WO2018115466A1 - Amanitin antibody conjugates - Google Patents

Abstract

The invention relates to a conjugate comprising (a) an amatoxin comprising (i) an amino acid 4 with a 6'-deoxy position; and (ii) an amino acid 8 with an S-deoxy position; (b) a BCMA-binding moiety comprising (i) the variable domains of humanized antibody J22.9-ISY, and (ii) a heavy chain constant region comprising a D265C mutation; and (c) a protease-cleavable linker linking said amatoxin and said target-binding moiety. The invention furthermore relates to a pharmaceutical composition comprising such conjugate, particularly for use in the treatment of multiple myeloma.

Description

AMANITIN ANTIBODY CONJUGATES

FIELD OF THE INVENTION

[001 ] The invention relates to a conjugate comprising (a) an amatoxin comprising (i) an amino acid 4 with a 6'-deoxy position; and (ii) an amino acid 8 with an S-deoxy position; (b) a BCMA-binding moiety comprising (i) the variable domains of humanized antibody J22.9-ISY, and (ii) a heavy chain constant region comprising a D265C mutation; and (c) a protease-cleavable linker linking said amatoxin and said target-binding moiety. The invention furthermore relates to a pharmaceutical composition comprising such conjugate, particularly for use in the treatment of multiple myeloma.

BACKGROUND OF THE INVENTION

[002] Multiple Myeloma (MM) is the second most prevalent hematopoietic malignancy characterized by the proliferation of a single clone of plasma cells derived from B cells in the bone marrow. MM ranks as the 15^th most common type of cancer in the US and is still considered incurable, with a median survival rate of about 30-60 months. Even though advances in the treatment of MM using conventional and novel therapeutics alone or in combination have resulted in improved response rate and thus prolonged median survival of many years, MM is still considered as incurable disease. MM represents about 0.8-1 % of all cancers worldwide and approximately 10% of all hematological cancers with an estimated global incidence of more than 100.000 new cases a year and a global mortality of more than 70.000 cases yearly (National Cancer Institute SEER database).

[003] The uncontrolled proliferation of plasma cells in MM results in the abnormal production of monoclonal immunoglobulins (also known as M-protein) that can lead to renal failure caused by immunoglobulin light chains or hyperviscosity from excessive amounts of M-proteins. In summary, clinically symptomatic MM is characterized by the presence of M-protein in serum and/or urine and end-organ damage, generally involving hypercalcemia, renal failure, anemia and bone marrow lesions (CRAB features) (Bird et al., 201 1 ; Rajkumar et al. 2014).

[004] Despite several treatment options, MM is an incurable disease and almost all patients will eventually develop resistant or refractory disease. Relapsed MM refers to progressive disease in which at least a partial response was previously achieved following first-line treatment of salvage therapy. Refractory MM indicates progressive disease when the patient is either unresponsive initially or following treatment within the last 60 days (Dimipoulos et al., 2015).

[005] Without any treatment, the median length of survival after diagnosis of a myeloma patient is about three years (Kyle and Rajkumar, 2004). There are now several classes of drugs used in the treatment of myeloma: immunomodulatory drugs (thalidomide, lenalidomide, pomalidomide), proteasome inhibitors (bortezomib, carfilzomib) and stem cell transplantation. Many of the notable drugs in these classes were approved over the past 10-15 years. Treatments using these novel drugs and high-dose chemotherapy in combination with autologous or allogenic hematopoietic stem-cell transplant (SCT) have been shown to improve median survival to five years, which makes this the current standard therapy. However, many patients suffer relapse or develop treatment resistance, leaving a need to develop safe and effective treatments that prolong the duration of remission and improve survival.

[006] About 30% of MM patients will develop renal insufficiency over the course of disease and 20% will present renal failure, caused by the accumulation and precipitation of light chains. Renal morbidity is a considerable burden for MM patients with incidence of renal replacement therapy for end-stage renal disease due to MM. Novel agents such as the proteasome inhibitor bortezomib can successfully restore renal function in a considerable proportion of patients and complications due to renal failure can be avoided when detected and treated early (Dimopoulos et al., 2008).

[007] Even though agents like proteasome inhibitors and/or immunomodulatory drugs have improved outcome for patients, these agents do not cure the disease. Thus, the need remains for new agents with novel modes of action. The knowledge of molecules expressed on the surface of myeloma cells resulted in a number of new monoclonal antibodies currently in development for new therapies. However, most of these molecules are not exclusively expressed on MM cells and their potential to treat MM needs to be demonstrated.

[008] Numerous drugs and therapies are currently in clinical phase: antibodies like daratumumab targeting CD38 and elotuzumab targeting SLAMF7 (Signaling Lymphocyte Activation Molecule Family Member 7), both used alone or in combination with other drugs; panabinostat, a histone deacetylase inhibitors (HDAC) to be used in combination therapy; chimeric antigen receptor engineered T (CAR-T) cells expressing an anti-BCMA single-chain variable fragment allowing specific targeting of MM cells.

[009] MM is characterized by the malignant proliferation of plasma cells, terminally differentiated B-cells which under normal circumstances are responsible for the mass production of immunoglobulins. This progression from plasma cells to malignant myeloma cell is associated with multiple genetic and oncogenic events including deregulation of cyclin D1 and c-Myc, and mutations like KRAS, BRAF, FGFR3 (fibroblast growth factor receptor 3) and TP53 (Shaffer et al., 2008; Kuehl and Bergsagel, 2012; Chesi and Bergsagel, 2015).

[0010] Despite these genetic alterations, the malignant plasma cell remains largely dependent on the bone marrow (BM) environment. Following binding of the MM cells in the BM, signalling cascades are activated including accessory growth factors and its ligands secreted by the BM accessory cells. Components released by these accessory cells play a critical role in MM to both promote disease and escape from immune surveillance.

[001 1 ] Antibody-drug conjugates (ADCs) aim to take advantage of the specificity of monoclonal antibodies to deliver potent cytotoxic drugs selectively to antigen-presenting tumor cells. [0012] The antibody should target a well-characterized antigen with high and specific expression at the tumor site and low or no expression on normal tissue, thus maximizing the efficacy of the ADC while limiting toxicity.

[0013] Members of the tumor necrosis factor receptor (TNFR) superfamily and their ligands play a critical role in controlling proliferation, differentiation and apoptosis of B-cells. In particular, B-cell activating factor (BAFF) ligand-receptor network play a central role in regulating B-cell maturation and differentiation into malignant plasma cells. The functionally related BAFF receptors BAFF-R, cyclophilin ligand interactor (TACI) and B-Cell maturation antigen (BCMA) are type III transmembrane proteins lacking a signal-peptide and containing cysteine-rich extracellular domains.

[0014] These receptors promote survival of B-cells at distinct stages of development and thus their expression patterns on B-cells differ depending on the type of B-cells and their stage of maturation and activation. BAFF-R expression is not detectable on B-cell precursors but is the dominant receptor expressed on mature B- cells. BAFF-R mediates the survival of B-cells, sustains the longevity of the germinal- center reaction and promotes immunoglobulin class-switch recombination. TACI is expressed on B-cells and activated T-cells and is important in class-switch recombination and supports T-cell dependent antibody response. BCMA, which is almost exclusively expressed on plasma cells but is absent from na^'fve and memory B-cells, promotes B-cell differentiation into plasma cells and thus promotes the maintenance of humoral immune response. Soluble BCMA is elevated in the serum of MM patients and the successful donor lymphocyte infusion is associated with the formation of antibodies against BCMA. The level of serum BCMA found among MM patients associates with clinical status and overall survival. (Belucci et al., 2005; Sanchez et al., 2012).

[0015] BAFF, APRIL (a proliferation-inducing ligand) and BCMA perform critical roles in decision-making process from B-cell to plasma and MM cell. As B- cells egress from the bone marrow, further maturation into marginal zone B-cells (MZ B) is dependent on BAFF. BAFF is also essential for the homeostasis of na^'fve recirculating B-cells and MZ B-cells. Downregulation of BAFF-R on plasma cells is coincident with the upregulation of BCMA, which can bind BAFF as well as APRIL. Transcription factor NFKB plays a crucial role in pathways regulation, B-cell proliferation, survival and differentiation. N FKB activity is modulated by the inhibitor of KB (IKB) proteins. Among these, ΙκΒα retains NFKB dimers inactive in the cytoplasm and, when degraded, releases bound N FKB dimers to translocate to the nucleus and drive gene expression. Regulation of IKB activity through degradation depends on serine phosphorylation by the IKB kinase (IKK) (Rickert et al., 201 1 ).

[0016] Various anti-BCMA antibodies have been generated in the past, including murine antibody J22.9 (WO 2014/068079) and humanized versions thereof (WO 2015/166073).

[0017] In order to obtain antibody-drug conjugates, bifunctional linkers with attachment sites for both the antibody and the drug are used to join the two components.

[0018] With respect to conjugation to the antibody, linker attachment strategies typically rely on cysteine or lysine residues on the antibody. Ideally, the linker must remain stable in systemic circulation to minimize adverse or toxic effects. Upon antigen recognition and binding, the resulting ADC receptor complex is internalized through receptor-mediated endocytosis. The unconjugated toxin should demonstrate high potency to enable efficient killing upon release from ADC. Release of the toxic payload should only occur inside the target cell and ADC must be stable in plasma. For the release of the toxin from the antibody inside the cell linkers can be either cleavable (either cleaved by lysosomal proteases, hydrolysed at low pH in the lysosomal compartment or released from disulfide bonds by intracellular gluthatione) or non-cleavable (depending on complete degradation of the antibody after internalization of the ADC).

[0019] Finally, with respect to the toxic payload of antibody-drug conjugates, two main classes of cytotoxic agents are currently used in clinical evaluations: drugs that disrupt microtubule assembly or compounds that bind to the minor groove of DNA causing double strand breaks. [0020] Amatoxins are cyclic peptides composed of 8 amino acids that are found in Amanita phalloides mushrooms (see Fig. 1). Amatoxins specifically inhibit the DNA-dependent RNA polymerase II of mammalian cells, and thereby also the transcription and protein biosynthesis of the affected cells. Inhibition of transcription in a cell causes stop of growth and proliferation. Though not covalently bound, the complex between amanitin and RNA-polymerase II is very tight (K_D = 3 nM). Dissociation of amanitin from the enzyme is a very slow process, thus making recovery of an affected cell unlikely. When the inhibition of transcription lasts too long, the cell will undergo programmed cell death (apoptosis).

[0021 ] The use of amatoxins as cytotoxic moieties for tumor therapy had already been explored in 1981 by coupling an anti-Thy 1 .2 antibody to a-amanitin using a linker attached to the indole ring of Trp (amino acid 4; see Fig. 1) via diazotation (Davis & Preston, Science 213 (1981 ) 1385-1388). Davis & Preston identified the site of attachment as position 7'. Morris & Venton demonstrated as well that substitution at position T results in a derivative, which maintains cytotoxic activity (Morris & Venton, Int. J. Peptide Protein Res. 21 (1983) 419-430).

[0022] Patent application EP 1 859 81 1 A1 (published November 28, 2007) described conjugates, in which the γ C-atom of amatoxin amino acid 1 of β-amanitin was directly coupled, i.e. without a linker structure, to albumin or to monoclonal antibody HEA125, OKT3, or PA-1 . Furthermore, the inhibitory effect of these conjugates on the proliferation of breast cancer cells (MCF-7), Burkitt's lymphoma cells (Raji) and T-lymphoma cells (Jurkat) was shown. The use of linkers was suggested, including linkers comprising elements such as amide, ester, ether, thioether, disulfide, urea, thiourea, hydrocarbon moieties and the like, but no such constructs were actually shown, and no more details, such as attachment sites on the amatoxins, were provided.

[0023] Patent applications WO 2010/1 15629 and WO 2010/1 15630 (both published October 14, 2010) describe conjugates, where antibodies, such as anti- EpCAM antibodies such as humanized antibody huHEA125, are coupled to amatoxins via (i) the γ C-atom of amatoxin amino acid 1 , (ii) the 6' C-atom of amatoxin amino acid 4, or (iii) via the δ C-atom of amatoxin amino acid 3, in each case either directly or via a linker between the antibody and the amatoxins. The suggested linkers comprise elements such as amide, ester, ether, thioether, disulfide, urea, thiourea, hydrocarbon moieties and the like. Furthermore, the inhibitory effects of these conjugates on the proliferation of breast cancer cells (cell line MCF-7), pancreatic carcinoma (cell line Capan-1 ), colon cancer (cell line Colo205), and cholangiocarcinoma (cell line OZ) were shown.

[0024] It is known that amatoxins are relatively non-toxic when coupled to large biomolecule carriers, such as antibody molecules, and that they exert their cytotoxic activity only after the biomolecule carrier is cleaved off. In light of the toxicity of amatoxins, particularly for liver cells, it is of outmost importance that amatoxin conjugates for targeted tumor therapy remain highly stable after administration in plasma, and that the release of the amatoxin occurs after internalization in the target cells. In this context, minor improvements of the conjugate stability may have drastic consequences for the therapeutic window and the safety of the amatoxin conjugates for therapeutic approaches.

[0025] Patent application WO 2016/142049 describes antibody-amanitin conjugates wherein the amanitin payload is attached to the antibody via specifically engineered cysteine residues in the antibody heavy chain constant region. Depending on the nature of the mutated residue, the replacement of the wild-type amino acid residue may influence the activity.

[0026] It has already been reported that amatoxin-based antibody conjugates directed at CD269 may offer advantages in the treatment of tumor patients (see Palfi et al.: "Preclinical evaluation of anti-CD269 antibody drug conjugates", European Journal of Cancer 69 (2016), S21 & 28^th EORTC-NCI-AACR Symposium on Molecular Targets and Cancer Therapeutics, Munich (Germany) November 29 - December 2, 2016; Palfi et al.: "CD269 - A promising target for amanitin based ADCs", Cancer Research 76 (No. 14 suppl) (2016) 2973 & 107^th Annual Meeting of the American Association for Cancer Research, New Orleans (U.S.A.) April 16 - April 20, 2016). However, no details were provided about the specific nature of the amatoxin derivative being used and of the specific linkage between that amatoxin derivative and the anti-CD269 antibody.

[0027] Thus, significant progress has already been made in the development of amatoxin-based conjugates for therapeutic uses. However, the present inventors have found that constructs based on a- and β-amatoxin were not fully stable under stress conditions in plasma and resulted in a substantial degree of cross-linked products (see Figure 2). This effect was completely unexpected, and the reasons for such instability and cross-linking were unclear. In Figure 2 it is additionally shown that the amount of cross-linked products formed from an amatoxin-based antibody conjugate can be reduced by the addition of cysteine, and in Figures 3 and 4 it is shown that a-amanitin reacts with cysteine. However, the stability of the conjugates comprising a highly toxic amatoxin is of utmost importance for the envisaged use as a therapeutic molecule for administration to human beings.

OBJECT OF THE INVENTION

[0028] Thus, there was still a great need for amatoxin-based antibody-drug conjugates having a high efficacy in target cells, with simultaneously having high stability. The solution to this problem, i.e. the identification of (a) an appropriate parental anti-BCMA antibody, (b) an appropriate strategy for attaching the toxic payload via linker to the antibody, (c) an appropriate linker and (d) an appropriate amatoxin variant was neither provided nor suggested by the prior art.

SUMMARY OF THE INVENTION

[0029] The present invention is based on the unexpected observation that a variant form of an amatoxin conjugated to an anti-BCMA antibody based on antibody J22.9, wherein the antibody and the amatoxin are linked by a cleavable linker shows an increased stability under stress conditions, particularly in human plasma, and an improved therapeutic index. [0030] Thus, in one aspect the present invention relates to a conjugate according to Formula I comprising (a) an amatoxin comprising (i) an amino acid 4 with a 6'-deoxy position; and (ii) an amino acid 8 with an S-deoxy position; (b) a BCMA-binding moiety comprising (i) the variable domains of the heavy chain according to SEQ ID NO: 1 and the light chain according to SEQ ID NO: 2 of antibody J22.9-ISY, and (ii) a heavy chain constant region comprising a D265C mutation; and (c) a protease-cleavable linker, wherein said BCMA-binding moiety is attached to said linker via the thiol group of the engineered cysteine residue at position 265 in the antibody heavy chain.

[0031 ] In a second aspect, the present invention relates to a pharmaceutical composition comprising the conjugate of the present invention.

[0032] In a third aspect, the present invention relates to the conjugate of the present invention, or the pharmaceutical composition of the present invention, for use in the treatment of cancer in a patient, particularly wherein the cancer is selected from the group consisting of multiple myeloma, diffuse large B-cell lymphoma (DLBCL), and chronic lymphocytic leukemia (CLL), particularly multiple myeloma. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Fig. 1 shows the structural formulae of different amatoxins. The numbers in bold type (1 to 8) designate the standard numbering of the eight amino acids forming the amatoxin. The standard designations of the atoms in amino acids 1 , 3 and 4 are also shown (Greek letters a to γ, Greek letters a to δ, and numbers from 1 ' to 7', respectively).

[0034] Fig. 2 shows the results of a stress testing experiment in an anti- amanitin Western blot. A trastuzumab-amanitin conjugate (Her-30.0643; lysine conjugation via 6^*-OH; stable linker) was incubated for 5 days at 37°C in PBS, pH 7.4, which led to extensive inter- and intrachain cross-linking; cross-linking of antibody chains could be reduced by addition of free cysteine.

[0035] Fig. 3 shows that a-amanitin shows a strong reactivity with cysteine in PBS buffer, pH 7.4. 1 mg/mL a -amanitin 10 mg/mL cysteine in PBS, pH 7.4 at 37 °C after 24 h, 48 h and 6 d RP-HPLC C18.

[0036] Fig. 4 shows that β-amanitin shows a strong reactivity with cysteine in PBS buffer, pH 7.4. 1 mg/mL β-amanitin 10 mg/mL cysteine in PBS, pH 7.4 at 37 °C after 24 h, 48 h and 6 d RP-HPLC C18.

[0037] Fig. 5 shows that a 6^'-deoxy variant at amino acid 4 („amanin") shows a reduced reactivity with cysteine. 1 mg/mL amanin 10 mg/mL cysteine in PBS, pH 7.4 at 37 °C after 24 h, 48 h and 6 d RP-HPLC C18.

[0038] Fig. 6 shows that a double deoxy variant HDP 30.2105 (6^'-deoxy at amino acid 4 and S-deoxy at amino acid 8; formula I with R³ = -OR⁵ and each R⁵ = H) shows complete absence of reactivity with cysteine. 1 mg/mL HDP 30.2105, 10 mg/mL cysteine in PBS, pH 7.4 at 37 °C after 24 h, 48 h and 6 d RP-HPLC C18; ^* impurity. [0039] Fig. 7 summarizes the results from Figures 3 to 6; x axis: reaction time in hours; y axis: amount of remaining amatoxin variant in %.

[0040] Fig. 8 shows alpha-amanitin derivative HDP 30.1699 with cleavable linker at AA4 - 6-OH moiety, alpha-amanitin derivative HDP 30.2060 with cleavable linker at AA1 γ-position and double deoxy amatoxin variant HDP 30.2115 with cleavable linker at AA1 γ-position.

[0041 ] Fig. 9 shows Western-Blot analysis of amatoxin derivatives HDP 30.1699, HDP 30.2060 and HDP 30.2115 conjugated to reference Thiomab antibody with D265C mutation after incubation at 37 °C in human plasma, mouse plasma and phosphate buffered saline (PBS) for 0, 4 and 10 days. Detection was done with a polyclonal anti-amanitin antibody from rabbit and an anti-rabbit antibody conjugated to horseradish peroxidase. HDP 30.1699 and HDP 30.2060 showed considerable cross-links and loss of the amatoxin moiety. Double deoxy amanitin variant HDP 30.2115 shows high stability and significantly reduced cross-links.

[0042] Fig. 10 shows cytotoxicity of amatoxin derivatives HDP 30.1699, HDP 30.2060 and HDP 30.2115 conjugated to reference Thiomab antibody with D265C mutation. Test items were incubated in human plasma at 37 °C for 0 an 4 days. Cytotoxicity assay were performed on SKBR-3 cells for 96 h. HDP 30.1699 and HDP 30.2060 based ADCs show remarkable loss of cytotoxicity after 4 days plasma stressing whereas deoxygenated derivative HDP 30.2115 shows still picomolar activity

[0043] Fig. 11 shows cytotoxicity of amatoxin derivatives HDP 30.1699, HDP 30.2060 and HDP 30.2115 conjugated to reference Thiomab antibody with D265C mutation. Test items were incubated in mouse plasma at 37 °C for 0 an 4 days. Cytotoxicity assay were performed on SKBR-3 cells for 96 h. HDP 30.1699 and HDP 30.2060 based ADCs show remarkable loss of cytotoxicity after plasma stressing whereas deoxygenated derivative HDP 30.2115 remains almost unchanged.

[0044] Fig. 12 shows cytotoxicity of amatoxin derivatives HDP 30.1699, HDP 30.2060 and HDP 30.2115 conjugated to reference Thiomab antibody with D265C mutation. Test items were incubated in PBS at 37 °C for 0 an 4 days. Cytotoxicity assay were performed on SKBR-3 cells for 96 h. All ADCs show adequate stability to non-enzymatic environment.

[0045] Fig. 13 compares the antitumoral activity of different amatoxin conjugates based on a tumor-targeting reference antibody (HDP_ref) -in a xenograft model - single dose experiment. Depending on linker and toxin structure significant differences in antitumoral activity have been observed. The deoxy-amanin variant HDP_ref-30.21 15 (6^'-deoxy at amino acid 4 and S-deoxy at amino acid 8) showed best antitumoral activity of all amatoxin ADCs, with a significantly better therapeutic index than corresponding cleavable linker ADC HDP_ref-30.1699 (lysine conjugation via 6^*- OH; S=O at amino acid 8).

[0046] Fig. 14 shows the Kaplan Meier survival analysis in a systemic tumor model - single dose experiment with reference antibodies. In brief, 2.5x10⁶ tumor cells in 200 μΙ_ PBS / mouse were inoculated intravenously on day 0. Therapy (single dose, iv) was initiated on day 3 post tumor cell inoculation. The deoxy-amanin variant HDP_ref-30.21 15 (6^'-deoxy at amino acid 4 and S-deoxy at amino acid 8) showed superior survival over HDP_ref-based conjugates with a-amanitin derivatives HDP 30.1699, HDP 30.0880 and HDP 30.0643 as well as with the corresponding MMAE- derivative.

[0047] Fig. 15 shows that J22.9-ISY and thiomab J22.9-ISY-D265C have comparable binding properties to BCMA-positive cells. The results of a flow cytometry-binding assays of antibody variants J22.9-ISY and J22.9-ISY-D265C revealed comparable binding properties to BCMA-expressing MM.1 S Luc and NCI- H929 cells.

[0048] Fig. 16 shows the QA results of the production of (A) J22.9-ISY- D265C-30.21 15, and (B) J22.9-ISY-D265C-30.1699.

[0049] Fig. 17 shows the results of stability experiments: J22.9-ISY-D265C- 30.21 15 (A) and J22.9-ISY-D265C-30.1699 (B) were incubated in human, cynomolgus, or mouse plasma or PBS for control for 0, 4, and 10 days at 37°C and analysed in reducing SDS-PAGE and subsequent anti-amanitin Western Blot.

[0050] Fig. 18 shows the results from experiments analysing plasma stability and cytotoxic potential. J22.9-ISY-D265C-30.21 15 (A) and J22.9-ISY-D265C- 30.1699 (B) were incubated in PBS for control, or human, cynomolgus or mouse plasma for 0, 4, and 10 days at 37°C and analysed for remaining cellular toxicity potential using BCMA-positive NCI-H929 cells.

[0051 ] Fig. 19 shows BCMA expression levels on different cell lines used for cytotoxicity assays. Western Blot analyses using antibodies against BCMA, CD138, RNA polymerase 2A (POLR2A), p53 and GAPDH (control). All cell lines express BCMA, except for CCRF-CEM which was used a BCMA-negative cell line. Highest BCMA levels are detected in NCI-H929 cells, lowest in RPMI-8226 cells.

[0052] Fig. 20 shows results from cellular cytotoxicity experiments: Viability of BCMA-positive (NCI-H929, MM.1 S, MM.1 S Luc, U266B1 , OPM-2) and BCMA- negative (CCRF-CEM) cells after incubation for 96 hours with J22.9-ISY-D265C- 30.1699 and J22.9-ISY-D265C-30.21 15 was monitored. Cytotoxicity was calculated as half maximal effective concentration (EC₅o).

[0053] Fig. 21 shows the antitumor activity in mouse NCI-H929 subcutaneous xenograft in single dose application: Female CB-17 Scid mice were subcutaneously inoculated with 5x10⁶ NCI-H929 cells. Once tumor volume reached 80 mm³, mice (8 animals per group) were treated intravenously with single doses (2 or 4 mg/kg) of J22.9-ISY-D265C-30.21 15 or J22.9-ISY-D265C-30.1699, or were treated with PBS as control.. Tumor volumes were monitored twice per week.

[0054] Fig. 22 shows the antitumor activity in mouse NCI-H929 subcutaneous xenograft in repeated dose application: Female CB-17 Scid mice were subcutaneously inoculated with 5x10⁶ NCI-H929 cells. Once tumor volume reached 80 mm³, mice (9 animals per group) were treated intravenously either once per week (1 x week), every two weeks (1 x 2weeks) or every 3 weeks (1 x 3weeks) with J22.9- ISY-D265C-30.21 15 (doses 0.25 mg/kg, 0.5 mg/kg or 1 mg/kg) or were treated with PBS as control. In addition, 2 mg/kg doses were applied twice per week (2x/week). Tumor volumes were monitored twice per week.

[0055] Fig. 23 shows the antitumor activity in disseminating MM.1 S Luc xenograft model: (A) Female SCID beige mice were intravenously injected with 1 x10⁷ MM.1 S Luc cells. When reaching a mean flux of 1 .5x10⁶ - 1 x10⁷ (10-14 days after implantation) 8-10 mice per group were treated with single doses (2 and 4 mg/kg) of J22.9-ISY-D265C-30.1699, J22.9-ISY-D265C-30.21 15 or PBS as control. Luciferase activity was monitored by non-invasive bio-imaging twice weekly, 10 minutes after administration of luciferin (10 μΙ/g body weight). (B) Individual tumor growth curves for each animal are depicted. Individual Growth curves are shown for groups treated with 2mg/kg (solid line) and 4mg/kg (dashed line) of J22.9-ISY-D265C-30.21 15. Data indicate that only one animal from each depicted group shows elevated tumor signals significantly above background level.

[0056] Fig. 24 shows the antitumor activity in a disseminating MM.1 S Luc xenograft model: dose reduction: (A) Female SCID beige mice were intravenously injected with 1 x10⁷ MM.1 S Luc cells. When reaching a mean flux of 1 .5x10⁶ - 1 x10⁷ (10-14 days after implantation) 8-10 mice per group were treated with single doses (0.1 , 0.3, 1 and 2 mg/kg) of J22.9-ISY-D265C-30.21 15 or PBS as control. Luciferase activity was monitored by non-invasive bio-imaging twice weekly, 10 minutes after administration of luciferin (10 μΙ/g body weight). (B) Individual tumor growth curves for each animal are depicted. Growth curves are shown for groups treated with 1 mg/kg (solid line) and 2mg/kg (dashed line) of J22.9-ISY-D265C-30.21 15. Data indicate that only one animal from each depicted group shows elevated tumor signals significantly above background level.

[0057] Fig. 25 shows the results from a tolerability study in cynomolgus monkeys: Groups of 3 animals were injected with J22.9-ISY-D265C-30.21 15 or J22.9-ISY-D265C-30.1699 in escalating doses (0.3, 1 and 3 mg/kg) followed by repeated dosing using 3 mg/kg. ALT: alanine transaminase; AST: aspartate transaminase; LDH: lactate dehydrogenase. [0058] Fig. 26 shows a comparison of the cytotoxic potential of J22.9-ISY- D265C-30.21 15 (·) with the monomethyl auristatin F derivative J22.9-ISY-MMAF (■) on MM.1 S Luc (left panel) and KMS-1 1 cells (right panel): (A) comparison at time point 96 h, EC₅₀ of J22.9-ISY-D265C-30.21 15: 5.575 10^"10 M on MM.1 S Luc cells; 1 .716 10^"9 M on KMS.-1 1 cells; EC₅₀ of J22.9-ISY-MMAF: 4.812 10^"10 M on MM.1 S Luc cells; 6.261 10^"10 M on KMS-1 1 cells; (B) time course 24 h to 120 h of cytotoxic potential on MM.1 S Luc cells (J22.9-ISY-D265C-30.21 15: ·; J22.9-ISY-MMAF: ■); (C) time course 24 h to 120 h of cytotoxic potential on KMS-1 1 cells (J22.9-ISY- D265C-30.21 15: ·; J22.9-ISY-MMAF:■).

[0059] Fig. 27 shows the results from a study of the efficacy of J22.9-ISY- D265C-30.21 15 in comparison with the monomethyl auristatin F derivative J22.9- ISY-MMAF and Carfilzomib in a disseminating Xenograft model (MM.1 S Luc).

DETAILED DESCRIPTION OF THE INVENTION

[0060] Before the present invention is described in detail below, it is to be understood that this invention is not limited to the particular methodology, protocols and reagents described herein as these may vary. It is also to be understood that the terminology used herein is for the purpose of describing particular embodiments only, and is not intended to limit the scope of the present invention which will be limited only by the appended claims. Unless defined otherwise, all technical and scientific terms used herein have the same meanings as commonly understood by one of ordinary skill in the art.

[0061 ] Particularly, the terms used herein are defined as described in "A multilingual glossary of biotechnological terms: (lUPAC Recommendations)", Leuenberger, H.G.W, Nagel, B. and Kolbl, H. eds. (1995), Helvetica Chimica Acta, CH-4010 Basel, Switzerland).

[0062] Throughout this specification and the claims which follow, unless the context requires otherwise, the word "comprise", and variations such as "comprises" and "comprising", will be understood to imply the inclusion of a stated integer, composition or step or group of integers or steps, while any additional integer, composition or step or group of integers, compositions or steps may optionally be present as well, including embodiments, where no additional integer, composition or step or group of integers, compositions or steps are present. In such latter embodiments, the term "comprising" is used coterminous with "consisting of.

[0063] Several documents are cited throughout the text of this specification. Each of the documents cited herein (including all patents, patent applications, scientific publications, manufacturer's specifications, instructions, GenBank Accession Number sequence submissions etc.), whether supra or infra, is hereby incorporated by reference in its entirety to the extent possible under the respective patent law.

[0064] The present invention will now be further described. In the following passages different aspects of the invention are defined in more detail. Each aspect so defined may be combined with any other aspect or aspects unless clearly indicated to the contrary. In particular, any feature indicated as being of particular relevance or advantageous may be combined with any other feature or features indicated as being of particular relevance or advantageous.

[0065] The present invention is based on a combination of different advantageous elements and features and in particular on the unexpected observation that a variant form of an amatoxin conjugated to an anti-BCMA antibody based on antibody J22.9, wherein the antibody and the amatoxin are linked by a cleavable linker shows an increased stability under stress conditions, particularly in human plasma, and an improved therapeutic index.

[0066] Thus, in one aspect the present invention relates to a conjugate comprising (a) an amatoxin comprising (i) an amino acid 4 with a 6'-deoxy position; and (ii) an amino acid 8 with an S-deoxy position; (b) a BCMA-binding moiety comprising the variable domains of the heavy chain according to SEQ ID NO: 1 and the light chain according to SEQ ID NO: 2 of antibody J22.9-ISY, and (c) a protease- cleavable linker, wherein said BCMA-binding moiety is attached to said linker via the thiol group of the cysteine residue at position 265 in the antibody heavy chain.

[0067] In a particular embodiment, the BCMA-binding moiety comprises a heavy chain constant region comprising a D265C mutation.

[0068] In a particular embodiment, said protease-cleavable linker is a self- immolative linker.

In a particular embodiment, the conjugate is the conjugate according to

Formula I

[0070] In the context of the present invention, the term "amatoxin" includes all cyclic peptides composed of 8 amino acids as isolated from the genus Amanita and described in Wieland, T. and Faulstich H. (1978), which comprise the specific positions according to (i) (i.e. where the indole moiety of the amino acid residue tryptophan has no oxygen-containing substituent at position 6', particularly where position 6' carries a hydrogen atom) and (ii) (i.e. in which the thioether sulfoxide moiety of naturally occurring amatoxins is replaced by a sulfide), and furthermore includes all chemical derivatives thereof; further all semisynthetic analogues thereof; further all synthetic analogues thereof built from building blocks according to the master structure of the natural compounds (cyclic, 8 amino acids), further all synthetic or semisynthetic analogues containing non-hydroxylated amino acids instead of the hydroxylated amino acids, further all synthetic or semisynthetic analogues, in each case wherein any such derivative or analogue carries at least the positions (i) and (ii) mentioned above and is functionally active by inhibiting mammalian RNA polymerase II.

[0071 ] Functionally, amatoxins are defined as peptides or depsipeptides that inhibit mammalian RNA polymerase II. Preferred amatoxins are those with a functional group (e.g. a carboxylic group or carboxylic acid derivative such as a carboxamide or hydroxamic acid, an amino group, a hydroxy group, a thiol or a thiol- capturing group) that can be reacted with linker molecules or target-binding moieties as defined above. Amatoxins which are particularly suitable for the conjugates of the present invention are di-deoxy variants of a-amanitin, β-amanitin, γ-amanitin, ε- amanitin, amanullin, or amanullinic acid, or mono-deoxy variants of amanin, amaninamide, γ-amanin, or γ-amaninamide as shown in Fig. 1 as well as salts, chemical derivatives, semisynthetic analogues, and synthetic analogues thereof.

[0072] An amatoxin variant comprising (i) an amino acid 4 with a 6'-deoxy position; and (ii) an amino acid 8 with an S-deoxy position was first mentioned in Zhou et al., ChemBioChem 16 (2015) 1420-1425, as one of four diastereomers being obtained in a total synthesis approach for amatoxins. However, the fact that such variants are more stable under stress conditions in plasma and result in a reduced degree of cross-linked products (as shown in Figure 2) was not publicly known at the effective filing date of the present application in December 2016 (see patent application WO 2017/149077 published on September 8, 2017).

[0073] In a particular embodiment, the conjugate of the present invention has a purity greater than 90%, particularly greater than 95%.

[0074] Monoclonal murine antibody J22.9 has been obtained using standard hybridoma technology from C57BL/6 mice immunized with purified human BCMA extracellular domain (residue 1 -54) N-terminally fused to glutathione S-transferase (GST). Due to instability of the hybridomas, variable regions of light and heavy chain were amplified and cloned upstream of the human kappa and lgG1 constant domain, respectively, resulting in chimeric J22.9-xi antibody (Oden et al., 2015).

[0075] Antibody J22.9-xi was humanized based on sequence alignments and data obtained from crystal structure in order to identify mutations that would potentially disrupt the binding to BCMA. Briefly, J22.9-xi Fab fragment was generated from full-length antibody by incubation with pepsin and combined with purified 54 amino acid residue BCMA extracellular domain. Isolated complexes were used for crystallization studies and J22.9-xi BCMA binding epitopes were analysed in detail (WO2014/068079, WO2015/166073, Oden et al., 2015, Marino et al. 2016).

[0076] Based on these analyses, various combinations of fully humanized gene cassettes were identified and expressed as full-length antibody. J22.9-H is the fully humanized version of J22.9-xi. J22.9-ISY and J22.9-FSY are fully humanized versions containing in addition mutations intended to remove potentially detrimental post-translational modification (PTM) motifs.

[0077] J22.9-xi binds to purified BCMA and also detects BCMA on human MM cell lines MM.1 S, NCI-H929, OPM-2 and RPMI-8226. No binding was detected on BCMA-negative cells. In addition, flow cytometry analyses revealed that cells from bone marrow of MM patients were detectable using J22.9-xi.The affinity of J22.9-xi to BCMA is very high with a mean Kd of 54 pM as determined using plasmon resonance (Oden et al., 2015).

[0078] BCMA activates nuclear factor κΒ (NFKB) pathways and triggers signals important for survival of MM and plasma cells through interaction with APRIL and/or BAFF. Affinity of J22.9 to BCMA was shown to be very high and exceeds that of April by 300- and of BAFF by 30.000-fold (Bossen and Schneider 2006). J22.9-xi efficiently blocks binding of APRIL and BAFF to BCMA. In addition, J22.9-xi interferes with APRIL-induced NFKB activation by blocking phosphorylation of ΙκΒ kinase (IKK) and subsequent ΙκΒα degradation, leading to reduced DNA-binding activity of NFKB. In summary, J22.9-xi interferes with APRIL-induced NFKB activation in BCMA-positive NCI-H929 cells (Oden et al., 2015). [0079] Cytotoxic activity of lgG1 antibodies is achieved through interaction of the lgG1 with Fey receptor (FcyR) on effector cells (e.g. natural killer cells) or with the C1 q protein of the complement cascade. This interaction is dependent on glycosylation of the antibody at position Asn297 in the heavy chain constant region. Glycans on lgG1 display some heterogeneity but core structure is usually a fucosylated bi-antennary structure with varying levels of sialic acid at the antennae. Numerous investigations have shown that glycosylation affects binding affinity and loss of glycosylation completely abrogates binding altogether. Absence of glycosylation disrupts the structural integrity of the Fc region which is required for optimal binding to the Fey receptor (for review see Hayes et al., 2014). However, absence of core fucose of IgG results in improved binding to Fey receptor and enhanced antibody-dependent cellular cytotoxicity (ADCC).

[0080] Binding of the antibody to effector cells or complement results in ADCC mediated by natural killer cells or complement-dependent cytotoxicity (CDC). J22.9-xi is able to induce strong ADCC and CDC on BCMA-positive MM.1 S cells when mixed with isolated Fc-bearing effector, peripheral blood mononuclear cells (PBMCs) from healthy donors.

[0081 ] Humanized and chimeric antibody variants bind to BCMA-expressing cell lines. No differences in binding characteristics were observed for J22.9-ISY and J22.9-FSY, whereas binding affinity of humanized variant J22.9-H was much lower.

[0082] In the context of the present invention, the term "purity" refers to the total amount of conjugates being present. A purity of greater than 90%, for example, means that in 1 mg of a composition comprising a conjugate of the present invention, there are more than 90%, i.e. more than 900 g, of such conjugate. The remaining part, i.e. the impurities may include unreacted starting material and other reactants, solvents, cleavage products and/or side products.

[0083] In a particular embodiment, a composition comprising a conjugate of the present invention comprises more than 100 mg of such conjugate. Thus, trace amount of a conjugate of the present invention that arguably may be present in complex preparations of conjugates of the prior art, e.g. from partial reduction of naturally occurring sulfoxides, are explicitly excluded.

[0084] As used herein, a first compound (e.g. an antibody) is considered to "specifically bind" to a second compound (e.g. an antigen, such as a target protein), if it has a dissociation constant K_D to said second compound of 100 μΜ or less, particularly 50 μΜ or less, particularly 30 μΜ or less, particularly 20 μΜ or less, particularly 10 μΜ or less, particularly 5 μΜ or less, more particularly 1 μΜ or less, more particularly 900 nM or less, more particularly 800 nM or less, more particularly 700 nM or less, more particularly 600 nM or less, more particularly 500 nM or less, more particularly 400 nM or less, more particularly 300 nM or less, more particularly 200 nM or less, even more particularly 100 nM or less, even more particularly 90 nM or less, even more particularly 80 nM or less, even more particularly 70 nM or less, even more particularly 60 nM or less, even more particularly 50 nM or less, even more particularly 40 nM or less, even more particularly 30 nM or less, even more particularly 20 nM or less, and even more particularly 10 nM or less.

[0085] The term "antibody or antigen binding fragment thereof, as used herein, refers to immunoglobulin molecules and immunologically active portions of immunoglobulin molecules, i.e. molecules that contain an antigen-binding site that immunospecifically binds the antigen BSMA. Thus, the term "antigen-binding fragments thereof refers to a fragment of an antibody comprising at least a functional antigen-binding domain. In a particular embodiment, functional antigen-binding domain comprises the variable domains of the heavy chain according to SEQ ID NO: 1 and the light chain according to SEQ ID NO: 2 of antibody J22.9-ISY. Also comprised are immunoglobulin-like proteins that are selected through techniques including, for example, phage display to specifically bind to the antigen BSMA. The immunoglobulin molecules of the invention can be of any type (e.g., IgG, IgE, IgM, IgD, IgA and IgY), class (e.g., lgG1 , lgG2, lgG3, lgG4, lgA1 and lgA2) or subclass of immunoglobulin molecule. "Antibodies and antigen-binding fragments thereof suitable for use in the present invention include, but are not limited to, polyclonal, monoclonal, monovalent, bispecific, heteroconjugate, multispecific, human, humanized (in particular CDR-grafted), deimmunized, or chimeric antibodies, single chain antibodies (e.g. scFv), Fab fragments, F(ab')2 fragments, fragments produced by a Fab expression library, diabodies or tetrabodies (Holliger P. et al., Proc Natl Acad Sci U S A. 90 (1993) 6444-8), nanobodies, anti-idiotypic (anti-Id) antibodies (including, e.g., anti-Id antibodies to antibodies of the invention), and epitope-binding fragments of any of the above.

[0086] In some embodiments the antigen-binding fragments are human antigen-binding antibody fragments of the present invention and include, but are not limited to, Fab, Fab' and F(ab')2, Fd, single-chain Fvs (scFv), single-chain antibodies, disulfide-linked Fvs (dsFv) and fragments comprising either a VL or VH domain. Antigen-binding antibody fragments, including single-chain antibodies, may comprise the variable domain(s) alone or in combination with the entirety or a portion of the following: hinge region, CL, CH1 , CH2, and CH3 domains. Also included in the invention are antigen-binding fragments also comprising any combination of variable domain(s) with a hinge region, CL, CH1 , CH2, and CH3 domains. In a particular embodiment, the antibody or antigen binding fragment thereof comprises a heavy chain constant region comprising a D265C mutation.

[0087] Antibodies usable in the invention may be from any animal origin including birds and mammals. Particularly, the antibodies are from human, rodent (e.g. mouse, rat, guinea pig, or rabbit), chicken, pig, sheep, goat, camel, cow, horse, donkey, cat, or dog origin. It is particularly preferred that the antibodies are of human or murine origin. As used herein, "human antibodies" include antibodies having the amino acid sequence of a human immunoglobulin and include antibodies isolated from human immunoglobulin libraries or from animals transgenic for one or more human immunoglobulin and that do not express endogenous immunoglobulins, as described for example in U.S. Patent No. 5,939,598 by Kucherlapati & Jakobovits.

[0088] In a second aspect, the present invention relates to a pharmaceutical composition comprising the conjugate of the present invention.

[0089] In a third aspect, the present invention relates to the conjugate of the present invention, or the pharmaceutical composition of the present invention, for use in the treatment of cancer in a patient, particularly wherein the cancer is selected from the group consisting of multiple myeloma, diffuse large B-cell lymphoma (DLBCL), and chronic lymphocytic leukemia (CLL), particularly multiple myeloma.

[0090] As used herein, "treat", "treating" or "treatment" of a disease or disorder means accomplishing one or more of the following: (a) reducing the severity of the disorder; (b) limiting or preventing development of symptoms characteristic of the disorder(s) being treated; (c) inhibiting worsening of symptoms characteristic of the disorder(s) being treated; (d) limiting or preventing recurrence of the disorder(s) in patients that have previously had the disorder(s); and (e) limiting or preventing recurrence of symptoms in patients that were previously symptomatic for the disorder(s).

[0091 ] As used herein, the treatment may comprise administering a conjugate or a pharmaceutical composition according to the present invention to a patient, wherein "administering" includes in vivo administration, as well as administration directly to tissue ex vivo, such as vein grafts.

[0092] In particular embodiments, a therapeutically effective amount of the conjugate of the present invention is used.

[0093] A "therapeutically effective amount" is an amount of a therapeutic agent sufficient to achieve the intended purpose. The effective amount of a given therapeutic agent will vary with factors such as the nature of the agent, the route of administration, the size and species of the animal to receive the therapeutic agent, and the purpose of the administration. The effective amount in each individual case may be determined empirically by a skilled artisan according to established methods in the art.

[0094] In another aspect the present invention relates to pharmaceutical composition comprising an amatoxin according to the present invention, or a conjugate of the present invention of an amatoxin with a target-binding moiety, and further comprising one or more pharmaceutically acceptable diluents, carriers, excipients, fillers, binders, lubricants, glidants, disintegrants, adsorbents; and/or preservatives.

[0095] "Pharmaceutically acceptable" means approved by a regulatory agency of the Federal or a state government or listed in the U.S. Pharmacopeia or other generally recognized pharmacopeia for use in animals, and more particularly in humans.

[0096] In particular embodiments, the pharmaceutical composition is used in the form of a systemically administered medicament. This includes parenterals, which comprise among others injectables and infusions. Injectables are formulated either in the form of ampoules or as so called ready-for-use injectables, e.g. ready-to- use syringes or single-use syringes and aside from this in puncturable flasks for multiple withdrawal. The administration of injectables can be in the form of subcutaneous (s.c), intramuscular (i.m.), intravenous (i.v.) or intracutaneous (i.e.) application. In particular, it is possible to produce the respectively suitable injection formulations as a suspension of crystals, solutions, nanoparticular or a colloid dispersed systems like, e.g. hydrosols.

[0097] Injectable formulations can further be produced as concentrates, which can be dissolved or dispersed with aqueous isotonic diluents. The infusion can also be prepared in form of isotonic solutions, fatty emulsions, liposomal formulations and micro-emulsions. Similar to injectables, infusion formulations can also be prepared in the form of concentrates for dilution. Injectable formulations can also be applied in the form of permanent infusions both in in-patient and ambulant therapy, e.g. by way of mini-pumps.

[0098] It is possible to add to parenteral drug formulations, for example, albumin, plasma, expander, surface-active substances, organic diluents, pH- influencing substances, complexing substances or polymeric substances, in particular as substances to influence the adsorption of the target-binding moiety toxin conjugates of the invention to proteins or polymers or they can also be added with the aim to reduce the adsorption of the target-binding moiety toxin conjugates of the invention to materials like injection instruments or packaging-materials, for example, plastic or glass.

[0099] The amatoxins of the present invention comprising a target-binding moiety can be bound to microcarriers or nanoparticles in parenterals like, for example, to finely dispersed particles based on poly(meth)acrylates, polylactates, polyglycolates, polyamino acids or polyether urethanes. Parenteral formulations can also be modified as depot preparations, e.g. based on the "multiple unit principle", if the target-binding moiety toxin conjugates of the invention are introduced in finely dispersed, dispersed and suspended form, respectively, or as a suspension of crystals in the medicament or based on the "single unit principle" if the target-binding moiety toxin conjugate of the invention is enclosed in a formulation, e.g. in a tablet or a rod which is subsequently implanted. These implants or depot medicaments in single unit and multiple unit formulations often consist of so called biodegradable polymers like e.g. polyesters of lactic acid and glycolic acid, polyether urethanes, polyamino acids, poly(meth)acrylates or polysaccharides.

[00100] Adjuvants and carriers added during the production of the pharmaceutical compositions of the present invention formulated as parenterals are particularly aqua sterilisata (sterilized water), pH value influencing substances like, e.g. organic or inorganic acids or bases as well as salts thereof, buffering substances for adjusting pH values, substances for isotonization like e.g. sodium chloride, sodium hydrogen carbonate, glucose and fructose, tensides and surfactants, respectively, and emulsifiers like, e.g. partial esters of fatty acids of polyoxyethylene sorbitans (for example, Tween^®) or, e.g. fatty acid esters of polyoxyethylenes (for example, Cremophor^®), fatty oils like, e.g. peanut oil, soybean oil or castor oil, synthetic esters of fatty acids like, e.g. ethyl oleate, isopropyl myristate and neutral oil (for example, Miglyol^®) as well as polymeric adjuvants like, e.g. gelatine, dextran, polyvinylpyrrolidone, additives which increase the solubility of organic solvents like, e.g. propylene glycol, ethanol, Ν,Ν-dimethylacetamide, propylene glycol or complex forming substances like, e.g. citrate and urea, preservatives like, e.g. benzoic acid hydroxypropyl ester and methyl ester, benzyl alcohol, antioxidants like e.g. sodium sulfite and stabilizers like e.g. EDTA. [00101 ] When formulating the pharmaceutical compositions of the present invention as suspensions in a preferred embodiment thickening agents to prevent the setting of the target-binding moiety toxin conjugates of the invention or, tensides and polyelectrolytes to assure the resuspendability of sediments and/or complex forming agents like, for example, EDTA are added. It is also possible to achieve complexes of the active ingredient with various polymers. Examples of such polymers are polyethylene glycol, polystyrene, carboxymethyl cellulose, Pluronics^® or polyethylene glycol sorbit fatty acid ester. The target-binding moiety toxin conjugates of the invention can also be incorporated in liquid formulations in the form of inclusion compounds e.g. with cyclodextrins. In particular embodiments dispersing agents can be added as further adjuvants. For the production of lyophilisates scaffolding agents like mannite, dextran, saccharose, human albumin, lactose, PVP or varieties of gelatine can be used.

EXAMPLES

[00102] In the following, the invention is explained in more detail by non-limiting examples:

Example 1 : Synthesis of Amatoxin-Linker HDP 30.2115

[00103] Linkage of antibody to toxin can occur either to cysteine or lysine residues, a specific tag or non-natural amino acids of the antibody. In order to obtain homogenous ADC products with specific DAR of 2, site-specific conjugation to cysteine in combination with genetic engineered Thiomabs is preferred. Key attributes of the linker include the requirement to be stable in plasma in order to prevent the uncontrolled release of the toxic payload into the circulation, and on the other hand the toxin needs to be released within the cell after internalization of the ADC upon target binding. Release of the toxin inside the target cell can occur either via cleavable or non-cleavable linkers. Cleavable linkers can be cleaved from the payload via a variety of mechanism including acidic degradation as consequence of lower pH inside the cell compared to circulation and protease cleavage by the protease cathepsin B or thiol-disulfide exchange attributed to the more reductive intracellular environment. Non-cleavable linkers require complete lysosomal proteolytic degradation generating a payload with a charged lysine or cysteine residue from the antibody.

[00104] The main advantage of non-cleavable compared to cleavable linkers is their increased plasma stability. While it is apparent that activities of ADCs containing non-cleavable linkers are less predictable, cleavable linkers raise the concerns of non-specific cytotoxicity.

[00105] However, in previous experiments, particularly in vitro experiments, with numerous different linkers, ADCs containing non-cleavable linkers have been shown to be less toxic than ADCs with a cleavable linker, and the majority of amanitin-based ADCs based on cleavable linkers were more effective compared to non-cleavable linker constructs. Thus, a cathepsin B-cleavable linker has been chosen for the drug substance.

[00106] The a-amanineamide-linker-toxin HDP 30.21 15 is synthesized by a multistep approach using solid phase peptide synthesis whereas the bicyclic octapeptide is initially assembled in a linear fashion. Starting point is a hydroxyproline resin immobilisation, followed by C-terminal coupling of dihydroxyisoleucin (HDP 30.0477). The remaining six amino acids are coupled by Fmoc strategy. The first cyclisation (right hand ring) occurs upon acidic cleavage from resin following the 'Savige-Fontana' mechanism. To make this happen, Tryptophan is incorporated in its oxidized form, ΉΡΓ (HDP 30.0079), as final amino acid. The second ring is formed by macrolactamisation using moderate to high dilution. The linker compound (HDP 30.2109) is synthesized in six linear steps and is finally introduced under standard coupling conditions, after deprotection. 1. Synthesis of synthetic dideoxy precursor molecule K

[00107] The synthesis of the dideoxy precursor molecule K is described in WO 2014/009025 in Example 5.5.

K

[00108] Compound K may be deprotected by treatment with 7 N methanolic NH₃ solution (3.0 ml) and stirring overnight.

2. Synthesis of synthetic dideoxy precursor HDP 30.2105

[00109] An alternative dideoxy precursor molecules comprising a -COOH group instead of the carboxamide group at amino acid 1 can be synthesized (HDP 30.1895) and deprotected to result in HDP 30.2105.

HDP 30.2105 Step l : 4-Hydroxy-pyrrolidine-1 ,2-dicarboxylic acid 2-allyl ester 1 -(9H- fluoren-9-ylmethyl) ester (HDP 30.0013)

[001 10] FmocHypOH (10.0 g, 28.3 mmol) was suspended in 100 ml 80% MeOH and Cs2CO3 (4.6 g, 14.1 mmol) was added. The suspension was stirred at 50°C for 30 minutes until complete dissolution. The reaction mixture was concentrated to dryness and resolved in 100 ml DMF. Allylbronnide (1 .6 ml, 3.6 g, 29.7 mmol) was added dropwise and the reaction was stirred over night at RT. DMF was distilled off and the residue dissolved in tert-butylmethyl ether. Precipitates were filtered and the clear solution was absorbed on Celite prior column chromatography. The compound was purified on 220 g Silicagel with an n-hexane/ethyl acetate gradient.

Yield: 1 1 .5 g, 100%

Step 2: Resin Loading (HDP 30.0400)

[001 1 1 ] HDP 30.0013 (5.0 g, 14.1 mmol), pyridinium 4-toluenesulfonate (1 .33 g, 5.3 mmol) were added to a suspension of 1 ,3-dihydro-2H-pyran-2-yl-methoxymethyl resin (5.0 g, 1 .0 mmol/g THP-resin) in 40 ml dichloroethane. The reaction was stirred at 80°C overnight. After cooling the resin was filtered and extensively washed with dichloroethane, dimethylformamide, acetonitrile, dichloromethane and tert- butylmethylether.

[001 12] Loading was 0.62 mmol/g (determined by UV-spectroscopy of the fluorene methyl group after deprotection).

Step 3: Solid Phase Precursor Synthesis (HDP 30.1894)

Resin pre-treatment:

[001 13] HDP 30.0400 (0.5 g, 0.31 mmol) was treated with N,N- dimethylbarbituric acid (483 mg, 3.1 mmol) and Pd(PPh3)4 (69 mg, 0.06 mmol). The resin was shaken over night at RT. Thereafter the resin was extensively washed with dichloromethane, N-methyl-2-pyrrolidone, acetonitrile, dichloromethane and tert- butylmethyl ether and dried under reduced pressure.

Coupling Procedure:

[001 14] All reactants and reagents were dissolved in dichloromethane/N- methyl-2-pyrrolidone containing 1 % Triton-X100 (Solvent A). [001 15] HDP 30.0477 (257 mg, 0.38 mmol) was dissolved in 3.0 ml Solvent A and treated with 3.0 ml of a 0.2 N solution PyBOP (333 mg, 0.63 mmol, 2.0 eq), 3.0 ml of a 0.2 N solution HOBt (130 mg, 0.63 mmol, 2.0 eq) and 439 μΙ DIEA (4.0 eq). The reaction was heated to 50°C for 8 minutes by microwave irradiation (20 W, CEM microwave reactor) and was washed with N-methyl-2-pyrrolidone after coupling.

Deprotection:

[001 16] Deprotection was performed by addition of 6.0 ml 20% piperidine in N- methyl-2-pyrrolidone at 50°C for 8 minutes. The resin was washed with N-methyl-2- pyrrolidone (Note: No deprotection after coupling of the final amino acid).

[001 17] All other amino acids were coupled following the above protocol, weightings are shown below:

0.63 mmol, 498 mg Fmoc Asp(OAII)OH

0.63 mmol, 738 mg Fmoc Cys(Tri)OH

0.63 mmol, 375 mg Fmoc GlyOH

0.63 mmol, 445 mg Fmoc lleOH

0.63 mmol, 375 mg Fmoc GlyOH

0.38 mmol, 242 mg N-Boc-HPIOH (HDP 30.0079)

[001 18] 4,5-Diacetoxy-2-amino-3-methyl-pentanoic acid tert-butyl ester; hydrochloride (HDP 30.0477) was synthesized as described in WO 2014/009025.

[001 19] N-Boc-HPIOH (HDP 30.0079) was prepared according to Zanotti, Giancarlo; Birr Christian; Wieland Theodor; International Journal of Peptide & Protein Research 18 (1981 ) 162-8. Step 4: HDP 30.1895

Elimination from resin and B-ring formation

[00120] The resin was shaken with 10 ml trifluoroacetic acid/dichloromethane 50:50 (v/v) plus 10% methanol for 30 min and finally eluted into a 50 ml flask. The resin was washed twice with methanol (10 ml each). The combined eluates were concentrated in vacuum and re-suspended in 2-4 ml methanol. The methanolic solution was dropped twice into 50 ml cold diethyl ether for peptide precipitation. After centrifugation the precipitate was washed with diethyl ether (2 times) and dried under reduced pressure. The white precipitate was solubilized in approx. 4-5 ml methanol (0.5 ml per 100 mg) and purified by preparative reverse phase column chromatography. Approximately 100 mg crude precipitate were purified per run. Fractions were analyzed by mass spectrometry, combined and methanol distilled off under reduced pressure. The aqueous phase was freeze dried.

Yield: 24.4 mg, 23.7 μηηοΙ

Mass spectrometry: [M+H]⁺, 1030.5

A-ring formation

[00121 ] The above freeze dried intermediate was dissolved in 25 ml dimethylformamide and treated with diphenylphosphorylazide (63 μΙ, 1 185 μιτιοΙ, 5 eq) and diisopropylethyl amine (201 μΙ, 1 185 μιτιοΙ, 5 eq). The reaction was stirred overnight (20 hours). Conversion was monitored by reverse phase chromatography and finally quenched with 100 μΙ water. The mixture was concentrated by reduced pressure and re-dissolved in 1 -2 ml methanol. Precipitation of the product was performed by dropwise addition to 20 ml diethyl ether. The precipitate was washed twice with diethyl ether and dried under reduced pressure. The next step was performed without further purification.

Mass spectrometry: [M+Na]⁺, 1034.6

Ester deprotection:

[00122] To the crude cyclisation product 2.5 ml dichloromethane, diethylbarbituric acid (22.3 mg, 1 18.5 μηηοΙ) and Pd(PPh₃) (27 mg, 23.7 μηηοΙ) were added. The reaction was stirred at RT overnight. The reaction can be monitored by RP-HPLC. After complete conversion, the mixture was added dropwise to 20 ml cooled diethyl ether and the precipitate washed twice with diethyl ether. After drying at reduced pressure the precipitate was dissolved in methanol (1 .0 ml) and purified by preparative reversed phase chromatography.

Yield: 15.0 mg

Mass spectrometry: [M+H]⁺, 972,3; [M+Na]⁺, 994.5

Step 5: HDP 30.2105

HDP 30.1895 HDP 30.2105

1012,1 1 887,97

C₄₆H_{6i 9}0₁₅S C39H53 9O13S [00123] HDP 30.1895 (15.0 mg, 15.3 μηηοΙ) was dissolved in 7 N methanolic NH₃ solution (3.0 ml) and stirred overnight. Conversion was checked by mass spectrometry. After complete conversion the reaction was concentrated in vacuum, suspended in 80% tert-butanol and lyophilized. Product was purified by preparative HPLC.

Yield: 12.1 mg

Mass spectrometry: [M+H]⁺, 888,0; [M+Na]⁺, 910.2

3. Synthesis of synthetic dideoxy precursor HDP 30.2115

[00124] A dideoxy precursor molecule comprising a thiol reactive group with cleavable linker can be synthesized from example 2 product in 7 steps as follows:

Step 1 : Fmoc-Val-OSu (HDP 30.1343)

HO

L-Valine

339,39 Fmoc-Val-OSu (HDP 30.1343)

C₂₀H₂₁NO₄ 436,47

C₂₄H₂₄N₂0₆ [00125] This compound is prepared according to R. A. Firestone et al, US 6,214,345. Fmoc-Val-OH (20.24 g; 59.64 mmol) and N-hydroxysuccinimide (6.86 g = 1 .0 eq.) in tetrahydrofuran (200 ml) at 0°C were treated with Ν,Ν'- dicyclohexylcarbodiimide (12.30 g; 1 .0 eq.). The mixture was stirred at RT under argon atmosphere for 6 h and then the solid dicyclohexyl urea (DCU) by-product was filtered off and washed with THF and the solvent was removed by rotavap. The residue was dissolved in 300 ml dichloromethane, cooled in an ice bath for 1 h and filtered again to remove additional DCU. The dichloromethane was evaporated and the solid foam (26.51 g) was used in the next step without further purification.

Step 2: Fmoc-Val-Ala-OH (HDP 30.1414)

Fmoc-Val-OSu (HDP 30.1343) Fmoc-Val-Ala-OH (HDP 30.1414)

436,47 410,47

C₂₄H₂₄N₂0₆ C₂₃H₂₆N₂0₅

[00126] Step 2 product is prepared in analogy to P.W. Howard et al. US 201 1/0256157. A solution of L-alanine (5.58 g; 1 .05 eq.) and sodium hydrogen carbonate (5.51 g; 1 .1 eq.) in 150 ml water was prepared and added to a solution of HDP 30.1343 (26.51 g; max. 59.6 mmol) in 225 ml tetrahydrofuran. The mixture was stirred for 50 h at RT. After consumption of starting material the solution was partitioned between 240 ml of 0.2 M citric acid and 200 ml of ethyl acetate. The aqueous layer was separated and extracted with ethyl acetate (3 x 200 ml). The combined organic layers were washed with water and brine (300 ml each) dried (MgSO₄) and the solvent was evaporated to approx. 200 ml. Pure product precipitated at this time and was filtered off. The mother liquor was evaporated to dryness and the residue was stirred 1 h with 100 ml MTBE to result additional crystalline material. The two crops of product were combined to 18.01 g (74%) white powder, (m.p.: 203-207°C) [M+Na]⁺ found: 410.94; calc: 41 1 .19 (C23H27N2O5) [M+Na]⁺ found: 433.14; calc: 433.17 (C23H27N2O5)

[2M+H]⁺ found: 842.70; calc: 843.36 (C₄₆H₅2N₄NaOio)

Step 3: Fmoc-Val-Ala-PAB-NHBoc (HDP 30.1713)

N-(tert.-Butoxycarbonyl)- HDP 30.1 1414 HDP 30.1713

4-aminobenzylamine 410,47 614,75

222,29 C₂₃H₂₆N₂0₅ C₃₅H₄₂N₄0₆

C₁₂H₁₈N₂0₂

[00127] Step 2 product HDP 30.1414 (1 .76 g; 4.28 mmol) and 4-[(N- Boc)aminomethyl]aniline (1 .00 g; 1 .05 eq.) were dissolved in 26 ml abs. tetrahydrofuran. 2-Ethoxy-N-(ethoxycarbonyl)-1 ,2-dihydroquinoline (EEDQ 1 .1 1 g; 1 .05 eq.) was added and the mixture was stirred at RT, protected from light. With ongoing reaction a gelatinous matter is formed from the initially clear solution. After 40 h the reaction mixture was diluted with 25 ml of tert-butylmethyl ether (MTBE) and stirred for 1 h. Subsequently the precipitation is filtered off with suction, washed with MTBE and dried in vacuo to 2.30 g (85% yield) of a white solid.

1H NMR (500 MHz, DMSO-d6) δ 9.87 (s, 1 H), 8.1 1 (d, J = 7.1 Hz, 1 H), 7.88 (d, J = 7.5 Hz, 2H), 7.74 (q, J = 8.4, 7.9 Hz, 2H), 7.51 (d, J = 8.2 Hz, 2H), 7.45 - 7.23 (m, 7H), 7.17 (d, J = 8.3 Hz, 2H), 4.44 (p, J = 7.0 Hz, 1 H), 4.36 - 4.17 (m, 3H), 3.96 - 3.89 (m, 1 H), 2.01 (hept, J = 6.9 Hz, 1 H), 1 .39 (s, 9H), 1 .31 (d, J = 7.1 Hz, 3H), 0.90 (d, J = 6.8 Hz, 3H), 0.87 (d, J = 6.8 Hz, 3H).

¹³C NMR (126 MHz, DMSO-d6) δ 170.84, 170.76, 156.04, 155.63, 143.77, 143.69, 140.60, 137.41 , 134.99, 127.50, 127.26, 126.93, 125.22, 1 19.95, 1 18.97, 77.60, 65.62, 59.95, 48.86, 46.62, 42.93, 30.28, 28.16, 19.06, 18.10, 18.03. Step 4: H-Val-Ala-PAB-NHBoc (HDP 30.1747)

HDP 30.1713 HDP 30.1747

[00128] Step 3 compound HDP 30.1713 (1 .230 g, 2.00 mmol) was placed in a 100 ml flask and dissolved in 40 ml dimethylformamide (DMF). Diethyl amine (7.5 ml) was added and the mixture was stirred at RT. The reaction was monitored by TLC (chloroform/methanol/HOAc 90:8:2). After consumption of starting material (30 min) the volatiles were evaporated and the residue was co-evaporated with 40 ml fresh DMF to remove traces of diethyl amine. The crude product was used without further purification for the next step.

MS (ESI+) [MH]⁺ found: 393.26 calc.: 393.25 (C₂oH₃₃N₄O₄)

[M+Na]⁺ found: 415.35 calc: 415.23 (C₂₀H₃2N₄NaO₄)

[2M+H]⁺ found: 785.37 calc: 785.49 (C^HesNsOs)

Step 5: BMP-Val-Ala-PAB-NHBoc (HDP 30.2108)

392,50 543,63 C₂₀H₃₂N₄O₄ C₂₇H₃₇N₅0₇

[00129] Crude step 4 product HDP 30.1747 (max 2.00 mmol) was dissolved in 40 ml DMF, 3-(maleimido)propionic acid N-hydroxysuccinimide ester (BMPS 532 mg; 1 .0 eq.) and N-ethyldiisopropylamine (510 μΙ, 1 .5 eq.) were added and the mixture was stirred 3 h at RT After consumption of starting material HDP 30.1747 (TLC: chloroform/ methanol/HOAc 90:8:2) the volatiles were evaporated and the residue is stirred with 50 ml MTBE until a fine suspension was formed (1 h). The precipitate was filtered off with suction, washed with MTBE and dried. The crude product (1 .10 g) was dissolved in 20 ml dichloromethane/methanol 1 :1 , kieselgur (15 g) was added and the solvents were stripped off. The solid material was placed on top of an 80 g silica gel column and eluted with a linear gradient of 0-10% methanol in dichloromethane. Product fractions were combined and evaporated to 793 mg (73% over two steps) amorphous solid.

MS (ESI⁺) [M+Na]⁺ found: 566.24; calc: 566.26

¹H NMR (500 MHz, DMSO-d₆) δ 9.75 (s, 1 H), 8.09 (d, J = 7.1 Hz, 1 H), 7.98 (d, J = 8.4 Hz, 1 H), 7.52 (d, J = 8.6 Hz, 2H), 7.29 - 7.23 (m, 1 H), 7.16 (d, J = 8.5 Hz, 2H), 6.98 (s, 2H), 4.39 (p, J = 7.1 Hz, 1 H), 4.13 (dd, J = 8.4, 6.7 Hz, 1 H), 4.06 (d, J = 6.1 Hz, 2H), 3.67 - 3.56 (m, 2H), 2.49 - 2.41 (m, 2H), 1 .96 (h, J = 6.8 Hz, 1 H), 1 .39 (s, 9H), 1 .30 (d, J = 7.1 Hz, 3H), 0.86 (d, J = 6.8 Hz, 3H), 0.82 (d, J = 6.8 Hz, 3H).

1³C NMR (126 MHz, DMSO-d₆) δ 170.80, 170.63, 170.60, 169.72, 155.65, 137.45, 134.94, 134.44, 127.26, 1 18.95, 77.62, 57.71 , 48.92, 42.95, 33.96, 33.64, 30.17, 28.17, 19.02, 18.06, 17.82.

Step 6: BMP-Val-Ala-PAB-NH₂ (HDP 30.2109)

HDP 30.2108 HDP 30.2109

543,63 443,51

C₂₇H₃₇N₅0₇ C ₅0₅

TFA salt: 557,53

[00130] Step 5 product HDP 30.2108 (400 mg, 736 μηηοΙ) was dissolved in 4,000 μΙ trifluoroacetic acid and stirred for 2 min. Subsequently the volatiles were evaporated at RT and the remainders were co-evaporated twice with 4,000 μΙ toluene. The residue was dissolved in 5,000 μΙ 1 ,4-dioxane/water 4:1 , solidified in liquid nitrogen and freeze-dried: 410 mg (quant.) colorless powder.

MS (ESI+) [M+Na]⁺ found: 415.35; calc: 466.21

[2M+H]⁺ found: 887.13; calc: 887.44 (C₄₄H₅9Ni₀Oio) ¹H NMR (500 MHz, DMSO-d₆) δ 9.89 (s, 1 H), 8.13 (d, J = 6.9 Hz, 1 H), 7.99 (d, J = 8.2 Hz, 1 H), 7.66 - 7.60 (m, 2H), 7.41 - 7.34 (m, 2H), 6.98 (s, 2H), 4.39 (p, J = 7.1 Hz, 1 H), 4.1 1 (dd, J = 8.2, 6.6 Hz, 1 H), 3.97 (q, J = 5.6 Hz, 2H), 3.69 - 3.58 (m, 2H), 2.49 - 2.40 (m, 2H), 1 .96 (h, J = 6.8 Hz, 1 H), 1 .32 (d, J = 7.1 Hz, 3H), 0.86 (d, J = 6.8 Hz, 3H), 0.83 (d, J = 6.7 Hz, 3H).

¹³C NMR (126 MHz, DMSO-d₆) δ 171 .24, 170.78, 170.72, 169.85, 158.12 (q, J = 33.2 Hz, TFA), 158.25, 157.99, 157.73, 139.19, 134.53, 129.45, 128.52, 1 19.02, 1 16.57 (q, J = 296.7 Hz, TFA), 57.78, 49.08, 41 .90, 34.00, 33.68, 30.21 , 19.07, 18.16, 17.76.

Step 6: HDP 30.2115

[00131 ] HDP 30.2105 (15.0 mg, 16.5 μηηοΙ) were treated with 429 μΙ of a 0.1 M solution of HDP 30.2109 (25.2 μΓηοΙ, 1 .5 eq), 492 μΙ of 0.1 M TBTU (25.2 μΓηοΙ, 1 .5 eq) and 492 μΙ of 0.2 M DIEA (49.1 μηηοΙ, 3.0 eq) at RT. The reaction was monitored by RP-HPLC. After completion the reaction was quenched with 100 μΙ H₂O stirred for 15 minutes and injected onto a preparative RP-HPLC.

Yield: 12.2 mg, 56%

Mass spectrometry: 1313.2 [M+H]⁺, 1335.5 [M+Na]⁺ Example 2: Properties of constructs based on HDP 30.2105

[00132] Various experiments were performed to compare the properties of constructs based on di-deoxy amanitin derivative HDP 30.2105 with those of other amanitin variants. The results of these experiments are shown in Figures 2 to 14.

Example 3: Generation and expression of Thiomab antibody J22.9-ISY-D265C

[00133] Conjugation of the antibody with linker and toxin can occur to lysine or cysteine residues, resulting in highly variable drug-antibody ratio (DAR). Since potency and toxicity is strongly influenced by DAR, homogeneity and comparability of the ADC with predictable DAR is favourable. Antibodies with engineered reactive cysteine residues (thiomabs) allow for site-specific conjugation and therefore, amino acid aspartic acid at position 265 has been exchanged to cysteine (D265C). Thus, the resulting antibody variant contains two introduced cysteines at each chain of the Fc region, which serves as coupling site for the toxin-linker compound and allows production of ADCs with DAR=2.

[00134] Sequence modifications in the Fc-region of the antibody can have dramatic influence in linking antibody-mediated immune responses with cellular effector functions, since residues in the Fc-region are responsible for interaction with IgG Fey receptor (FcyR). Interactions of IgG with the FcyR play crucial role in cellular effector functions including release of inflammatory mediators, endocytosis of immune complexes, antibody-dependent cellular cytotoxicity (ADCC) and complement-dependent cytotoxicity (CDC). Preserving these effector functions in an ADC might contribute to the anti-tumor activity of the ADC. However, this assumption might prove to be incorrect as these effector functions are generally weak and, importantly, might compete with ADC internalization, reduce or abolish their specific toxin-related activity and increase off-target toxicity due to ADC uptake in antigen- negative cells via Fc-receptor. Several ADCs currently in pre-clinical and clinical development are based on lgG2 and lgG4 which are very inefficient in these effector functions (Trail, 2013; Peters and Brown, 2015). [00135] Detailed mapping of binding sites on human lgG1 to FcyRI, FcyRII, FcyRIII and to neonatal Fc receptor (FcRn) revealed a crucial role of aspartic acid at position 265. Replacement of aspartic acid in lgG2 and lgG1 completely abolished interaction with FcyR (Baudino et al., 2008; Shields et al., 2001 ).

[00136] The efficiency of antibody-induced effector functions in Trastuzumab (Trademark Herceptin^®) and corresponding thiomab antibodies with amino acid exchanges at position 265 (D265C) or 1 18 (T1 18C) for control were tested on a macrophage cell line THP-1 . This monocytic cell line expresses multiple FcyRs and allows for testing antibody-mediated phagocytosis (Ackerman et al. 201 1 ). Cytotoxicity on THP-1 cells was reduced by at least two orders of magnitude with thiomab D265C compared to thiomab A1 18C, containing a cysteine mutation at position A1 18 not reducing the effector function, and non-engineered antibody. This effect of reduced cytotoxicity with thiomab engineered at amino acid 265 is not due to conjugation of the toxin to the engineered cysteine, since, in contrast to T-D265C- 30.1699 containing a cleavable linker conjugated to the engineered cysteine, in the case of T-D265C-30.0643 a stable linker is conjugated to lysine residues of the antibody. Thus, antibody-dependent effector functions were surprisingly highly reduced due to D265C exchange in thiomabs (see Table 1 ).

Table 1 :

[00137] Results from characterization of the chimeric and sequence-optimized variants of J22.9 revealed improved binding properties of the sequence optimized variants J22.9-ISY and J22.9-FSY to BCMA compared to J22.9-H, with no obvious difference for ISY- or FSY-variant. Since use of plasmids expressing J22.9-ISY antibody variant resulted in improved titers when transiently expressed in cell culture, variant J22.9-ISY was chosen for conjugation and further evaluation.

[00138] The nucleic acid coding sequence for the heavy chain of monoclonal antibody J22.9-ISY-D265C was obtained from the heavy chain sequence of humanized antibody J22.9-ISY (see WO 2014/068079 and WO 2015/166073) as described in WO 2016/142049. In addition to the exchange D265C, J22.9-ISY- D265C additionally contains a mutation R214K (see SEQ ID NO: 1 ).

[00139] In order to produce amanitin-conjugate J22.9-ISY-D265C-30.21 15, antibody J22.9-ISY-D265C was transiently expressed in Expi293F™ cells co- transfected with plasmids for the heavy and light chain. In an additional approach, antibody J22.9-ISY-D265C was produced in CHO cells stably transfected with plasmids expressing heavy and light chain. The antibody was purified from the cell culture supernatant using protein A chromatography followed by gel filtration. Antibody produced transiently from Expi293F™ or from stable CHO cells are comparable conjugation to payload, in binding to BCMA and cytotoxicity on BCMA- expressing cells.

[00140] In order to test if amino acid exchange in Thiomab J22.9-ISY-D265C did not interfere with binding to BCMA, binding of J22.9-ISY and J22.9-ISY-D265C to BCMA-expressing cells was compared. The binding property of the engineered thiomab J22.9-ISY-D265C to BCMA-positive NCI-H929 and MM.1 S-Luc cells was identical to J22.9-ISY antibody (see Figure 15).

Example 4: Synthesis of conjugate J22.9-ISY-D265C-30.2115

[00141 ] Conjugation of HDP 30.2115 to 10 mg J22.9-ISY-D265C

[00142] 10 mg Thiomab J22.9-ISY-D265C in PBS buffer will be used for conjugation to HDP 30.21 15.

Adjust antibody solution to 1 mM EDTA:

2 ml antibody solution (10.0 mg) + 20 μΙ 100 mM EDTA, pH 8.0 Amount antibody: 10 mg = 6.8 x 10 mol

[00143] Uncapping of cysteines by reaction of antibody with 40 eq. TCEP:

- 2 ml antibody solution (6.8x10^"8 mol) + 54.5 μΙ 50 mM TCEP solution (2.72 x 10^"6 mol)

- Incubate for 3_h at 37°C on a shaker.

- Two consecutive dialyses at 4°C in 2.0 I 1 x PBS, 1 mM EDTA, pH 7.4 in a Slide-A- Lyzer Dialysis Cassette 20Ό00 MWCO, first dialysis ca. 4 h, second dialysis overnight

- Concentrate to ca. 4.0 ml using Amicon Ultra Centrifugal Filters 50Ό00 MWCO.

[00144] Oxidation by reaction of antibody with 20 eq. dehydroascorbic acid (dhAA):

- ca. 2 ml antibody solution (6.8 x 10^"8 mol) + 27.2 μΙ fresh 50 mM dhAA solution (1 .36 x 10^"6 mol)

- Incubate for 3_h at RT on a shaker.

[00145] Conjugation with amanitin using 4 eq. HDP 30.21 15 and quenching with 25 eq. N-acetyl-L-cysteine:

Solubilize 0.7 mg HDP 30.21 15 in 70 μΙ DMSO = 10 μς/μΙ

- ca. 2 ml antibody solution (= 9.5 mg; 6.54 x 10^"8 mol) + 34.4 μΙ HDP 30.21 15 (= 344 μg; 2.62 x 10^"7 mol).

- Incubate 1 h at RT.

- Quench by addition of 16.4 μΙ 100 mM N-acetyl-L-cysteine (1 .64 x 10^"6 mol).

- Incubate 15 min at RT (or overnight at 4°C).

- Purify reaction mix with 1 x PD-10 columns equilibrated with 1 x PBS, pH 7.4.

Identify protein-containing fractions with Bradford reagent on parafilm and bring protein-containing fractions together.

- Dialysis of antibody solution at 4°C overnight in 2.0 I PBS, pH 7.4 and Slide-A- Lyzer Dialysis Cassettes 20Ό00 MWCO.

Determination of protein concentration by UV-spectra (absorption [00147] Determination of DAR by LC-ESI-MS-analysis.

[00148] Adjust protein concentration to 5.0 mg/ml (3.4 x 10-5 M) and bring to sterile conditions by filtration. Store at 4°C.

Example 5: Characterization of Drug Substance J22.9-ISY-D265C-30.2115

1 . Production and Release Testing

[00149] In order to characterize an ADC variant based on HDP 30.21 15, it was compared to a variant based on HDP 30.1699 (see Table 1 ).

HDP 30.1699

[00150] HDP 30.1699 contains amanitin from natural source in contrast to fully synthetic amanitin derivative HDP 30.21 15. Due to the chemical synthesis, there are the following differences between the two compounds: HDP 30.21 15 contains a thioether bridge and a core tryptophan moiety, whereas HDP 30.1699 contains a sulfoxide bridge and a 6-hydroxytryptophan moiety. The absence of 6- hydroxytryptophan in HDP 30.21 15 requires linkage of the antibody to the aspartic acid in contrast to HDP 30.1699 were 6-hydroxytryptophan is used for the linkage. In both compounds a cathepsin B cleavable linker was used. [00151 ] Antibody J22.9-ISY-D265C was conjugated to compounds HDP 30.1699 and HDP 30.21 15 using maleimido-chemistry as described above and tested for aggregates, DAR and product quality in analytical SEC-HPLC, Mass spectrometry, SDS-Page and Western Blot.

[00152] Test results from B16-0040 and B16-0049 are presented exemplarily in Figure 16.

[00153] Stability of J22.9-ISY-D265C-30.21 15 and J22.9-ISY-D265C-30.1699 was tested following incubation at 37°C for 0, 4, and 10 days in PBS, human, cynomolgus or mouse plasma. In contrast to J22.9-ISY-D265C-30.1699, Western Blot analyses surprisingly revealed a particularly high stability throughout the time course of the experiment for J22.9-ISY-D265C-30.21 15 in PBS or plasma (see Figure 17). Thus the ADC-containing the synthetic amanitin-derivative showed superior stability compared to the ADC with amanitin from natural source.

[00154] Plasma stability of both ADC variants was confirmed in a cell-based cytotoxicity assay on BCMA-positive NCI-H929 cells (see Table 2) after incubation for 0, 4 and 10 days in PBS, human, cynomolgus or mouse plasma. Incubation in PBS or plasma over up to 10 days has almost no influence on the cytotoxic potential of J22.9-ISY-D265C-30.21 15. In contrast, the cytotoxicity and thus the stability of J22.9-ISY-D265C-30.1699 was clearly reduced after incubation in human, cynomolgus or mouse plasma, as already observed in Western Blot analyses (see Figure 18).

Table 2:

EC~5o^"" human c nomol us mouse PBS

plasma

J22.9-ISY- day 0 2.8 x 10 -10

D265C- day 4 2.7 x 10^"1'

30.2115 day 10 1 .0 x 10^"9

^!-J227§-iSY- day 0 1 .8 x 10^"11

D265C- day 4 3.7 x 10^"7

30.1699 day 10

[00155] Cytotoxicity potentials of both ADCs were tested on five BCMA-positive MM cell lines (NCI-H929, MM.1 S Luc, MM.1 S, U266B1 , and OPM-2) and one BCMA- negative control cell line (CCRF-CEM). Both ADCs show high cytotoxic activity in the picomolar range with slightly superior toxicity of J22.9-ISY-D265C-30.1699 on some cell lines. Cytotoxicity of J22.9-ADCs is strongly dependent on BCMA-expression levels on cell surface with no toxicity observed in non-BCMA expressing control cells. In addition, with both amanitin-ADCs low EC₅o values on OPM-2 cells correlate with poor BCMA-expression (see Table 3 / Figure 19 and Table 4 / Figure 20).

Table 2:

J22.9-ISY binding

(FACS)

Multiple myeloma cell (B

lymphoblast) established from

the bone marrow of a 55-year-

MM.1S ++ +++ ++

old man with plasma cell

leukemia at relapse after

chemotherapy

Multiple myeloma cell (B

lymphoblast) cell established

RPMI- from the peripheral blood of a +/- 8226

61 -year-old man with multiple

myeloma

T lymphoblastoid line

established from the peripheral

CCRF- blood of a 3-year-old Caucasian

- CEM girl with acute lymphoblastic

leukemia (ALL) at relapse

(control cell line)

Table 4:

Example 6: ln-vivo Pharmacological Activity in Multiple Myeloma Xenograft Model

1. Subcutaneous Xenograft Model Single Dosing

[00156] Antitumor activity of J22.9-ISY-D265C-30.21 15 and J22.9-ISY-D265C- 30.1699 was tested in a subcutaneous NCI-H929 (see Table 3 for details of cell line) mouse xenograft model.

[00157] Single treatment with 2 mg/kg doses resulted in initial response followed by regrowth of tumor in 7 of 8 mice for J22.9-ISY-D265C-30.1699 and 6 of 8 mice for J22.9-ISY-D265C-30.21 15. At 4 mg/kg doses, complete tumor remission was reached in 8 of 8 mice for J22.9-ISY-D265C-30.1699 and 7 of 8 mice for J22.9- ISY-D265C-30.21 15 and mice stayed tumor free for more than 100 days (the time course of the experiment). At doses of 2 mg/kg J22.9-ISY-D265C-30.21 15 showed slightly superior anti-tumor efficacy compared to J22.9-ISY-D265C-30.1699. No statistically significant body weight reduction was observed, except for J22.9-ISY- D265C-30.1699 at 4 mg/kg doses (see Figure 21 ).

Repeated dosing

[00158] The efficacy of drug substance J22.9-ISY-D265C-30.21 15 was further evaluated in a repeated dose setting in the subcutaneous NCI-H929 xenograft model. Compared to the single dose application described above, doses were reduced to 1 , 0.5 and 0.25 mg/kg and applied either once per week (1x/week), every two weeks (1 x/2 weeks) or every three weeks (1x/3 weeks). Repeatedly applied doses of 1 mg/kg already resulted in tumor regression, and with repeated 2 mg/kg dosing complete tumor remission could be observed (see Figure 22).

2. Intravenous Xenograft Model Single Dosing

[00159] The antitumor activity of J22.9-ISY-D265C-30.1699 and J22.9-ISY- D265C-30.21 15 was further tested in a disseminating intravenous MM.1 S Luc xenograft model at doses of 2 and 4 mg/kg. Both compounds showed very effective and comparable antitumor activity with complete tumor remission for more than 100 days (see Figure 23A). Only 1 of 10 mice for both compounds showed slight regrowth of tumor after initial response for over 90 days(see Figure 23B).

[00160] In the disseminating MM.1 S Luc xenograft model doses of drug substance J22.9-ISY-D265C-30.21 15 were further reduced to single dose application of 1 , 0.3 and 0.1 mg/kg. Already single dose application of 0.3 mg/kg resulted in complete tumor eradication (see Figure 24). Example 7: Non-Human Primate (NHP) Tolerability Study

[00161 ] J22.9-ISY-D265C-30.21 15 and J22.9-ISY-D265C-30.1699 were assessed for a dose-escalating tolerability study in cynomolgus monkeys. Groups of 3 animals were injected with J22.9-ISY-D265C-30.1699 at days 1 (0.3 mg/kg), 22 (1 mg/kg), and 44 (3 mg/kg), or with J22.9-ISY-D265C-30.21 15 at days 1 (0.3 mg/kg), 21 (1 mg/kg), and 42, 64, 84,106 (3 mg/kg each time point). Animals were monitored over time for biochemical and haematological blood parameters (see Figure 25), urinalysis, body weight, food consumption, clinical signs and mortality. In addition, blood samples were collected for pharmacokinetic studies. At the end of the experiments tissue samples are used for histopathological examinations.

[00162] Up to 3 mg/kg doses of both compounds were well tolerated with no signs of kidney damage by serum parameters and unaffected body weight and food consumption. Doses of 3 mg/kg resulted in increased but transient levels of liver enzymes (ALT, AST) and the unspecific inflammatory marker LDH. Repeated dosing of 3 mg/kg J22.9-ISY-D265C-30.21 15 did not result in further increase of liver enzymes or LDH.

Example 7: Cytotoxic Potential of J22.9-ISY-D265C-30.2115 in comparison to monomethyl auristatin F derivative

[00163] Cytotoxicity potentials of J22.9-ISY-D265C-30.21 15 was compared to the interchain conjugate J22.9-ISY-MMAF on two BCMA-positive MM cell lines (MM.1 S Luc, and KMS-1 1 ). Both ADCs show high cytotoxic activity in the picomolar range after 96h with slightly superior toxicity of J22.9-ISY-MMAF (see Figure 26A). The efficacy of both compounds was compared in a timeline experiment on the two cell lines. Cytotoxicity was measured after 24, 46, 72, 96, and 120 hours (see Figures 26B + 26C). J22.9-ISY-D265C-30.21 15 unfolded the full cytotoxicity after 96 and 120 hours, whereas the MMAF conjugate showed already after 48 and 72 hours a full blown cytotoxic potential. This indicates that the amanitin derivative needs more time to unfold its cytotoxic potential and to drive the cells into apoptosis due to its mode of action than MMAF. Example 8: Efficacy of J22.9-ISY-D265C-30.2115 in a disseminating Xenograft model

[00164] The antitumor activity of J22.9-ISY-D265C-30.21 15 was compared to the interchain conjugate J22.9-ISY-MMAF in a disseminating intravenous MM.1 S Luc xenograft model at doses of 1 and 4 mg/kg. The MMAF conjugate showed a very fast initial response after five days at both doses and a regrowth of the tumor after 20 and 50 days, respectively. J22.9-ISY-D265C-30.21 15 in contrast showed a slower initial response after 10 days at both doses, but a longer period until regrowth of the tumor after 50 and 100 days, respectively (see Figure 27). Thus, the amatoxin conjugate showed a better efficacy in terms of tumor free survival despite its slower onset of action (see Figures 26(A) to (C)).

References

• Ackerman ME et al. A robust, high-throughput assay to determine the phagocytic activity of clinical antibody samples. J Immunol Methods. 201 1 Mar 7; 366(1 -2): 8-19.

• Baudino L et al. Crucial role of aspartic acid at position 265 in the CH2 domain for murine lgG2a and lgG2b Fc-associated effector functions. J Immunol. 2008 Nov 1 ;181 (9):6664-9.

• Belucci R et al. Graft-versus-tumor response in patients with multiple myeloma is associated with antibody response to BCMA, a plasma-cell membrane receptor. Blood. 2005 May 15;105(10):3945-5

• Bird JM et al. Guidelines for the diagnosis and management of multiple myeloma 201 1 . Br J Haematol. 201 1 Jul;154(1 ):32-75

• Bossen C, Schneider P. BAFF, APRIL and their receptors: structure, function and signaling. Semin Immunol. 2006 Oct;18(5):263-75

• Chesi M, Bergsagel PL. Advances in the pathogenesis of multiple myeloma. Int J Lab Hematol. 2015 May;37 Suppl 1 :108-14

• Davis MB, Preston JF. A Conjugate of a -amanitin and Monoclonal

Immunoglobulin G to Thy 1 .2 Antigen is Selectively Toxic to T Lymphoma Cells.

Science 213 (1981 ) 1385-1388 • Dimopoulos MA et al. Current treatment landscape for relapsed and/or refractory multiple myeloma. Nat Rev Clin Oncol. 2015 Jan;12(1 ):42-54

• Dimopoulos MA et al. Pathogenesis and treatment of renal failure in multiple myeloma. Leukemia. 2008 Aug;22(8):1485-93

• Hayes JM et al. Glycosylation and Fc receptors. Curr Top Microbiol Immunol. 2014;382:165-99

• Kuehl WM, Bergsagel PL. Molecular pathogenesis of multiple myeloma and its premalignant precursor. J Clin Invest. 2012 Oct;122(10):3456-63

• Leukemia. 2009 Jan;23(1 ):3-9

• Kyle RA, Rajkumar SV. Multiple myeloma. N Engl J Med. 2004 Oct

28;351 (18):1860-73.

• Marino SF, Olal D, Daumke O. A complex water network contributes to high- affinity binding in an antibody-antigen interface. Data Brief. 2015 Dec 19;6:394-7

• Morris PW, Venton DL. Regiospecific amine substitution into a-amanitin with retention of inhibitory properties against eukaryotic class II RNA polymerase. Int. J. Peptide Protein Res. 21 (1983) 419-430

• Oden et al. Potent anti-tumor response by targeting B cell maturation antigen (BCMA) in a mouse model of multiple myeloma. Mol Oncol. 2015 Aug;9(7):1348-5

• Peters C, Brown S. Antibody-drug conjugates as novel anti-cancer

chemotherapeutics. Biosci Rep. 2015 Jun 12;35(4).

• Rajkumar SV et al. International Myeloma Working Group updated criteria for the diagnosis of multiple myeloma. Lancet Oncol. 2014 Nov;15(12):e538-4

• Rickert RC et al. Signaling by the tumor necrosis factor receptor superfamily in B-cell biology and disease. Immunol Rev. 201 1 Nov;244(1 ):1 15-33

• Sanchez et al. Serum B-cell maturation antigen is elevated in multiple myeloma and correlates with disease status and survival. Br J Haematol. 2012

Sep;158(6):727-38

• Shaffer AL et al. IRF4 addiction in multiple myeloma. Nature. 2008 Jul 10;454(7201 ):226-31

• Shields RL et al. High resolution mapping of the binding site on human lgG1 for Fc gamma Rl, Fc gamma Rll, Fc gamma RIM, and FcRn and design of lgG1 variants with improved binding to the Fc gamma R. J Biol Chem. 2001 Mar

2;276(9):6591 -604

• Trail PA. Antibody drug conjugates as cancer therapeutics. Antibodies 2013, 2(1 ), 1 13-129. • Wieland T, Faulstich H. Amatoxins, phallotoxins, phallolysin, and antamanide: the biologically active components of poisonous Amanita mushrooms. CRC Crit Rev Biochem. 1978 Dec;5(3):185-260

• Zhao L, May JP, Blanc A, Dr. Dietrich DJ, Loonchanta A, Matinkhoo K, Pryyma A, and Perrin DM. Synthesis of a Cytotoxic Amanitin for Biorthogonal Conjugation. ChemBioChem 16 (2015) 1420-1425

Sequences:

Humanized J22.9 heavy chain J22.9-ISY-D265C (SEQ ID NO: 1 ):

EVQLVESGGG LVQPGGSLRL SCAASGFTFS RYWISWVRQA PGKGLVWVGE INPSSSTINY APSLKDKFTI SRDNAKNTLY LQMNSLRAED TAVYYCASLY YDYGDAYDYW GQGTLVTVSS ASTKGPSVFP LAPSSKSTSG GTAALGCLVK DYFPEPVTVS WNSGALTSGV HTFPAVLQSS GLYSLSSVVT VPSSSLGTQT YICNVNHKPS NTKVDKKVEP KSCDKTHTCP PCPAPELLGG PSVFLFPPKP KDTLMI SRTP EVTCVVVCVS HEDPEVKFNW YVDGVEVHNA KTKPREEQYN STYRVVSVLT VLHQDWLNGK EYKCKVSNKA LPAPIEKTIS KAKGQPREPQ VYTLPPSREE MTKNQVSLTC LVKGFYPSDI AVEWESNGQP ENNYKTTPPV LDSDGSFFLY SKLTVDKSRW QQGNVFSCSV MHEALHNHYT QKSLSLSPGK

Humanized J22.9 light chain (SEQ ID NO: 2):

EIVMTQSPAT LSVSPGERAT LSCKASQSVE SNVAWYQQKP GQAPRALIYS ASLRFSGI PA RFSGSGSGTE FTLTISSLQS EDFAVYYCQQ YNNYPLTFGA GTKLELKRTV AAPSVFIFPP SDEQLKSGTA SVVCLLNNFY PREAKVQWKV DNALQSGNSQ ESVTEQDSKD STYSLSSTLT LSKADYEKHK VYACEVTHQG LSSPVTKSFN RGEC

Claims

A conjugate according to Formula I comprising (a) an amatoxin comprising (i) an amino acid 4 with a 6'-deoxy position; and (ii) an amino acid 8 with an S- deoxy position; (b) a BCMA-binding moiety comprising (i) the variable domains of the heavy chain according to SEQ ID NO: 1 and the light chain according to SEQ ID NO: 2 of antibody J22.9-ISY, and (ii) a heavy chain constant region comprising a D265C mutation; and (c) a protease-cleavable linker, wherein said BCMA-binding moiety is attached to said linker via the thiol group of the cysteine residue at position 265 in the antibody heavy chain.

Formula I

The conjugate of claim 1 , wherein said BCMA-binding moiety comprises the heavy chain according to SEQ ID NO: 1 and the light chain according to SEQ ID NO: 2.

The conjugate of claims 1 or 2, wherein said BCMA-binding moiety is an lgG1 .

A pharmaceutical composition comprising the conjugate of any one of claims 1 to 3.

5. The conjugate of any one of claims 1 to 3 or the pharmaceutical composition of claim 4 for use in the treatment of cancer in a patient, particularly wherein the cancer is selected from the group consisting of multiple myeloma, diffuse large B-cell lymphoma (DLBCL), and chronic lymphocytic leukemia (CLL), particularly multiple myeloma.