM of three independent experiments, n=5 mice per group. One-way ANOVA with Tukey's multiple comparison test. (B) Cell counts of splenic CD4.sup.+ CD25.sup.+ Foxp3.sup.+, CD8.sup.+ T and NK cells displayed as fold change to PBS. Data are represented as mean±SEM of 3 independent experiments, n=5 mice per group. One-way ANOVA with Tukey's multiple comparison test.

[0291] FIG. 5 Wild-type (WT) C57BL/6 mice received a single injection of IL-2 (1.5 μg; open symbols), IL-2/UFKA-20 complexes (IL-2/UFKA-20cx; closed black symbols) or IL-2/UFKA-22 complexes (IL-2/UFKA-22cx; closed red symbols). IL-2/UFKA-20cx and IL-2/UFKA-22cx were generated by complexing human IL-2 (1.5 μg) with UFKA-20 (15 μg) and UFKA-22 (15 μg), respectively, at a 1:1 molar ratio. Blood samples were collected at indicated time points (in hours, hr) after injection, and human IL-2 was detected in serum using a sandwich enzyme-linked immunosorbent assay (ELISA). Half-life (t.sub.1/2) values were calculated by fitting exponential, one-phase decay curves. Shown are mean values of n=7 for IL-2, n=9 for IL-2/UFKA-20cx, and n=3 for IL-2/UFKA-22cx.

[0292] FIG. 6 Quantification of IL-2 cx stimulation of pSTAT5 in human CD4.sup.+ CD127.sup.low Foxp3.sup.+ T.sub.reg and CD8.sup.+ T cells measured by flow cytometry. 100 ng/ml IL-2 was complexed at a 1:1 molar ratio with UFKA-20 or UFKA-22. The pSTAT5 levels on indicated human T cell subsets responding to titrated IL-2 (left graphs), IL-2/UFKA-20cx (middle graphs) and IL-2/UFKA-22cx (right graphs). Half maximal effective concentrations (EC50) were calculated for each condition in both CD4.sup.+ CD127.sup.low Foxp3.sup.+ T.sub.reg (open symbols) and CD8.sup.+ T cells (closed symbols). Fitted dose-response curves are shown as lines. Data are represented as mean±SEM of three independent experiments.

[0293] FIG. 7 (A) Experimental setup as described in FIG. 2. Mice received three injections of PBS, IL-2 (1 μg) alone or in complex (1:1 molar ratio) with a chimeric UFKA-20 mAb (chUFKA-20) or the humanized UFKA-20 variants UFKA-22-00 (referred to as UFKA-22), UFKA-22-02, and UFKA-22-07. Mice were euthanized on day 4, frequencies of CD4.sup.+ CD25.sup.+ Foxp3.sup.+ T cells and CD8.sup.+ CD44.sup.hi CD122.sup.+ T cells were analysed in spleens using flow cytometry. Frequency and counts of CD4.sup.+ CD25.sup.+ Foxp3.sup.+ T cells in spleens were quantified on day 4. (B) Ratios of splenic CD4.sup.+ CD25.sup.+ Foxp3.sup.+ T and CD8.sup.+ CD44.sup.hi CD122.sup.+ T cell counts are plotted. Mean values±SEM from 3 experiments, n=3 for PBS, IL-2, and UFKA-22-07; n=5 for UFKA-20, chUFKA-20, UFKA-22-02 and UFKA-22-07.

[0294] FIG. 8 (A) Rhesus macaques received daily injections of IL-2 until day 6 or two injections of IL-2/UFKA-22 complexes or UFKA-22 mAb only on days 0 and 3. Blood was drawn and analysed at indicated timepoints. IL-2 and IL-2/UFKA-22 complex treatment groups were subdivided into two groups receiving low-dose (LD) or high-dose (HD) treatment, corresponding to 10 μg/kg or 33 μg/kg IL-2, respectively, whereas UFKA-22 mAb alone was injected at a dose of 330 μg/kg. EC50s of UFKA-20 (left table) and UFKA-22 (right table) towards human IL-2 and rhesus macaque IL-2 were determined using an IL-2 sandwich ELISA. (B) Representative flow cytometry plots of pSTAT5.sup.+ cells in CD4.sup.+ CD25.sup.+ and CD8.sup.+ CD25.sup.+ T cells in blood. Histograms of pSTAT5 levels at baseline measured on day −8 (top), and upon LD IL-2 (middle) and LD IL-2/UFKA-22 (bottom) treatment on day 1, n=3 per group. (C) Frequency of CD4.sup.+ CD25.sup.+ Foxp3.sup.+ and CD4.sup.+ CD25.sup.+ T cell populations in blood during the study period measured by flow cytometry. (D) gMFI of Foxp3, CTLA-4 and Ki-67 on CD4.sup.+ CD25.sup.+ T cells at indicated time points in blood measured by flow cytometry. For (C) and (D) means±SEM are plotted as white dots (IL-2 treatment group), grey dots (IL-2/UFKA-22 complex treatment group), and dark grey dots (UFKA-22 mAb treatment group). Significance was determined by unpaired t-test, two-tailed, on day 6. ns, not significant. (E) Ratio of T.sub.reg (CD4.sup.+ CD25.sup.+ Foxp3.sup.+ CD4 T) cells to CD8.sup.+ T, NK and B cells counts on day 6. Means±SEM and individual values are plotted. Significance determined by one-way ANOVA, Dunnett's multiple comparison. ns, not significant.

[0295] FIG. 9 (A) Front and 90°-turned side view of IL-2/anti-IL-2 mAb complex structures superimposed on the human quaternary IL-2R complex comprised of IL-2 (purple), CD25 (IL-2Rα, light orange), CD122 (IL-2Rβ, white) and CD132 (IL-2Rγ, gray). Comparison of IL-2-IL-2R complex (PDB: 2651); IL-2/UFKA-20 complex (PDB: 6YE3); IL-2/F5111 complex (PDB: 5UTZ); IL-2/JES6-1 complex (PDB: 4YQX); IL-2/NARA1 complex (PDB: 5LQB). (B) Equilibrium surface plasmon resonance (SPR) quantification of IL-2-IL-2R binding epitope overlaps with IL-2/anti-IL-2 complexes. CD25−, CD122− and CD132-binding sites. Bar graphs indicate the relative overlap between antibody and receptor epitope, calculated based on the buried surface area (Å.sup.2) using PDBePISA. (C) SPR titration of CD25 and CD122 on IL-2 captured by immobilized UFKA-20 or NARA1, as indicated. Data are representative of 3 experiments. RU, resonance units. (D) IL-2 competition between UFKA-20 and CD25−, CD122+CD132− and CD25+CD122 +CD132-expressing HEK293T cells. A fixed concentration of IL-2.sub.Rhod (0.2 μg/ml) was mixed with titrated amounts of UFKA-20 and incubated with IL-2R-expressing HEK293T cells. Mean±SEM of 2 experiments are plotted.

[0296] FIG. 10 Buried surface area (BSA) in square angstrom (Å.sup.2) between human IL-2 and CD25, CD122, CD132, UFKA-20-00, F5111 and NARA1. Calculated using PDBePISA server and crystal structures of the quaternary IL-2-IL-2R complex (PDB: 2651), IL-2/UFKA-20-00 complex (PDB: 6YE3), and IL-2/F5111 complex (PDB: 5UTZ). 3-letter amino acid code of the human IL-2 protein sequence with annotated helices A, A′, B, B′, C and D (Arenas-Ramirez et al. 2015). BSA values (above 0.00 Å.sup.2) for amino acid residues of human IL-2 and their ligands (CD25, CD122, CD132, UFKA-20-00, F5111, NARA1) are shown. In addition, the one-letter amino acid code of the affected amino acid residue is indicated.

[0297] FIG. 11 Predicted Buried surface area in square angstrom (Å.sup.2) showing the predicted UFKA20 binding site on hIL-2.

[0298] FIG. 12 shows optimal UFKA-20 affinity influences the ability to stimulate CD25.sup.+ Foxp3.sup.+ Treg cells in vivo. (A) Binding affinity Dissociation constants (K.sub.D) of indicated UFKA-20 variants were determined using single-cycle surface plasmon resonance (SPR) measurement. (B) C57BL/6 wildtype mice received a single injection of IL-2/anti-IL-2cx (1 ug IL-2:10 ug UFKA-20 variant). Mean values±SD of the frequency of CD4.sup.+ CD25.sup.+ Foxp3.sup.+ T.sub.reg cells in spleens was measured by flow cytometry at day 4. The dotted line indicates the mean value from PBS treated mice. (C) Correlation between antibody K.sub.D and the ability to induce CD4.sup.+ CD25.sup.+ Foxp3.sup.+ Treg cells (mean values±SD of the CD4.sup.+ CD25.sup.+ Foxp3.sup.+ frequency).

[0299] FIG. 13 shows distance measured using PyMOL software in Angstrom (Å) between IL-2's carboxy- (C—) terminus and the amino- (N—) terminus of UFKA-20's variable heavy (vH) or variable light (vL) chain. (A) Crystal structure of IL-2/UFKA-20 Fab complex (PDB: 6YE3). (B) Distance (dotted black line) between the C-terminus of IL-2 and the N-terminus of UFKA-20 vH (top panel) and UFKA-20 vL (bottom panel). UFKA-20 vH is shown in black, UFKA-20 vL in gray and IL-2 in white. (C) Schematic illustration of IL-2/UFKA-22 fusion proteins (FPs), where IL-2 is N-terminally linked to UFKA-22 light or heavy chain. (D) UFKA-22FP vL (G.sub.4S).sub.6 at two different states. (Left) IL-2 is not associated with the binding pocket of the UFKA-22, or the IL-2 is associated with the binding pocket of the UFKA-22 (right).

[0300] FIG. 14 shows comparison of IL-2/UFKA-22 and UFKA-22FPs in vivo in mice. (A) Experimental design: C57BL/6 mice were injected with IL-2/UFKA-22cx (12 μg [2 μg IL-2 and 10 μg UFKA-22]), UFKA-22FP vH (G.sub.4S).sub.5 (12 μg or 24 μg) and UFKA-22FP vL (G.sub.4S).sub.6 (12 μg or 24 μg) on days 0, 1, and 2. Mice were euthanized on day 4, and frequencies of CD4.sup.+ CD25.sup.+ Foxp3.sup.+ Treg cells and CD8.sup.+ CD122.sup.+ CD44.sup.hi T cells were analysed in spleens using flow cytometry. (B) Frequency of CD25.sup.+ Foxp3.sup.+ among CD4.sup.+ T cells, (C) Ki-67.sup.+ on CD25.sup.+ Foxp3.sup.+ Treg cells, and (D) CD8.sup.+ T cells in the spleens of PBS, IL-2/UFKA-22 and UFKA-22FP treated mice as in (A), mean values±SEM. P values were calculated using one-way ANOVA with Tukey's multiple comparison test; ns indicates not significant.

[0301] FIG. 15 shows (A) flow cytometry gating strategy for mouse splenic DC subsets including plasmacytoid DCs (pDC) and conventional type I and type II DCs (cDC1 and cDC2). (B) Splenic cDC expansion from mice treated with IL-2, CD25-biased IL-2/5344 complexes, or CD122-biased IL-2cx (IL-2/NARA1 complexes) (n=7 to 9 mice per group). (C) Proliferation of splenic cDCs from mice receiving the indicated treatments from (B) measured by BrdU incorporation over 3 days. (n=7 to 9 mice per group). (D) abundance of CD40, CD80, CD86, MHC-I, and MHC-II on splenic cDCs of untreated and IL-2cx-treated mice (IL-2/NARA1 complexes) displayed as representative histograms (left panel) and fold change of gMFI normalized to untreated (right panel). Data are presented as mean+/−SEM (n=9 mice per group).

[0302] FIG. 16 shows (A) study design of the investigator-initiated clinical trial testing 1.5 million international units (MIU) aldesleukin injected subcutaneously daily on 5 consecutive days with blood draws before and after aldesleukin injection (upper panel). Corresponding gating strategy of human cDC subsets identifying CD141.sup.+ cDC1 and CD1 c+cDC2 (lower panel). (B) Percentages of Ki67.sup.+ proliferating cDC1 (n=8) and cDC2 (n=10) in peripheral blood from patients before and after aldesleukin treatment.

[0303] FIG. 17 shows (A) quantification of splenic cDCs after treatment with suppressive IL-2cx (IL-2/UFKA-20 complexes) on three consecutive days. (B) Proliferation of splenic cDCs from mice in (A) measured by BrdU incorporation over 3 days. Data are mean+/−SEM (n=7 mice per group). Surface expression of (C) MHC-II, and (D) CD80, on cDC by UFKA-20 complexes measured by flow cytometry on splenic cDCs of untreated and UFKA-20 complex-treated mice as in (A) displayed as fold change of geometric mean fluorescence intensity (gMFI) normalized to untreated. Data are presented as mean+/−SEM (n=5 to 8 mice per group). RNA sequencing of sorted conventional DCs (cDCs) isolated from mice 1 day following three days of UFKA-20 complexes treatment as in (E). Volcano plot of differentially expressed genes, with genes enriched in UFKA-20 complex treated mice on the right, and genes enriched in untreated mice depicted on the left. Each dot represents a single gene, with light grey marking genes that are non-significantly changed compared to untreated and black dots showing genes significantly (P<0.05) different from untreated. Genes coding representative pro- or anti-inflammatory proteins are indicated. Cutoffs were set at log2 ratio of 0.5. Tgfbi, transforming growth factor beta-induced; II1 m, Interleukin 1 receptor antagonist; Tab1, TGF-beta activated kinase 1/MAP3K7 binding protein 1; II6st, interleukin 6 signal transducer; Ltb, lymphotoxin beta; Tnfsf14, tumor necrosis factor (ligand) superfamily, member 14; Csf1, colony stimulating factor 1; Fas, TNF receptor superfamily member 6.

[0304] Table 1. SPR analysis of anti-IL-2 mAbs UFKA-20, UFKA-22-00 (briefly, UFKA-22), UFKA-22-02, and UFKA-22-07, in comparison to the previously-reported anti-IL-2 mAbs JES6-1, F5111, and NARA1.

[0305] Table 2 shows IL-2 binding characteristics of UFKA-22 variants with framework mutations measured by surface plasmon resonance (SPR).

[0306] Table 3 shows V.sub.H (SEQ ID NO 019) and VL (SEQ ID NO 020) alterations in UFKA-20 variants.

[0307] Table 4 shows predicted role of amino acid substitutions in UFKA20 variants from Table 3.

EXAMPLES

Material and Methods

Cell Lines and Primary Cells

[0308] HEK293T cells obtained from the American Type Culture Collection (ATCC) were maintained in Dulbecco's modified Eagle medium supplemented with fetal calf serum (10% v/v, Thermo Scientific) and penicillin-streptomycin (100 U/ml, Thermo Scientific). Peripheral blood mononuclear cells (PBMCs) were isolated by Ficoll-Paque PLUS (GE Healthcare) gradient centrifugation from human peripheral blood collected from healthy individuals after prior informed consent and with approval of the Ethical Committee of the Canton of Zurich (BASEC no. 2016-01440).

Generation of Fluorescently-Tagged IL-2R Subunits

[0309] IL-2R subunits were C-terminally linked to fluorescent proteins with a flexible 15 amino acid (GGGGS).sub.3-spacer (motif SEQ ID NO 021). Sequences encoding CyPet, YPet or RFP657 (RFP) derived from plasmids pCEP4CyPet-MAMM and pCEP4yPET-MAMM (Addgene plasmids 14033 and 14032, respectively, kindly donated by P. Daugherty) and pSG4OC-RFP657 (kindly provided by D. Hecki, Hannover Medical School). CyPet was amplified using specific primers: forward 5′-CGTCTCGTGGTGGTGGTTCTGGTGGTGGTGGTTC-TGTGACAAGG-3′ (SEQ ID NO 022) and reverse 5′-GGTGGTCTCGAGTTATTTGTACA-GTTCGTCCATGCCG TG-3′ (SEQ ID NO 023). The gene sequence of human CD25 was amplified from human PBMC RNA (RNeasy Plus Mini Kit, Qiagen), transcribed to complementary DNA (cDNA) using QuantiTect Reverse Transcription Kit (Qiagen), following PCR amplification using specific primer pairs for human CD25: forward 5′-CTAGGAAGCTTATCTATGGATTCATACCTGCTG-3′ (SEQ ID NO 024) and reverse 5′-ACCAGAACCACCACCACCAGAACCACCACCACCGATTGTTCTTCTA-CTCTTCCTCTG-3′ (SEQ ID NO 025). PCR products were purified from 0.5-1% agarose gels by gel extraction (New England BioLabs), fragments were annealed using overlap extension PCR and cloned into a mammalian expression vector pCSCMV (Addgene plasmid 30530, kindly donated by G. Ryffel). Human CD122-CyPet and human CD132-RFP657 (termed CD132-RFP) were synthesized by GeneArt service (Thermo Scientific) and cloned into the mammalian expression vector pcDNA3.1.

Cell-Based IL-2R Binding Assay

[0310] 0.75×10.sup.6 HEK293T cells were co-transfected in six-well plates with each 1.3 μg pCD25-CyPet, pCD122-YPet and pCD132-RFP using a 1:3 DNA to ViaFect (Promega) ratio in Opti-MEM (Thermo Scientific). Total DNA amount was adjusted to 3.9 μg using empty vector pcDNA3.1, when one or two IL-2R subunits were transfected and culture at 37° C., 5% CO.sub.2. Cells were detached 48 hours after transfection using enzyme-free cell dissociation buffer (Thermo Scientific) and collected in FACS buffer (PBS containing 2% FBS plus 2 mM EDTA). Rhodamine-labelled IL-2 (IL-2.sub.Rhod) and anti-IL-2 mAbs were mixed in a 1:1 molar ratio and incubated for 15 minutes at room temperature (RT). To generate IL-2.sub.Rhod, human IL-2 was reconstituted with sterile water and dialyzed into a 50 mM phosphate buffer (pH=6.5) to optimize for preferred N-terminal rhodamine coupling, followed by incubation with N-hydroxy-succinimidyl (NHS)-rhodamine (Thermo Scientific) for two hours on ice. Non-reacted NHS-rhodamine was removed by gel filtration (Zeba Spin Desalting Columns, 7K MWCO, Thermo Scientific). IL-2.sub.Rhod/anti-IL-2 mAb complexes were incubated with 0.3×10.sup.6 HEK293T cells (expressing IL-2R subunits or mock control) in V-bottom, 96-well plates for 10 minutes at 37° C., washed twice with cold FACS buffer and incubated for 20 minutes with BV605 rat anti-mouse IgG1 (BD Biosciences, clone X56) in the fridge. Following surface staining, cells were washed with PBS, fixed with 2% paraformaldehyde, acquired with a BD LSRFortessa and analysed using FlowJo software (both BD Biosciences).

Mice

[0311] C5761/6J mice were purchased from Charles River Laboratories. Female mice were used for experiments at two to five months of age. Experiments were approved by the Veterinary Office of the Canton of Zurich (license 246/2016) and conducted in accordance with Swiss Federal and Cantonal laws. Mice were randomized by unblinded investigators and held in a specific pathogen-free facility at the University Hospital Zurich following institutional guidelines.

Rhesus Macaques

[0312] The study with rhesus macaques (Macaca mulatta) was carried out at the Biomedical Primate Research Centre (BPRC) in 15 healthy female adults, aged four to 15 years and weighing five to 15 kg. Animals did not show circulating antibodies specific to STLV or SRV and had not received any immunosuppressive or antibody therapy before the study. All procedures and protocols complied with all relevant ethical regulations for animal testing of BPRC's Animal Experiments Committee. Animals were randomized into five groups of three animals each: group 1: LD IL-2 (10 μg/kg); group 2: HD IL-2 (33 μg/kg); group 3: LD IL-2/UFKA-22cx (10/100 μg/kg); group 4: HD IL-2/UFKA-22cx (33/330 μg/kg); and group 5: UFKA-22 (330 μg/kg). IL-2 was given daily by subcutaneous injection, while IL-2/UFKA-22cx and UFKA-22 were injected intravenously on days 0 and 3. Animals were sedated for injections and bleedings.

Clinical Trial and Human Samples

[0313] Human samples were collected within the clinical trial “Open-label, Monocentric, Phase II, Investigator-initiated Clinical Trial on Unbiased Characterization of Immunological Parameters in Interleukin-2-treated Systemic Lupus Erythematosus” (Charact-IL-2, ClinicalTrials identifier: NCT03312335) and the “Fundamental research project for phenotypical and functional characterization of different leukocyte subsets in healthy and diseased individuals” (PFCL-1, BASEC no. 2016-01440). Both projects have been reviewed and approved by the competent Swiss authorities and have been carried out in accordance with principles enunciated in the current version of the Declaration of Helsinki, the guidelines of Good Clinical Practice, and Swiss legal requirements. Prior to enrolment into the clinical trial or sample collection, written informed consent was obtained. Human blood was collected into EDTA Vacutainer tubes (BD Biosciences) followed by Ficoll-Paque PLUS (GE Healthcare) gradient centrifugation for peripheral blood mononuclear cell (PBMC) isolation. Isolated PBMCs were frozen in foetal calf serum (FCS, Gibco) containing 10% dimethyl sulfoxide (Sigma) and stored for less than 1 year in liquid nitrogen prior to analysis. Serum was isolated from blood collected with Clot Activator Vacutainer tubes (BD Biosciences) and stored for less than 18 months at −80° C. prior to analysis. For evaluation of IL-2-mediated expansion of cDCs and lymphocytes, blood from patients with systemic lupus erythematosus (SLE) was collected prior and after a 5-day course of daily 1.5 million international units (MIU) of aldesleukin (Proleukin®, Novartis Pharma), according to the study protocol.

IL-2 mAb Complex Formation

[0314] For the HEK cell-based assay IL-2.sub.Rhod was mixed with anti-hIL-2 antibodies at a 1:1 ratio in FACS buffer (1×PBS, 2% FBS, 2 mM EDTA) and incubated at room temperature for at least 15 minutes. For in vivo applications, hIL-2 was mixed with anti-hIL-2 antibodies at a 1:1 ratio in sterile PBS and incubated at room temperature for at least 15 minutes. Injection volume was 200 microliter per intraperitoneal injection. Recombinant human IL-2 (teceleukin, Roche) was obtained from the National Cancer Institute of the National Institutes of Health. Antibody complexes were prepared by mixing 15,000 IU IL-2 and 15 μg anti-IL-2 monoclonal antibodies (mAbs) per injection, as previously described (Arenas-Rameriz N. Sci Transl Med 2016, 8:367ra166). IL-2cx, or 200,000 IU IL-2 were injected daily for three consecutive days. BrdU-incorporated cells were measured using the FITC BrdU Flow Kit (BD Biosciences) according to manufacturer's instructions.

Flow cytometry

[0315] Single cell suspensions of LNs and spleens were prepared and stained for surface markers and intracellular Foxp3 and Ki-67, using a Foxp3/transcription factor intracellular staining kit according the manufacturers' instructions (Thermo Fisher). To detect pSTAT5 in mice or macaques, cells were immediately fixed using Phosflow Lyse/Fix Buffer (BD Biosciences) or lysing solution (Becton Dickinson), and further processed for intracellular staining according to manufacturer's instructions. To measure pSTAT5 in vitro, 10.sup.6 magnetically-purified human CD3.sup.+ T cells (BioLegend) were seeded in 96-well, V-bottom plates and stimulated for 15 minutes at 37° C. using IL-2, IL-2/UFKA-20cx, or IL-2/UFKA-22cx. Intracellular pSTAT5 was stained as aforementioned using anti-STAT5 (pY694) mAb (Thermo Fisher). For surfaces staining of macaque cells, we incubated mAbs with 200 μl EDTA blood, followed by red blood cells lysis, fixation and permeabilization for intracellular staining of Foxp3 and Ki-67, according to standard protocols. Samples were acquired on a BD LSRFortessa and analysed using FlowJo. Antibodies and fluorescent dyes used for flow cytometry were purchased from ebioscience, BD Biosciences, Biolegend or Miltenyi.

ELISA

[0316] Flat-bottom Nunc MaxiSorp 96-well plates (Thermo Scientific) were coated overnight at 4° C. with NARA1 anti-human IL-2 mAb (capture). After washing the plates with PBS, 0.1% Tween 20 (Sigma-Aldrich), wells were blocked for >1 hour at RT with PBS, 1% BSA (Sigma-Aldrich), 0.1% Tween 20 solution, shaking at 450 rpm. Cell supernatants or purified UFKA mAbs were incubated for one to two hours on plates, where IL-2 was directly coated or captured by plate-coated NARA1. After washing the plates, IL-2 or competitive binding was assessed by incubating the plates with anti-mouse IgG (BioLegend) or biotinylated anti-IL-2 detection mAb (clone 5344.111, BD Biosciences) for one hour at RT and 450 rpm. After an additional wash, plates were incubated with streptavidin-conjugated horseradish peroxidase (BD Biosciences) for 45 minutes at RT in the dark. Finally, after a last wash, plates were developed with TMB Peroxidase EIA substrate (BioRad) for two to five minutes, and stopped by adding H.sub.2SO.sub.4 (1.8 M, Sigma-Aldrich). Absorbance at 450 nm was read using an iMark microplate reader (BioRad). Serum half-life of IL-2 or IL-2/UFKA-20cx was measured using a sandwich ELISA, where NARA1 served as capture and a biotinylated anti-IL-2 mAb (clone 5334, R&D Systems) as detection mAb, followed by development, as above.

Surface Plasmon Resonance

[0317] For SPR studies, UFKA-20 or NARA1 were directly immobilized onto a CMD200 chip (XanTec bioanalytics) and titrated IL-2 concentrations starting from 300 nM followed by 2-fold dilutions were injected. To measure CD25 and CD122 binding, IL-2 (1000 nM) was captured for 60 seconds on the anti-IL-2 mAb-coated chip, followed by serial injections of recombinant CD25 or CD122 (R&D Systems), starting with 333 nM and followed by three-fold dilutions. Chip surface was regenerated after every cycle using glycine buffer pH 1.5. Measurements were acquired at 20° C. and analysed on a Biacore T100 (GE Healthcare).

Structural Analysis of the IL-2/UFKA-20cx

[0318] Fab fragments of UFKA-20 were generated by papain cleavage of the full-length mAb followed by Protein A purification. 1.5 ml UFKA-20 (15.3 mg/ml in 50 mM with 90 mM NaCl at pH 7.0) was mixed with dichlorodiphenyltrichloroethane (DDT) and papain (Roche) to reach a final concentration of 5 mM and 1.5 mg/ml, respectively. After 16 hours of digestion at RT, papain was deactivated using 56 mM E64 solution (Roche) and diluted ten times with Tris/NaCl buffer (25 mM Tris, 25 mM NaCl, pH 8.0). The mixture was loaded on a Protein A column equilibrated with Tris/NaCl buffer, and the flow-through fraction harbouring the Fab fragments was collected and further purified by sized exclusion chromatography (SEC). IL-2/UFKA-20 Fab complexes, formed by mixing purified UFKA-20 Fab with a 10-fold molar excess of human IL-2 dissolved in water, were purified by SEC using a Superdex 200 10/300 GL column on an Akta pure chromatography system (GE Healthcare). Fractions containing the complexes were pooled and dialyzed overnight at 4° C. against Tris/NaCl buffer (pH 7.4) and concentrated using Amicon Ultra-Centrifugal Filter Units (10-kDa, MerckMillipore) to a final protein concentration of 10 mg/ml as measured by absorption at 280 nm. Various crystallization buffers were screened and refined to find optimal crystallization conditions. Finally, the IL-2/UFKA-20 Fab complex solution was mixed 1:1 with a crystallization buffer comprising 10.86% (v/v) PEG 8000, 5.76% (v/v) ethylene glycol, 100 mM HEPES (pH 7.48). Crystals were grown by sitting-drop vapor diffusion in 96-well plated at 20° C., harvested and cryoprotected using reservoir solution supplemented with 30% (v/v) ethylene glycol and immediately frozen in liquid nitrogen. Diffraction data were collected at a wavelength of 1 Å at beamline X06DA (Swiss Light Source, Paul Scherrer Institut, Villigen, Switzerland), which is equipped with a Pilatus 2M detector (Dectris, Baden-Wättwil, Switzerland). Data processing was done using XDS and Aimless. The IL-2/UFKA-20 Fab complex structure was solved by molecular replacement using MOLREP, first with the structure of a Fab fragment of an anti-leukotriene antibody (PDB: 5B6F) and subsequently with the structure of human IL-2 (PDB: 1M47) (Arkin M. R. et al. PNAS 2003 100:1603) as search models. Model building was done in Coot and refined using REFMAC5, BUSTER, and PHENIX. We used TLS refinement where each domain was defined as an individual TLS group. The final structure contained three IL-2/UFKA-20 complexes in the asymmetric unit. Epitope overlaps of IL-2R subunits with anti-IL-2 mAbs were quantified using the protein interfaces, surfaces and assemblies' service (PISA) at the European Bioinformatics Institute (http://www.ebi.ac.uk/pdbe/prot_int/pistart.html), and further computed using Excel (Microsoft).

RNA Sequencing (RNA-Seq)

[0319] Forty thousand splenic mouse cDCs from untreated and UFKA20 complex-treated wild-type mice were separated by FACS in RLT Plus lysis buffer (Qiagen) containing 1% 2-mercaptoethanol (Sigma-Aldrich). Subsequently, RNA was isolated using the RNeasy Plus Micro Kit (Qiagen). The RNA extracted from sorted cells was quantified for quality and concentration using the TapeStation RNA high sensitivity kit (Agilent). SMARTer Stranded Total RNA Seq Kit v2 (Takara Bio) was used to prepare cDNA by universal priming (with 3 min fragmentation) and to deplete ribosomal cDNA with ZapR v2 and R Probes v2. The libraries were quantified by Tapestation D1000 (Agilent) measurements, and sequenced on a HiSeq 4000 platform using 125 cycles single-read targeting ˜40M reads per sample. Adapters and low-quality tails were trimmed from reads prior to read alignment. STAR aligner (v2.5.4b) was used to align the RNA-seq dataset to Ensembl genome build GRCh38. p10 (Release 91). Gene expression counts were calculated with feature counts from Bioconductor package Rsubread (v1.32.1). A gene was considered as expressed if, in at least one group of the comparison, it had more than 10 counts in more than half of the samples. Differentially expressed genes were detected using Bioconductor package EdgeR (v3.20.6). Gene set enrichment analysis was done with Gene Ontology analyser for RNA-seq and other length biased data (goseq, v1.30.0).

Quantification and Statistical Analysis

[0320] Statistical testing was performed using the Prism software (GraphPad). As indicated in the figure legends, most experiments were analysed by one-way ANOVA with Tukey's or Dunnett's multiple comparison, or with two-tailed unpaired Student's t-test. For datasets where the count was too small for the normality test, normal distribution was assumed based on data distribution. p<0.05 was considered significant.

Example 1

Generation and Selection of Anti-Human IL-2 Monoclonal Antibodies

[0321] Balb/c mice were immunized with human IL-2 in complete Freund's adjuvant (Sigma-Aldrich) and boosted twice with IL-2 emulsified in incomplete Freund's adjuvant (Sigma-Aldrich). Four to five weeks after the first immunization, mice were sacrificed to collect spleens. Splenocytes were mixed with myeloma cells at a 5:1 ratio with polyethylene glycol 1500 (Roche). Clones were cultured in Iscove's modified Dulbecco's medium supplemented with 10% fetal bovine serum (FBS), 50 mM mercaptoethanol, 1:100 insulin-transferrin-selenium, 2% IL-6-conditioned medium, penicillin-streptomycin, gentamicin (all from Life Technologies), and hypoxanthine-aminopterin-thymidine (HAT) (Sigma-Aldrich). B cell hybridoma supernatants were screened for IL-2 reactivity using a direct IL-2-binding ELISA and for specificity using a competition ELISA, followed by subcloning of positive hits. mAbs were expanded in hypoxanthine-thymidine (HT) medium (LifeTechnologies). After retesting, anti-IL-2 mAbs were purified from cell supernatants using Protein G agarose purification (Thermo Fisher). Antibodies were produced by transiently transfected HEK293F cells, affinity-purified using Protein A MabSelect SuRe resin (GE Healthcare) and fractionated. Purity was analysed by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis.

Example 2

Biasing Anti-IL-2 mAbs have Distinct Properties of IL-2R-Binding and IL-2 Delivery

[0322] Using a competition enzyme-linked immunosorbent assay (ELISA), over ten thousand anti-human IL-2 mAbs, including those generated in Example 1, proprietary antibodies and publicly available clones from mouse hybridoma libraries were assessed for their IL-2 and IL-2R binding properties. Unless otherwise specified, all IL-2 and IL-2R subunits in the examples refer to human molecules. To identify and compare CD25-biasing anti-IL-2 mAbs, a novel cell-based in vitro screening platform where monomeric CD25, dimeric CD122+CD132 and trimeric CD25+CD122+CD132 were expressed on human cells was developed. Fluorescently-tagged IL-2R subunits were generated, and transiently expressed in human embryonic kidney (HEK) 293T cells, enabling precise identification by flow cytometry of cells expressing defined IL-2R subsets and quantification of the binding of rhodamine-labelled IL-2 (IL-2.sub.Rhod), either alone or in complex with an anti-IL-2 mAb. CD25-biased IL-2cx associated with CD25 but not with CD122+CD132, whereas CD122-biased IL-2cx showed the opposite pattern (FIG. 1A). Five anti-IL-2 mAbs were selected on their different binding patterns, termed UFKA-10, UFKA-20 (heavy chain SEQ ID N019, light chain SEQ ID NO 020), UFKA-30, UFKA-40, and UFKA-50, which fell into three broad categories: unbiased (UFKA-10), CD25-biased (UFKA-20, UFKA-30, and UFKA-40), and CD122-biased (UFKA-50) (FIG. 1B). As expected, IL-2.sub.Rhod alone showed low binding to CD25, intermediate association with CD122+CD132, and strong binding to CD25+CD122.sup.+CD132 (FIG. 10). Next, IL-2.sub.Rhod was complexed at a 1:1 ratio with our different anti-IL-2 mAbs and tested them on IL-2R subunit-expressing HEK293T cells. Compared to IL-2.sub.Rhod, complexes of IL-2.sub.Rhod with UFKA-10 showed a slightly reduced binding to CD25 and CD122+CD132, suggesting mild interference of UFKA-10 with these receptor subunits, whereas association of IL-2.sub.Rhod with CD25+CD122 +CD132 was unaltered by this mAb (FIG. 1C). A clear pattern of CD25 bias emerged when testing the mAbs UFKA-20, UFKA-30 and UFKA-40, although with distinct differences between the mAbs (FIGS. 10 and 1D). Thus, IL-2.sub.Rhod/UFKA-30cx and IL-2.sub.Rhod/UFKA-40cx bound preferentially to CD25 and CD25+CD122+CD132 and remained bound to 74.5% and 93.2% as complexes, respectively, whereas their association with CD122+CD132 was either slightly reduced (as with UFKA-30) or remained unchanged (as with UFKA-40) (FIG. 1C-1E). Notably, IL-2.sub.Rhod/UFKA-20cx preferentially associated with CD25, with about two thirds of the measured interactions being made by IL-2.sub.Rhod/UFKA-20cx, while in less than a third IL-2.sub.Rhod alone was detected on CD25. However, IL-2.sub.Rhod/UFKA-20cx appeared to rapidly dissociate upon interaction with trimeric CD25+CD122+CD132, as evidenced by less than 8.4% of this interaction being formed by IL-2.sub.Rhod/UFKA-20cx and 91.6% by uncomplexed, free IL-2.sub.Rhod binding to CD25+CD122+CD132 (FIG. 1C-1E); yet, binding of IL-2.sub.Rhod/UFKA-20cx to CD122+CD132 was disfavoured by the presence of UFKA-20 compared to IL-2.sub.Rhod (FIG. 10), suggesting that UFKA-20 conferred a “CD25 bias” to IL-2.sub.Rhod and, simultaneously, allowed IL-2.sub.Rhod to dissociate from UFKA-20 and bind to trimeric CD25+CD122+CD132, thereby allowing “IL-2 delivery” to a signalling IL-2R complex (FIG. 1F-1H). Conversely, UFKA-30 and UFKA-40 enforced an even stronger CD25 bias upon IL-2.sub.Rhod, but they failed to dissociate and deliver IL-2 to trimeric IL-2R (FIG. 1F-1H). Unlike the aforementioned CD25-biasing anti-IL-2 mAbs, IL-2.sub.Rhod/UFKA-50cx showed reduced CD25 binding and clearly favoured association with dimeric CD122+CD132 and trimeric CD25+CD122+CD132 IL-2Rs (FIG. 1C-1H), similar to IL-2cx with the well-characterized CD122-biasing NARA1 mAb (Arena-Ramirez et al. Sci. Transl. Med. 2006 8:367). Taken together, distinct differences were observed in mechanisms of CD25-biasing mAbs in terms of the two features CD25 bias, and IL-2 delivery to signalling IL-2Rs.

Example 3

CD25 Bias and IL-2 Delivery are Key for Selective Stimulation of Mouse T.SUB.reg .Cells

[0323] The in vivo activity our CD25-biasing mAbs in mice was then assessed. C57BL/6 WT mice received daily injections for three days of IL-2 alone or in complex with UFKA-10, UFKA-20, UFKA-30, UFKA-40, and NARA1, followed by flow cytometry analysis of CD4.sup.+ CD25.sup.+ Foxp3.sup.+ T.sub.reg, CD8.sup.+ CD44.sup.hi CD122.sup.+ memory T, and CD3.sup.− NK1.1.sup.+ CD122.sup.+ NK cells in lymph nodes (LNs) and spleens of treated animals (FIGS. 2A, 3A, and 3B). Controls receiving saline showed on average 8.7% CD4.sup.+ CD25.sup.+ Foxp3.sup.+ T.sub.reg cells and 7% CD8.sup.+ CD44.sup.hi CD122.sup.+ memory T cells in their LNs, as well as 8.6% CD4.sup.+ CD25.sup.+ Foxp3.sup.+ T.sub.reg cells, 13.4% CD8.sup.+ CD44.sup.hi CD122.sup.+ memory T cells and 2.8% NK cells in their spleens (FIGS. 2A, 3A, and 3B). Upon LD IL-2 treatment (1.5 μg per mouse daily), percentages and counts of CD4.sup.+ CD25.sup.+ Foxp3.sup.+ T.sub.reg and CD8.sup.+ CD122.sup.+ memory T cells increased about two to three-fold in both LNs and spleens (FIGS. 2A, 2B and 3A). IL-2cx made with UFKA-10, UFKA-30, and UFKA-40 only discretely improved the stimulation of CD4.sup.+ CD25.sup.+ Foxp3.sup.+ T.sub.reg cells over that seen with LD IL-2 (FIGS. 2A and 2B), although these IL-2cx were able to curtail the expansion of CD8.sup.+ CD44.sup.hi CD122.sup.+ memory T cells (FIG. 2A-2C). However, IL-2/UFKA-20cx induced a vigorous expansion of CD4.sup.+ CD25.sup.+ Foxp3.sup.+ T.sub.reg cells, reaching a total of almost 20×10.sup.6 T.sub.reg cells in LNs and spleens, compared to 1.5×10.sup.6 in saline- and 3×10.sup.6 in LD IL-2-treated animals, whereas percentages and counts of CD8.sup.+ CD44.sup.hi CD122.sup.+ memory T cells remained unaltered compared to IL-2 alone (FIG. 2A-2C). As expected, the CD122-biased IL-2/NARA1cx induced preferential stimulation of CD8.sup.+ CD44.sup.hi CD122.sup.+ memory T cells, although some CD4.sup.+ CD25.sup.+ Foxp3.sup.+ T.sub.reg cell expansion was observed (FIG. 2A-2C), probably due to dissociation of IL-2 from NARA1 (Arenas-Ramirez N. et al. Trends Immunol. 2015 36:763). Collectively, these data demonstrate that CD25 bias alone, as seen with IL-2/UFKA-30cx and IL-2/UFKA-40cx in vitro, is insufficient to stimulate T.sub.reg cells in vivo. Rather, an IL-2cx's capacity to confer mild CD25 bias and deliver IL-2 efficiently to trimeric IL-2Rs, as with IL-2/UFKA-20cx, appear to be necessary features for their in vivo selectivity and efficacy. Hence, these screening assays identified UFKA-20 as a candidate mAb for further characterization as an improved method to increase IL-2 signalling to T.sub.reg cells to inhibit inflammation.

Example 4

IL-2/UFKA-20cx Provide Improved Signalling to Mouse T.SUB.reg .Cells In Vivo

[0324] In a time-course experiment, we compared the capacity of a single intraperitoneal injection of LD IL-2 (1 μg) versus IL-2/UFKA-20cx (1 μg/10 μg) to induce signalling as measured by intracellular staining of phosphorylated STAT5(pSTAT5) in CD4.sup.+ CD25.sup.+ T, CD8.sup.+ T, and NK cells in spleens of mice (FIG. 4). In mice receiving LD IL-2, we observed a preferential stimulation of CD4.sup.+ CD25.sup.+ T cells at two hours after injection, but this effect had already disappeared on day 1 (FIG. 4A). CD8.sup.+ T and NK cells did not phosphorylate STAT5 in response to LD IL-2, however, using a single injection of HD IL-2 (30 μg) resulted in robust signalling in all three lymphocyte subsets, including CD8.sup.+ T and NK cells (FIG. 4A). IL-2/UFKA-20cx showed a high selectivity for CD4.sup.+ CD25.sup.+ T cells that was apparent at two hours after injection and lasted for at least two days, whereas pSTAT5 levels in CD8.sup.+ T and NK cells remained unaffected by this treatment (FIG. 4A). As a consequence, cell counts of CD4.sup.+ CD25.sup.+ T cells increased on day two, peaked on day 4 and had returned to baseline on day 8 after a single injection of IL-2/UFKA-20cx, whereas CD8.sup.+ T and NK cell counts did not change significantly over the course of the experiment, compared to IL-2 (FIG. 4B). The pSTAT5 signalling profile not only confirmed the selectivity of IL-2/UFKA-20cx for CD4.sup.+ CD25.sup.+ T cells but also suggested IL-2/UFKA-20cx had a much longer in vivo half-life than IL-2. To test this, a single injection of IL-2 or IL-2/UFKA-20cx was given to mice, followed by measurement of free or UFKA-20-complexed IL-2 by a sandwich ELISA using NARA1 as capture mAb and MAB202 as detection mAb. IL-2 was only detectable at 30 minutes after injection and had disappeared within four hours, whereas IL-2/UFKA-20cx peaked at four hours and were present for over 24 hours after injection (FIG. 5). The estimated in vivo half-life of IL-2/UFKA-20cx was 30 hours compared to roughly 30 minutes for IL-2. Serum half-life of IL-2/UFKA-22cx was about 21 hours.

Example 5

IL-2/UFKA-20cx Selectively Stimulate Human T.SUB.reg .Cells In Vitro

[0325] The activity of IL-2/UFKA-20cx was assessed on freshly-isolated resting human T cell subsets, including CD4.sup.+ CD25.sup.+ T cells that carry trimeric IL-2Rs and CD8.sup.+ T cells that are equipped with dimeric (CD122+CD132) IL-2Rs, as previously shown (Arena-Ramirez, 2006). CD3.sup.+ T cells were purified from peripheral blood of healthy human donors and stimulated with titrated IL-2 and IL-2/UFKA-20cx (at a 1:1 molar ratio of IL-2 and UFKA-20) for 15 minutes, followed by flow cytometry assessment of intracellular pSTAT5 on gated CD4.sup.+ CD25.sup.+ CD127I° Foxp3.sup.+ T.sub.reg and CD8.sup.+ T cells. IL-2 at a concentration as low as 0.1 ng/ml was able to induce half-maximal STAT5 activation in CD4.sup.+ T.sub.reg cells, whereas about 1000-fold higher concentrations were needed to achieve a comparable STAT5 activation in CD8.sup.+ T cells (FIG. 6), consistent with a previous publication (Yu, Diabetes 2015 64:2172). IL-2/UFKA-20cx were comparably efficient as IL-2 in stimulating pSTAT5 in human CD4.sup.+ T.sub.reg cells (FIG. 6). Contrarily, roughly 17-fold higher concentrations of IL-2/UFKA-20cx were necessary to achieve 50% pSTAT5.sup.+ CD8.sup.+ T cells, based on half maximal effective concentrations (EC50) (FIG. 6). The demonstrable improved CD25-biasing selectivity and efficacy of UFKA-20 in vitro measured using human IL-2R-bearing cell lines, in different in vivo experiments in mice, and in vitro using freshly-isolated primary T cell subsets from different healthy donors was followed by generation of several humanized versions of UFKA-20. Human germline genes sharing the highest level of identity with the V.sub.L (SEQ ID NO 020) and V.sub.H (SEQ ID NO 019) framework sequences of UFKA-20 were identified, codon optimized, and synthesized by GeneScript Custom Gene Synthesis and cloned into an expression vector containing Fc-silent (N297A) human IgG1. The complementarity-determining regions (CDR) of UFKA-20 were transferred onto a human immunoglobulin G1 (IgG1) backbone carrying an N297A mutation, which prevents glycosylation at this site and hence largely diminishes Fc γ receptor binding and effector functions (Park H. I. et al. Trends Biotenchol 2016 34:895; Arnold J. N. et al. Annu. Rev. Immunol. 2007 25:21). The best humanized candidate was named UFKA-22-00, shortened to UFKA-22, bearing a heavy chain of SEQ ID NO 017, and a light chain of SEQ ID NO 018. IL-2/UFKA-22cx and a similar clones sharing the same CDR but bearing framework mutations showed comparable stimulatory activity and selectivity as IL-2/UFKA-20cx in kinetic binding analysis with IL-2 (Table 2), in vitro using freshly-isolated human T cells (FIG. 6), or injected in mice (FIG. 7). Altogether, complexes of IL-2 with UFKA-20 and its humanized version UFKA-22 show strong in vitro activity on human T.sub.reg cells, whereas their activity on human CD8.sup.+ T cells is markedly reduced compared to uncomplexed, free IL-2.

Example 6

Humanized IL-2/UFKA-22cx Show In Vivo Selectivity for T.SUB.reg .Cells in Rhesus Macaques

[0326] IL-2R subunits share a high degree of homology between humans and rhesus macaques. Accordingly, a homology search on National Center for Biotechnology Information (NCBI) was performed using the Basic Local Alignment Search Tool (BLAST), finding that identity of CD25, CD122 and CD132 was 91.9% (accession number NP 001028089.1), 94.2% (NP 001244989.1) and 97.3% (NP001030606.1), respectively, between these two species. Both the mouse antibody UFKA-20 and humanized UFKA-22 clone showed similar binding to either macaque or human IL-2 in vitro (data not shown). To compensate for the difference in in vivo half-lives between IL-2 and IL-2/UFKA-22cx, animals were injected daily on days 0 to 6 with IL-2 (aldesleukin) at 10 μg/kg (LD) or 33 μgpg/kg (HD), whereas IL-2/UFKA-22cx at 10 μg/kg IL-2 and 100 μg/kg mAb (LD) or 33 μg/kg IL-2 and 330 μg/kg mAb (HD) were administered on days 0 and 3 (FIG. 8A). Injection of UFKA-22 at 330 μg/kg (without IL-2) on days 0 and 3 served to assess whether it bound endogenous macaque IL-2. Assessment of parameters on day −8 served as baseline and untreated control. pSTAT5 levels were measured one day after the first injection, significantly increased pSTAT5 levels were observed in CD4.sup.+ CD25.sup.+ T cells following HD IL-2 as well as LD and HD IL-2/UFKA-22cx (FIG. 8B), whereas LD IL-2 or UFKA-22 alone did not alter pSTAT5 levels of CD4.sup.+ CD25.sup.+ T cells over what was measured at baseline on day −8 (FIG. 8B). Overall, the increase in pSTAT5 levels in CD4.sup.+ CD25.sup.+ T cells was more pronounced with IL-2/UFKA-22cx than with IL-2. In contrast to CD4.sup.+ CD25.sup.+ T cells, pSTAT5 levels in CD4.sup.+ CD25.sup.−, CD8.sup.+ CD25+, and CD8.sup.+ CD25.sup.− T cells were not significantly changed by IL-2, IL-2/UFKA-22cx or UFKA-22 compared to baseline (FIG. 8B). To assess selectivity of IL-2/UFKA-22cx for T.sub.reg cells, dose- and time-dependent changes of T.sub.reg cells in the blood of macaques were quantified. We observed the strongest changes on day 6 where percentages of CD4.sup.+ CD25.sup.+ Foxp3.sup.+ T cells and CD4.sup.+ CD25.sup.+ T cells were significantly higher in animals receiving LD IL-2/UFKA-22cx compared to LD IL-2 (FIG. 8C). CD4.sup.+ CD25.sup.+ Foxp3.sup.+ T cells increased on average up to 29% of total CD4.sup.+ T cells after two injections of LD IL-2/UFKA-22cx, whereas seven daily low-dose IL-2 injections only resulted in 4.8% CD4.sup.+ CD25.sup.+ Foxp3.sup.+ T cells (FIG. 8C). HD IL-2 and HD IL-2/UFKA-22cx resulted in comparable T.sub.reg cell responses, although they did not surpass the effects seen with LD IL-2/UFKA-22cx (FIG. 8C). Levels of Foxp3 and cytotoxic T-lymphocyte-associated antigen 4 (CTLA-4) significantly increased in CD4.sup.+ CD25.sup.+ T cells on days 3 and 6 after injection of LD IL-2/UFKA-22cx compared to LD IL-2, whereas HD IL-2/UFKA-22cx and HD IL-2 did not provide any further advantage (FIG. 8D). Similarly, CD4.sup.+ CD25.sup.+ T cells became Ki-67.sup.+ from day 2 on, with Ki-67 levels peaking on days 3 and 6 (FIG. 8D), which most likely reflected IL-2 signalling-induced cell cycle and proliferation. Injection of UFKA-22 without IL-2 led to a small but distinct reduction of CD4.sup.+ CD25.sup.+ T cell frequencies (FIG. 8C), which was probably caused by mild neutralization of endogenous macaque IL-2 by UFKA-22. Foxp3, CTLA-4 and Ki-67 expression levels remained unchanged in the UFKA-22 group (FIG. 8D). Calculating the ratios of CD4.sup.+ CD25.sup.+ Foxp3.sup.+ T.sub.reg cells to CD8.sup.+ T cells, NK cells, and B cells, showed that LD IL-2/UFKA-22cx achieved the best T.sub.reg cell selectivity (FIG. 8E). Humanized IL-2/UFKA-22cx in rhesus macaques confirmed the selectivity of these IL-2cx for CD4.sup.+ CD25.sup.+ Foxp3.sup.+ T.sub.reg cells, and their superiority over IL-2.

Example 7

UFKA-20 Sterically Interferes with IL-2 Binding to CD122 and CD25

[0327] To obtain structural and further mechanistic insight into the IL-2/UFKA-20 interaction, a fragment antigen-binding (Fab) variant of UFKA-20 was generated and complexed with IL-2. The IL-2/UFKA-20 Fab complex was then crystalized for structural analysis. Crystals grew at physiological pH (pH 7.48) and diffracted to 2.89 A resolution. The structure was solved by molecular replacement and comprised three IL-2/UFKA-20 Fab complexes in the asymmetric unit. Compared to the crystal structure of the human IL-2 quaternary, UFKA-20 bound IL-2 dorsolaterally, with an angle of about 55° anti-clockwise to the vertical axis, and differed markedly from those of IL-2 in complex with F5111 (worldwide protein databank PDB: 5UTZ), JES6-1 (PDB: 4YQX), or NARA1 (PDB: 5LQB) (FIG. 9A). Next, a close analysis of the CD25−, CD122−, and CD132-binding sites was performed. Epitope overlaps were assessed between the anti-IL-2 mAbs and the IL-2R subunits, quantified based on buried surface areas within IL-2cx and the quaternary IL-2R complex using the protein interfaces, surfaces and assemblies software tool. UFKA-20 overlaid strongly with the CD122-binding site of IL-2, resulting in an overlap calculated at 40%, whereas UFKA-20 interference with the CD25-binding site of IL-2 was rather mild and amounted to only 6.3% (FIG. 9B). There was no overlap evident of UFKA-20 with IL-2's CD132-binding site. Thus, via its CDR1-3 of the variable heavy (V.sub.H) and CDR1 and CDR3 of the variable light (V.sub.L) chain, UFKA-20 predominantly contacted the C- and B-helices of IL-2 thereby forming a ‘clamp’ around the C-helix (FIG. 9B). The IL-2 residues D84, N88 and V91, usually involved in CD122 interaction, were closely engaged with UFKA-20 (FIG. 9A). These interactions very likely abrogated IL-2 binding to CD122, as shown by surface plasmon resonance (SPR) measurements (FIG. 9C). Furthermore, UFKA-20's V.sub.H chain clashed to a minor but significant extent with CD25 through close contact with IL-2 residues E60, E61 and K64, which locate to the IL-2-CD25 interface (FIG. 9B). However, as measured by SPR, IL-2/UFKA-20cx efficiently bound to recombinant human CD25 in a dose-dependent manner (FIG. 9C). As a result of its multiple contacts, UFKA-20 associated with a high affinity with IL-2, resulting in a K.sub.d of about 10.sup.−9 M (Table 1). F5111 bound to IL-2 at a different angle than UFKA-20, and the F5111 and CD122 epitopes overlapped significantly (calculated at 48.5%), whereas for the CD132 epitope this overlap was very mild (2.5%) and there was no significant overlap with the CD25 epitope (0.85%) (FIG. 9A and B). JES6-1 interacted with mouse IL-2 in a manner very different from UFKA-20 and F5111. Compared to UFKA-20, JES6-1 bound to the opposite side of IL-2, mainly interfering with CD132 (18%), followed by CD25 (16%) and CD122 (8%) (FIG. 9A and B). As the quaternary mouse IL-2R complex is unavailable, the IL-2R overlaps of the mouse IL-2/JES6-1cx were calculated using the quaternary human IL-2R crystal. IL-2/NARA1 fully overlapped with CD25, thus ‘mimicking’ CD25's binding to IL-2. At close observation, NARA1 largely overlapped with the CD25-binding site of IL-2 (52.5%), whereas the CD122- and CD132-binding sites remained fully accessible (0%) (FIG. 9A and B). Consequently, on SPR, IL-2/NARA1cx very efficiently bound to recombinant human CD122 but not to CD25 (FIG. 9C).

[0328] To assess whether UFKA-20 functionally competed with both CD122− and CD25-binding sites of IL-2 as suggested by structural analysis, a competition assay was performed on HEK293T cells expressing different IL-2R subunits using a set concentration of IL-2.sub.Rhod and titrated concentrations of UFKA-20. IL-2.sub.Rhod binding to CD122+CD132 was diminished already at a 10:1 and 1:1 molar ratio of IL-2 to UFKA-20, thus confirming UFKA-20's functional interference with CD122 (FIG. 9D, middle panel). Conversely, IL-2.sub.Rhod binding to CD25 was unperturbed at molar ratios of IL-2 to UFKA-20 of 10:1 and 1:1, whereas five to 50-fold higher UFKA-20 concentrations were needed to compete with CD25-binding (9D, left panel); a similar pattern emerged with HEK293T cells expressing CD25+ CD122 +CD132 (FIG. 9D, right panel). The results of the UFKA20 epitope overlap and buried surface areas within IL-2cx and the quaternary IL-2R complex suggest binding was occurring at three sites of the hIL-2 sequence: [0329] Epitope A comprising H16, D20. [0330] Epitope B comprising Q57, E60, E61, L63, K64, E67, E68. [0331] Epitope C comprising L80, R81, R83, D84, 186, S87, N88, N90, V91, L94, E95, K97, T101, T102, M104. [0332] Epitope A and C overlap the region of IL-2 important for binding with CD122 and CD132 (targeted by F5111), whereas epitope B overlapping the CD25 binding site is uniquely targeted by UFKA-20 compared to existing antibody clones (FIGS. 9, 10 and 11) and is likely to be associated with enhanced effect generated by this clone compared to the >10000 alternatives screened.

Example 8

CDR Mutations Altering UFKA20 Binding to Specific hIL-2 Epitopes

[0333] UFKA-20 variants that contain specific amino acid substitutions in VH (SEQ ID NO 019) and VL (SEQ ID NO 020) chain were created to investigate the effect of weaker or stronger polar and non-polar interactions between specific CDR loops and the proposed epitopes on hIL-2 (Table 3 and 4). The 7 VH chain variants contained between 1-3 and 4 V.sub.L chain variants with 2-4 amino acid substitutions. Collectively, 12 different UFKA-20 variants, including the original UFKA-20 mAb, were expressed, purified and subsequently their affinity and in vivo activity (FIG. 12) was determined. The affinities of the variant antibodies ranged from 6.403×10.sup.−8 M to 1.856×10.sup.−10 M, 2+9 was the lowest and 5+9 the highest among variants tested. Most of the antibody variants bind in the 10.sup.−10 M range, clustering around the affinity of the unmodified original 1+6 (UFKA-20) antibody. C57BL/6 wildtype mice were administered a single dose of IL-2/anti-IL-2cx of the UFKA-20 variants. Frequencies of CD4.sup.+ CD25.sup.+ Foxp3.sup.+ cells were not significantly altered by antibodies comprising variants 5+9; 4+9; 4+10; 2+9 (FIG. 12B), indicating that these variants lacked optimal interactions with residues in the hIL-2 epitopes B (for example E61, Q57, R83) and C (L94, and E95) which were affected by the residue substitutions in these variant chains (Table 4), and that interaction with both is essential for high efficacy of hIL-2 complexes formed with the antibody according to the invention. Of note, epitope B is uniquely targeted by UFKA-20 derived antibodies compared to known clones with capacity to form CD25-targeted complexes (FIG. 10). Most of the tested antibodies, including 105+6, 105+9, 2+9, 103+6, and 103+9 were very similar to UFKA-20 (1+6) in terms of CD4.sup.+ CD25.sup.+ Foxp3.sup.+ T.sub.reg cell stimulation, showing the sequence is tolerant to certain amino acid changes in the CDR regions, provided the residues S56, M100, Y102, in humanized antibodies derived from SEQ ID NO 019, and K36, F56, S32, A33, and A100 of SEQ ID NO 020 maintain their optimal orientation to interact with IL-2 epitopes B and C.

[0334] A correlation could be observed between K.sub.D values and a capacity to increase the frequency of CD4.sup.+ CD25.sup.+ Foxp3.sup.+ T.sub.reg cells in the mediated by T.sub.reg cell stimulation (FIG. 12C). Excepting antibodies 2+9 and 4+10, the antibodies' activity clustered around the UFKA-20 (1+6) antibody, suggesting that the optimal affinity lies in the 10.sup.−10M range. However, the 5+9 antibody that bound IL-2 with the highest affinity had reduced in vivo activity with regards to CD4.sup.+ CD25.sup.+ Foxp3.sup.+ T.sub.reg cell stimulation, suggesting an upper affinity limit of 1.856.sup.−10.

Example 9

Fusion Proteins of IL-2 and UFKA-22 Antibody

[0335] Although CD25-biased immune complexes have excellent immunomodulatory potential, they are yet to gain approval for use inhibiting inflammatory responses in humans, as several aspects of their biology create problems which have impeded clinical development. Firstly, IL-2 antibody complexes formed by incubating IL-2 with anti-IL-2 antibodies at 37 degrees must be prepared immediately prior to administration to avoid degradation into separate components. This is inconvenient in a clinical setting, and can lead to small differences in activity between batches. In addition, the complexes may dissociate in vivo, producing soluble IL-2 with the potential to produce undesirable off-target signalling. To overcome these problems, a single-agent drug compound was developed to replace IL-2/UFKA-22cx therapy—a two component immunotherapy consisting of recombinant human IL-2 and the humanized CD25-biased anti-IL-2 antibody UFKA-22—with IL-2/UFKA-22 fusion proteins retaining optimal signalling through CD25, combined with improved stability (UFKA-22FP).

[0336] For the UFKA-22FP design, the IL-2 protein and the UFKA-22 antibody must be connected by a flexible linker facilitating optimal rates of not just IL-2-association, but importantly, dissociation from the IL-2-binding groove of the UFKA-22 antibody, such that IL-2 signalling through the dimeric IL-2R (CD122+CD132) is not impeded by the joined antibody structure. The crystal structure of the IL-2/UFKA-20cx (PDB: 6YE3) was analysed to determine the distance between N-terminus of the UFKA-20 variable heavy (V.sub.H) and variable light (V.sub.L) chain to the C-terminus of IL-2 with 32.2 Å and 43.5 Å, respectively (FIG. 12A and B). Subsequently, UFKA-22FPs joining an N-terminal IL-12 polypeptide (Uniprot P60568) to the UFKA-22 V.sub.H (SEQ ID NO 017), or VL chains (SEQ ID NO 018) with flexible glycine (G) serine (S) linkers consisting of (G.sub.4S).sub.n, (SEQ ID NO 023), with n number of repeats ranging from 3 to 6 were generated to test which linker offers optimal CD25-targeted signalling though IL-2R (FIGS. 13C and D, and 14). UFKA-22FP vH (G.sub.4S).sub.3, UFKA-22FP vH (G.sub.4S).sub.4, UFKA-22FP vH (G.sub.4S).sub.5 and UFKA-22FP vL (G.sub.4S).sub.6, were expressed downstream of a secretion signal (SEQ ID NO 027), and obtained from suspension cultures of HEK293 FreeStyle cells purified over a HiTrap® Protein G column. The UFKA-22FPs were tested for binding the CD25 mimobody—NARA1—in an ELISA. All four UFKA-22FP variants bound NARA1, indicating that the IL-2 domain of UFKA-22FPs was folded correctly. IL-2 bioactivity was measured in a cell proliferation assay using mouse CTLL-2 cells that express all three IL-2R subunits, an induction of STAT5 signalling assay in HEK-Blue IL-2 reporter cells that express human IL-2Rαβγ, and T.sub.reg stimulation was confirmed in human PBMCs. Once the activity and T.sub.reg-selectivity of fusion proteins was confirmed for all fusion proteins, and the most promising candidates were selected for further testing in mice.

[0337] C57BL/6 WT mice received daily injections for three days of IL-2/UFKA-22cx, UFKA-22FP vH (G.sub.4S).sub.5 and UFKA-22FP vL (G.sub.4S).sub.6 (comprising hIL-2 LC fusion SEQ ID NO 028), followed by flow cytometry analysis of CD4.sup.+ CD25.sup.+ Foxp3.sup.+ T.sub.reg, CD8.sup.+ CD44.sup.hi CD122.sup.+ memory T, and CD3.sup.− NK1.1.sup.+ CD122.sup.+ NK cells in spleens of treated animals. Because UFKA-22FP molecules constitute one UFKA-22 antibody and two IL-2 molecules, a 2:1 molar ratio of IL-2 to UFKA-22 antibody was used for the IL-2/UFKA-22cx formulation. Three injections of UFKA-22FP vH (G.sub.4S).sub.5 slightly increased the CD25.sup.+ Foxp3.sup.+ T.sub.reg cell compartment but the changes were not significant at the applied doses. In contrast, UFKA-22FP vL (G.sub.4S).sub.6 significantly increased CD25.sup.+ Foxp3.sup.+ T.sub.reg cell frequencies to 15.2±1.2% in the spleen at the 12 μg dose, and the 24 μg dose reached 20.6±1.4%. Treatment with UFKA-22FP vL (G.sub.4S).sub.6 induced a dose-dependent increase in Ki-67 expression in CD4.sup.+ CD25.sup.+ Foxp3.sup.+ T.sub.reg cells, with 42.0±5.5% and 64.1±3.7% of CD4.sup.+ CD25.sup.+ Foxp3.sup.+ T.sub.reg cells upregulating Ki-67 in response to 12 μg and 24 μg UFKA-22FP vL (G.sub.4S).sub.6, respectively. A significant increase in Ki-67.sup.+ CD4.sup.+ CD25.sup.+ Foxp3.sup.+ T.sub.reg cells was observed in mice that received 24 μg UFKA-22FP vL (G.sub.4S).sub.6 was very comparable to the 70.9±1.1% Ki-67.sup.+ observed in mice injected with 12 μg of IL-2/UFKA-22cx. No significant changes in the CD8.sup.+ T cell frequencies were observed, demonstrating the fusion protein did not create off-target effects on cytotoxic CD8.sup.+ T cells. UFKA-22FP activity was reduced compared to IL-2/UFKA-22cx, but UFKA-22FP vL (G.sub.4S).sub.6 approached a similar activity at a slightly higher dose (FIG. 14). In summary, a peptide linker length of 30 amino acids to the light chain is preferable. The linkage of IL-2 to the antibody heavy chain is feasible as well. These results suggest that a fusion protein with desirable manufacturing and pharmaceutical safety properties can achieve equivalent effect in vivo in comparison to a two-component IL-2cx, albeit at a moderately increased dosage.

Example 10

IL-2 Immunotherapy Expands cDCs in Mice and Humans

[0338] DCs are characterized by the absence of lineage (Lin) markers, have intermediate (int) or high (hi) CD11c, and can further be subdivided into CD11c.sup.int B220.sup.hi pDCs, CD11c.sup.hi major histocompatibility class II (MHC-II).sup.hi cDCs, CD11b.sup.low XCR1.sup.+ CD8α.sup.+ DNGR-1 (CLEC9A).sup.+ cDC1s, and CD11b.sup.hi XCR1.sup.−1 cDC2s (FIG. 15A). A short course of three injections of recombinant human IL-2 (IL-2; teceleukin) increased total counts of cDCs in spleens of adult wild-type (WT) mice (FIG. 15B). This expansion was due to active proliferation of cDCs, as evidenced by increased incorporation of the thymidine analogue bromodeoxyuridine (BrdU) into cDCs (FIG. 15C). To assess whether this IL-2 effect was caused by binding of IL-2 to CD25.sup.hi or CD122.sup.hi cells, CD25-biased IL-2/anti-IL-2 (5344) antibody complexes (IL-2/5344) and CD122-biased IL-2/anti-IL-2 (NARA1) antibody complexes (IL-2/NARA1) were tested (Letourneau E. M. PNAS 2010, 107:11906; Krieg C. PNAS 2010, 107:11906, Arenas-Ramirez N. Sci Transl Med 2016). Both IL-2/5344 and IL-2/NARA1 mouse IL-2/antibody complexes stimulated quantitatively comparable expansion and proliferation of splenic DCs as unbiased IL-2 (FIG. 15B, C): Treatment of mice with IL-2cx comprising an inflammatory, CD122-targeted IL-2 antibody induced upregulation of CD40, CD80, CD86, and MHC class I (MHC-I), but not MHC-II, on cDCs (FIG. 15D), indicating maturation of cDCs with increased potential of cross-presentation and co-stimulation for T cell activation.

[0339] Human CD11c.sup.+ MHC-II (HLA-DR).sup.+ DCs were examined within an investigator-initiated clinical trial (termed Charact-IL-2; NCT 03312335) using recombinant hIL-2 (aldesleukin) immunotherapy (FIG. 16A). Clinical trials testing aldesleukin have reported the proliferation of CD4.sup.+ and CD8.sup.+ T cells and NK cells upon aldesleukin immunotherapy (Klatzmann D. et al. Nat Rev Immunol 2015, 15:283; Humrich J. Y et al. Lancet Rheumatol 2019, 1:e44). However, by comparing Ki67.sup.+ DCs on day 0 (before) and day 5 (1 day after the last injection) of a 5-day course of daily aldesleukin, an increase of proliferating cDC1s and cDC2s was observed (FIG. 16B).

[0340] The parenteral administration of UFKA20 complexes (IL-2 bound to the CD25-biased antibody UFKA20) also expanded cDCs in the spleen of murine recipients (FIG. 17A), inducing proliferation of cDCs, as measured by BrdU incorporation (FIG. 17B). Treatment with a pharmaceutical composition comprising UFKA20 complexes downregulated MHC-II and CD80 (FIG. 17C and D), indicating complexes with CD25-targeted antibodies reduce the capacity of cDCs to present antigen and transmit co-stimulatory signals to T cells. Treatment with UFKA20 complexes further induced upregulation of the immune regulatory protein programmed cell death protein 1 (PD-1), programmed cell death ligand 1 (PD-L1) and PD-L2 expressed on cDCs. Immunoregulatory genes were upregulated in UFA20 complex treated mice (positive values on x-axis) including transforming growth factor beta-induced (Tgfbi), interleukin 1 receptor antagonist (II1 m), and TGF-beta activated kinase 1/MAP3K7 binding protein 1 (Tab1). Immunostimulatory genes including interleukin 6 signal transducer (II6st), lymphotoxin beta (Ltb), tumor necrosis factor (ligand) superfamily member 14 (Tnfsf14), and colony stimulating factor 1 (Csf1) were downregulated in UFKA20 complex treated mice (negative values on x-axis). Additionally, UFKA20 complex treated cDCs downregulated TNF receptor superfamily member 6 (Fas), suggesting prolonged survival (FIG. 17E). In summary, the presented data shows that a pharmaceutical compound comprising UFKA20 promotes proliferation of cDCs with a tolerogenic phenotype and immunoregulatory properties.

TABLE-US-00001 TABLE 1 Anti-IL-2 mAb Anti-IL-2 mAb clone ka (1/Ms) kd (1/s) KD (M) JES6-1 5.60 × 10.sup.−9 UFKA-20 2.57 × 10.sup.5 3.91 × 10.sup.−4 2.30 × 10.sup.−9 UFKA-22-00 4.66 × 10.sup.5 2.39 × 10.sup.−3 5.13 × 10.sup.−9 UFKA-22-02 4.12 × 10.sup.5 2.27 × 10.sup.−3 5.51 × 10.sup.−9 UFKA-22-07 4.86 × 10.sup.5 2.51 × 10.sup.−3 4.45 × 10.sup.−9 NARA1 3.00 × 10.sup.4 1.17 × 10.sup.−4 4.20 × 10.sup.−9 F5111 3.78 × 10.sup.6 7.11 × 10.sup.−3 1.88 × 10.sup.−9

TABLE-US-00002 TABLE 2 IL-2 binding properties of UFKA-22 variants Variants in bold were evaluated in vivo, in mice or rhesus macaques. SEQ SEQ ka kd KD mAb ID VH NO ID VL NO ID (1/Ms) (1/s) (M) UFKA-22-00 VH1 007 zVK1 015 4.66E+05 2.39E−03 5.13E−09 (UFKA-22) UFKA-22-01 VH2 008 VK1 015 4.75E+05 2.52E−03 5.31E−09 UFKA-22-02 VH3 009 VK1 015 4.12E+05 2.27E−03 5.51E−09 UFKA-22-03 VH4 010 VK1 015 5.90E+05 2.79E−03 4.73E−09 UFKA-22-04 VH5 011 VK1 015 4.79E+05 2.41E−03 5.03E−09 UFKA-22-05 VH6 012 VK1 015 4.52E+05 2.20E−03 4.86E−09 UFKA-22-06 VH7 013 VK1 015 5.67E+05 2.52E−03 4.45E−09 UFKA-22-07 VH1 007 VK2 016 4.86E+05 2.51E−03 5.17E−09 UFKA-22-08 VH2 008 VK2 016 5.15E+05 2.75E−03 5.33E−09 UFKA-22-09 VH3 009 VK2 016 4.66E+05 2.46E−03 5.27E−09 UFKA-22-10 VH4 010 VK2 016 5.48E+05 2.83E−03 5.16E−09 UFKA-22-11 VH5 011 VK2 016 5.17E+05 2.75E−03 5.31E−09 UFKA-22-12 VH6 012 VK2 016 4.74E+05 2.40E−03 5.05E−09 UFKA-22-13 VH7 013 VK2 016 6.07E+05 2.77E−03 4.56E−09 UFKA-22-14 VH8 014 VK2 016 5.60E+05 2.40E−03 4.30E−09 V.sub.H heavy variable chain, V.sub.L light variable chain

TABLE-US-00003 TABLE 3 VH (SEQ ID NO 019) and VL (SEQ ID NO 020) alterations in UFKA-20 variants Variant Variable chain Amino acid substitution 1 vH — 2 vH S56R-Y102K 3 vH S56R-M100L-Y102K 4 vH S56R-Y102H 5 vH S56R-M100L-Y102H 6 vL — 7 vL S32V-A33S-A100C 8 vL S32V-A33S-K36R-F56W-A100C 9 vL K36R-F56W 10 vL S32V-A33S-K36R-A100C 103 vH N103A 104 vH D104L 105 vH Y105A

TABLE-US-00004 TABLE 4 Role of amino acid substitutions in UFKA20 variants Amino acid substitution Purpose of substitution vH S56R Positively charged side chain to strengthen interaction with epitope B (e.g. E61) on hIL-2 vH M100L Hydrophobic side chain to optimize interaction with epitope C (e.g. L94) on hIL-2 vH Y102K Positively charged side chain to strengthen interaction with epitope B (e.g. E61, Q57) on hIL-2 vH Y102H Positively charged side chain to strengthen interaction with epitope B (e.g. E61, Q57) on hIL-2 vH N103A Hydrophobic side chain to reduce interaction with epitope B on hIL-2 vH D104L Hydrophobic side chain to reduce interaction with epitope B on hIL-2 vH Y105A Hydrophobic side chain to reduce interaction with epitope B on hIL-2 vL S32V Hydrophobic side chain to optimize interaction with epitope C on hIL-2 vL A33S Polar uncharged side chain to optimize interaction with epitope C on hIL-2 vL K36R Positively charged side chain to strengthen interaction with epitope C (e.g. E95) on hIL-2 vL F56W Hydrophobic side chain to optimize interaction with epitope C on hIL-2 vL A100C Side chain to optimize interaction with epitope B (e.g. R83) on hIL-2

TABLE-US-00005 SEQ ID NO NAME SEQUENCE 001 CDRH1 GFSFSNYAMS 002 CDRH2 LISGGGSYSYYPDSLKG 003 CDRH3 HMGYNDYLAWFAY 004 CDRL1 KSSQSLLNSANQKNYLA 005 CDRL2 FASTRES 006 CDRL3 QQYYSAPPWT 007 V.sub.H1 EVQLVESGGGLVKPGGSLRLSCAASGFSFSNYAMSWVRQAPGKGLEWVSLI SGGGSYSYYPDSLKGRFTISRDNAKNSLYLQMNSLRAEDTAVYYCARHMGY NDYLAWFAYWGQGTLVTVSS 008 V.sub.H2 EVQLVESGGGLVKPGGSLRLSCAASGFSFSNYAMSWVRQAPGKGLEWVSLI SGGGSYSYYPDSLKGRFTISRDSAKNSLYLQMNSLRAEDTAVYYCARHMGY NDYLAWFAYWGQGTLVTVSS 009 V.sub.H3 EVQLVESGGGLVKPGGSLRLSCAASGFSFSNYAMSWVRQAPGKGLEWVSLI SGGGSYSYYPDSLKGRFTISRDNAKNSLYLQMSSLRAEDTAVYYCARHMGY NDYLAWFAYWGQGTLVTVSS 010 V.sub.H4 EVQLVESGGGLVKPGGSLRLSCAASGFSFSNYAMSWVRQAPGKGLEWVSLI SGGGSYSYYPDSLKGRFTISRDSAKNSLYLQMSSLRAEDTAVYYCARHMGY NDYLAWFAYWGQGTLVTVSS 011 V.sub.H5 EVQLVESGGGLVKPGGSLRLSCAASGFSFSNYAMSWVRQAPGKGLEWVSLI SGGGSYSYYPDSLKGRFTISRDSAKNSLYLQMNSLRAEDTAMYYCARHMGY NDYLAWFAYWGQGTLVTVSS 012 V.sub.H6 EVQLVESGGGLVKPGGSLRLSCAASGFSFSNYAMSWVRQAPGKGLEWVSLI SGGGSYSYYPDSLKGRFTISRDNAKNSLYLQMSSLRAEDTAMYYCARHMGY NDYLAWFAYWGQGTLVTVSS 013 V.sub.H7 EVQLVESGGGLVKPGGSLRLSCAASGFSFSNYAMSWVRQAPGKGLEWVSLI SGGGSYSYYPDSLKGRFTISRDSAKNSLYLQMSSLRAEDTAMYYCARHMGY NDYLAWFAYWGQGTLVTVSS 014 V.sub.H8 EVQLVESGGGLVKPGGSLRLSCAASGFSFSNYAMSWVRQAPGKGLEWVSLI SGGGSYSYYPDSLKGRFTISRDNAKNSLYLQMNSLRAEDTAVYYCARHMGY NDYLAWFAYWGQGTLVTVSA 015 V.sub.L1 DIVMTQSPDSLAVSLGERATINCKSSQSLLNSANQKNYLAWYQQKPGQPPK LLIYFASTRESGVPDRFSGSGSGTDFTLTISSLQAEDVAVYYCQQYYSAPP WTFGGGTKVEIK 016 V.sub.L2 DIVMTQSPDSLAVSLGERATINCKSSQSLLNSANQKNYLAWYQQKPGQPPK LLIYFASTRESGVPDRFIGSGSGTDFTLTISSLQAEDVAVYYCQQYYSAPP WTFGGGTKVEIK 017 V.sub.H EVQLVESGGGLVKPGGSLRLSCAASGFSFSNYAMSWVRQAPGKGLEWVSLI SGGGSYSYYPDSLKGRFTISRDNAKNSLYLQMNSLRAEDTAVYYCARHMGY NDYLAWFAYWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKD YFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYSLSSVVTVPSSSLGTQTYI CNVNHKPSNTKVDKRVEPKSCDKTHTCPPCPAPELLGGPSVFLFPPKPKDT LMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYASTYR VVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLP PSREEMTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGS FFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPGK 018 V.sub.L DIVMTQSPDSLAVSLGERATINCKSSQSLLNSANQKNYLAWYQQKPGQPPK LLIYFASTRESGVPDRFSGSGSGTDFTLTISSLQAEDVAVYYCQQYYSAPP WTFGGGTKVEIKRTVAAPSVFIFPPSDEQLKSGTASVVCLLNNFYPREAKV QWKVDNALQSGNSQESVTEQDSKDSTYSLSSTLTLSKADYEKHKVYACEVT HQGLSSPVTKSFNRGEC 019 UFKA-20 EVMLVESGGGLVKPGGSLKLSCAASGFSFSNYAMSWVRQTPERRLEWVALI HC SGGGSYSYYPDSLKGRFTISRDSARNSLYLQMSSLRSEDTAMYYCARHMGY NDYLAWFAYWGQGTLVTVSSAKTTPPSVYPLAPGSAAQTNSMVTLGCLVKG YFPEPVTVTWNSGSLSSGVHTFPAVLQSDLYTLSSSVTVPSSTWPSQTVTC NVAHPASSTKVDKKIVPRDCGCKPCICTVPEVSSVFIFPPKPKDVLMISLT PKVTCVVVDISKDDPEVQFSWFVDDVEVHTAQTKPREEQINSTFRSVSELP ILHQDWLNGKEFKCRVNSAAFPAPIEKTISKTKGRPKAPQVYTIPPPKEQM AKDKVSLTCMITNFFPEDITVEWQWNGQPAENYKNTQPIMDTDGSYFVYSK LNVQKSNWEAGNTFTCSVLHEGLHNHHTEKSLSHSPGK 020 UFKA-20 DIVMTQSPSSLAVSVGQKVTMSCKSSQSLLNSANQKNYLAWYQQKPGQSPK LC LLIYFASTRESGVPDRFIGSGSGTDFTLNISSVQAEDLADYFCQQYYSAPP WTFGGGTKLEIKRADAAPTVSIFPPSSEQLTSGGASVVCFLNNFYPKDINV KWKIDGSERQNGVLNSWTDQDSKDSTYSMSSTLTLTKDEYERHNSYTCEAT HKTSTSPIVKSFNRNEC 021 Spacer GGGGS motif 022 Cypet CGTCTCGTGGTGGTGGTTCTGGTGGTGGTGGTTCTGTGACAAGG primer f 023 Cypet GGTGGTCTCGAGTTATTTGTACAGTTCGTCCATGCCG primer r 024 CD25 CTAGGAAGCTTATCTATGGATTCATACCTGCTG primer f 025 CD25 ACCAGAACCACCACCACCAGAACCACCACCACCGATTGTTCTTCTACTCTT primer r CCTCTG 026 linker GGGGSGGGGSGGGGSGGGGSGGGGSGGGGS 027 Signal MPLLLLLPLLWAGALA peptide 028 hIL2 PTSSSTKKTQLQLEHLLLDLQMILNGINNYKNPKLTRMLTFKFYMPKKATE fusion LKHLQCLEEELKPLEEVLNLAQSKNFHLRPRDLISNINVIVLELKGSETTF protein MCEYADETATIVEFLNRWITFSQSIISTLTGGGGSGGGGSGGGGSGGGGSG (G.sub.4S).sub.6 GGGSGGGGSDIVMTQSPDSLAVSLGERATINCKSSQSLLNSANQKNYLAWY QQKPGQPPKLLIYFASTRESGVPDRFSGSGSGTDFTLTISSLQAEDVAVYY CQQYYSAPPWTFGGGTKVEIKRTVAAPSVFIFPPSDEQLKSGTASVVCLLN NFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLSSTLTLSKADYEK HKVYACEVTHQGLSSPVTKSFNRGEC