Novel masking antibodies
By using antigen peptides and sdAb as the masking part, the antibody masking method solves the problem of high molecular weight aggregates formed during the production process of antibody masking technology in the existing technology, achieving efficient production and effectiveness of antibodies in ADCs.
Patent Information
- Application Number
- CN202380040891.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2022-05-16
- Filing Date
- 2023-05-15
- Publication Date
- 2025-01-21
AI Technical Summary
Existing antibody masking techniques are prone to forming high molecular weight aggregates during production, resulting in low yields and induction of anti-drug antibodies, and are not suitable for use in antibody-drug conjugates (ADCs).
An antibody masking method is used that contains an antigenic peptide and a single domain antibody (sdAb) that specifically binds to the antigenic peptide as the masking part, and is connected through a cleavable linker to ensure that the mask is only cleaved at the target site and activates the binding ability of the antibody.
It achieves efficient production of masking antibodies without forming high molecular weight aggregates, avoids the induction of anti-drug antibodies, and ensures the effectiveness and safety of antibodies in ADCs.
Smart Images

Figure BDA0005136760030000431 
Figure BDA0005136760030000451 
Figure BDA0005136760030000481
Abstract
Description
Technical Field
[0001] The present invention relates to a new masked antibody or an antigen-binding fragment thereof, which comprises at least one light chain (LC) and heavy chain (HC) variable domain pair with masking, wherein the masking comprises two peptide masking parts that are specifically bound to each other and are connected to the N-termini of the light chain and the heavy chain, respectively, wherein at least one masking part is connected by a cleavable linker. The present invention also provides an antibody-drug conjugate (ADC) comprising the masked antibody and a pharmaceutical composition comprising the masked antibody or ADC. Background Art
[0002] Antibodies are widely used in the treatment of diseases such as cancer, autoimmune diseases or inflammatory diseases. Antibodies specifically recognize target antigens and bind to target antigens, which can trigger therapeutic effects. For example, antibodies can bind to receptors that are crucial in biological processes, and can have exciting or antagonistic effects on antigen receptors. Receptor binding can interfere with, for example, cell growth (e.g., tumor growth), or can have immunomodulatory effects, such as triggering the immune system. Antibodies can have their own therapeutic effects and / or can be used as carriers for delivering drugs to target sites (e.g., cancer cells), for example, as part of ADC.
[0003] In all cases, the therapeutic effect depends on the ability of the antibody to bind to its target antigen. Therapeutic antibodies should bind to their antigens with sufficient affinity and specificity to ensure that the desired therapeutic effect can be achieved with minimal side effects and / or toxicity. Not only is the selectivity of the antibody to its target important, but the specificity of the target to the disease to be treated (site) is also important. Some target antigens may also be present on non-target cells, such as in healthy tissues, and the antibody binding at these sites should be minimized to avoid undesirable therapeutic side effects.
[0004] In order to avoid the binding of therapeutic antibodies at sites other than the disease site, the antibody can be modified in such a way that the binding site of the antibody is "masked" and the binding of the antibody to its target antigen is inhibited by the mask until the antibody reaches the desired target site. Only at the target site, by removing the "mask", the ability of the antibody to bind to its target antigen is activated. In order to ensure that the mask is removed only at the desired target site (e.g., the diseased cells of a tumor), the masked component (masking portion) can be bound to the antibody through a cleavable linker, which is only cleaved when the antibody reaches its target destination. The cleavable linker may contain a cleavage site for an enzyme that is specific to certain tissues (e.g., a cleavage site recognized by a tumor-specific proteolytic enzyme, such as a tumor-specific matrix metalloproteinase (e.g., MMP2, MMP9 or MMP14), a protein lysing enzyme (matriptase) and a urokinase plasminogen activator present in the extracellular matrix).
[0005] The mask is removed from the antibody by cleavage of the cleavable linker at the target site (eg, a tumor site) and the antigen binding site of the antibody becomes available for binding to its target antigen.
[0006] Several antibody masking techniques are known in the art and are reviewed in, for example, Lucchi et al., 2021, ACSCent. Sci., 7, 724-738 and Lin et al., 2020, Journal of Biomedical Science, 27:76.
[0007] "Affinity-based" masking strategies known in the art rely on the use of a masking moiety that binds to the antigen binding site of an antibody, thereby preventing the masked antibody from binding to its target antigen until the mask is removed by cleavage of a cleavable linker connecting the masking moiety to the antibody. Such a masking moiety is designed to bind to the antigen binding site of the masking antibody with a lower affinity than the actual target antigen of the antibody. Once the linker holding the masking moiety to the antibody is cleaved, the masking moiety is replaced by the target antigen because the antigen binds to the targeting antibody with a higher affinity than the masking moiety.
[0008] Affinity-based masking methods are described in WO2004 / 009638 (ISIS Innovation Ltd.). In the method, the therapeutic antibody has a masking moiety that can be a peptide mimetic epitope that reversibly binds to the antigen binding site of the antibody (called antibody combining site (ACS)) in the initial stage after administration. The masking moiety is connected to the antibody heavy chain or light chain through an enzymatically cleavable linker (a linker carrying an enzymatically degradable motif that is easily cleaved by host enzymes in vivo). In WO2009 / 025846 (Cytomx, Univ. California), a masked anti-VEGF antibody is described, wherein the masking moiety is a peptide carrying a cysteine residue that is bound to the antibody through a peptide linker that is cleavable by the MMP2 protease. The cysteine in the masking moiety forms a disulfide bridge with a cysteine near (or within) the target binding site of the antibody, thereby masking the antigen binding site of the antibody. In the cleaved state, the antigen binding site of the antibody becomes available for binding to the target protein. In US8895702 B2 (Williams et al.), cross-masked antibody pairs are described, wherein a cross-masked epitope that binds to the antigen binding site of the other antibody is linked to each antibody via a cleavable linker, thereby forming a cross-masked complex until the cleavable linker is cleaved by MMP9 at the target site. A disadvantage of affinity-based masking strategies, in which the masking moiety specifically binds to the antigen binding site of the antibody, is that each antibody requires a different specific mask for its own unique antigen binding site, and the affinity of the mask needs to be carefully optimized for the antigen binding site.
[0009] Such antibody masking techniques have also been developed, which provide more generally applicable masking that can be applied to different antibodies and is unrelated to their target antigens. In these methods, the antibody has one or more masking parts that do not work by directly binding to the antigen binding site of the antibody. On the contrary, the antibody has one or more masking parts that prevent the antibody from binding to its target antigen by spatially interfering with the binding of the antibody rather than by occupying the antigen binding site of the antibody. In these methods, the N-termini of both the antibody light chain (LC) and the antibody heavy chain (HC) of the LC-HC pair can be extended with peptide masking parts, which have affinity for each other and form a mask by binding to each other rather than binding to the antigen binding site of the antibody. These masking parts can be combined with the N-termini of the antibody LC and HC in such a way that in the uncut state, the two masking parts bind to each other to form a masking that shields the antigen binding site of the antibody, and in the cut state, the masking is removed and the antibody binding site becomes available for binding to its target antigen again.
[0010] In WO2014 / 193973 (DCB USA LLC, Univ.Kaohsiung Medical), such an antibody masking method is described, in which the hinge region of a human antibody is used as a masking portion. The closed hinge region consists of two peptide arms interconnected by a disulfide bond. Each arm of the hinge region is connected to the heavy chain and light chain of the antibody binding domain with a cleavable linker (each antibody has four arms and four cleavable links). In WO2020 / 229553 (Ultrahuman Six Ltd.), such a masking method is described, in which two antigen-binding moieties (both of which can be antibodies (parts) bound to disease-specific antigens) are connected to each other by a cleavable peptide linker, which comprises an amino acid sequence from a human hinge region. In an uncut state, the binding of this part to its respective target antigen is blocked. In WO2018 / 107125 (Seattle Genetics Inc.), the use of a coiled-coil to form a peptide as a masking portion is described. The coiled-coil forming peptide is linked to the N-terminus of at least one of the antibody light and heavy chain pairs via a linker comprising a protease cleavage site. In the uncleaved state, the peptides associate to form a coiled-coil structure and thereby reduce the binding affinity of the light-heavy chain pair of the masked antibody to the target antigen. When exposed to tumor-associated proteases (e.g., MMP2 and MMP9), the coiled-coil peptide is cleaved and the binding of the antibody to its target antigen is restored.
[0011] A problem with existing masking techniques may be that the inclusion of masking in antibody products may lead to the formation of high molecular weight (HMW) aggregates during production. For example, HMW aggregates may form when masking moieties on different antibodies bind to each other rather than to their counterpart masking moieties on the same antibody. Therefore, the yield of masked antibodies available in production may be low, and / or additional steps must be taken to remove HMW aggregates from the product. Excessive aggregation may also lead to the induction of anti-drug antibodies (ADA).
[0012] For several masking methods known in the art, separate steps are proposed to solve the problem of HMW aggregate formation. In WO2020 / 247572 and WO2020 / 247574 (Seattle Genetics Inc.), the problem of HMW aggregate formation associated with masking methods using (hydrophobic) coiled coil masking during purification, processing and storage is described. For example, it is recommended to store masked antibodies in an aqueous buffer with a pH of 3.5 to 4.5. In WO2020 / 028401 (Amgen Inc.), a solution to the formation of HMW aggregates in the context of producing masked antibodies is also sought in terms of changing the pH of the drug antibody composition.
[0013] Despite the fact that several general masking techniques already exist, there is still a need for generally applicable antibody masking techniques that are easy to produce (e.g., without or limited HMW aggregate formation). This is particularly true for antibodies used in antibody-drug conjugates (ADCs), where the antibody masking should also not interfere with the attachment of the linker-drug molecule to the masking antibody. The linker-drug molecule is typically attached via (modified) cysteines within the antibody protein molecule. In particular, masking techniques that rely on the binding of a peptide masking moiety to cysteines or the binding of a peptide masking moiety via the formation of an -SS-bridge between cysteines may not be well suited for use in masked antibody-based ADCs due to the risk of interfering with the linker-drug conjugation. Summary of the invention
[0014] The present invention relates to a new method for masking antibodies. The present invention provides a masked antibody or an antigen-binding fragment thereof, wherein the masked antibody or the antigen-binding fragment thereof comprises at least one light chain and heavy chain pair having a mask, wherein the mask comprises two peptide masking parts that specifically bind to each other and are connected to the N-termini of the light chain and the heavy chain of the pair, respectively, wherein at least one masking part is connected by a cleavable linker, and one of the two masking parts comprises an antigenic peptide and the other comprises a single domain antibody (sdAb) that specifically binds to the antigenic peptide.
[0015] The present invention is based on the novel use of antigenic peptides and sdAbs that specifically bind to said antigenic peptides in masking the masking part of an antibody.
[0016] The present invention also provides an antibody-drug conjugate (ADC) based on the masked antibody according to the present invention, the use of the masked antibody or ADC as a drug, and a pharmaceutical composition comprising the masked antibody or ADC.
[0017] More specifically, the present invention is set out as follows.
[0018] An activatable antibody or antigen-binding fragment thereof, comprising an antigen-binding site, the antigen-binding site comprising a heavy chain variable domain and a light chain variable domain capable of binding to a target protein, wherein the antigen-binding site has a mask suitable for inhibiting binding of the antigen-binding site to the target protein, the mask comprising two masking parts:
[0019] i. a masking moiety connected to the N-terminus of the variable heavy chain domain, and
[0020] ii. a masking moiety connected to the N-terminus of the variable light chain domain,
[0021] wherein at least one of the masking moieties is linked by a cleavable linker, and wherein one of the masking moieties comprises an antigenic peptide and the other masking moiety comprises a single domain antibody (sdAb) that specifically binds to the antigenic peptide.
[0022] The activatable antibodies according to the present invention can be used for treatment by administering them to a subject in need of treatment (eg, a cancer patient receiving immunotherapy with a therapeutic anti-cancer antibody or antibody drug conjugate (ADC)).
[0023] The activatable antibody or antigen-binding fragment thereof according to the present invention comprises an antigen binding site that is capable of binding to a target protein in a subject. After administration of the antibody to a subject, the masking is removed in vivo, preferably at the target (disease, e.g., tumor) site, by cleavage of at least one cleavable linker.
[0024] In a preferred embodiment, the activatable antibody or antigen-binding fragment thereof comprises two antigen-binding sites, wherein both antigen-binding sites are masked.
[0025] Alternatively or in combination with the preferred embodiments of the foregoing, in a preferred embodiment, the antigenic peptide is 4 to 20 amino acid residues in length, preferably wherein the antigenic peptide is 12 to 20 amino acid residues in length, and comprises or consists of an amino acid sequence as defined in SEQ ID NO:51. More preferably, the antigenic peptide comprises or consists of an amino acid sequence as defined in SEQ ID NO:1, SEQ ID NO:2 (SA mutant) or SEQ ID NO:3 (ST mutant). Even more preferably, the antigenic peptide consists of an amino acid sequence as defined in SEQ ID NO:1, SEQ ID NO:2 (SA mutant) or SEQ ID NO:3 (ST mutant).
[0026] Alternatively or in combination with the preferred embodiments above, in a preferred embodiment, the Nanobody comprises or consists of an amino acid sequence as defined in SEQ ID NO: 4, a humanized form thereof, or a variant of SEQ ID NO: 4, such as a variant described in SEQ ID NO: 76 (RA mutant), SEQ ID NO: 77 (RE mutant) or SEQ ID NO: 78 (RS mutant), or a humanized form of said variant. In the RA, RE and RS mutants of the B2C Nanobody, the original CDR3 (SEQ ID NO: 5) is replaced by the CDR3 described in SEQ ID NO: 6, SEQ ID NO: 7 or SEQ ID NO: 8, respectively.
[0027] The humanized form of the BC2 Nanobody may be a Nanobody having SEQ ID NO: 79 (BC2 Nb人源化1 )、SEQ ID NO:80(BC2 Nb人源化2 ) or SEQ ID NO:81(BC2 Nb人源化3 ) in a Nanobody having the amino acid sequence described in .
[0028] The humanized form of the RA mutant Nanobody may be a variant having SEQ ID NO: 82 (RA h1 )、SEQ ID NO:83(RA h2 ) or SEQ ID NO:84 (RA h3 ) A Nanobody having an amino acid sequence as described in ). A humanized form of a RE mutant Nanobody may be one having SEQ ID NO: 85 (RE h1 )、SEQ ID NO:86(RE h2 ) or SEQ ID NO:87 (RE h3 ) A Nanobody with an amino acid sequence as described in ). A humanized form of a RS mutant Nanobody may be one having SEQ ID NO: 88 (RS h1 )、SEQ ID NO:89(RS h2 ) or SEQ ID NO:90 (RS h3 ) in a Nanobody having the amino acid sequence described in .
[0029] In a preferred embodiment, the antigenic peptide comprises or consists of a peptide having the amino acid sequence defined in SEQ ID NO: 2 (SA mutant), or a humanized form thereof, and the Nanobody comprises the amino acid sequence as defined in SEQ ID NO: 78 (RS mutant).
[0030] Preferably the humanized Nanobody is RS h1 .
[0031] Alternatively or in combination with the preceding preferred embodiments, in a preferred embodiment the masking moiety comprising the sdAb is linked to the N-terminus of the light chain variable domain and the masking moiety comprising the antigenic peptide is linked to the N-terminus of the heavy chain variable domain.
[0032] In a preferred embodiment, the antigenic peptide is preceded by a spacer amino acid sequence as defined in SEQ ID NO: 9 linked to the N-terminus of the antigenic peptide.
[0033] Alternatively or in combination with the preferred embodiments of the foregoing, in a preferred embodiment, the cleavable linker comprises one or more cleavage sites recognized by one or more tumor-specific proteases. This means that each cleavable linker may comprise one or more cleavage sites. Each cleavage site may be recognized by only one tumor-specific protease or by multiple tumor-specific proteases. Preferably, at least one cleavable linker comprises a cleavage site recognized by a protease or a cleavage site recognized by a metalloprotease.
[0034] In a second aspect, the present invention relates to an antibody-drug conjugate (ADC) comprising an activatable antibody or antigen-binding fragment thereof according to the present invention, and a linker-drug.
[0035] In a third aspect, the present invention relates to a pharmaceutical composition comprising an activatable antibody or antigen-binding fragment thereof according to the present invention, or an ADC according to the present invention, and a pharmaceutically acceptable excipient.
[0036] In a fourth aspect, the present invention relates to an activatable antibody or antigen-binding fragment thereof according to the invention, an ADC according to the invention or a pharmaceutical composition according to the invention for use as a medicament.
[0037] In a fifth aspect, the present invention relates to an activatable antibody or antigen-binding fragment thereof according to the present invention, an ADC according to the present invention or a pharmaceutical composition according to the present invention for use in treating cancer, an autoimmune disease or an infectious disease, preferably for treating cancer.
[0038] In a sixth aspect, the present invention relates to a nucleic acid construct comprising: (a) a nucleotide sequence encoding a heavy chain variable domain and a masking moiety; and a nucleotide sequence encoding a light chain variable domain and a masking moiety; wherein the nucleotide sequence is operably linked to an expression control sequence for expression in a host cell, preferably in a mammalian host cell. Preferably, the heavy chain variable domain, the light chain variable domain and the masking moiety are as defined with respect to the first aspect of the present invention. In a preferred embodiment, the nucleic acid construct is a vector compatible with the host cell.
[0039] In a seventh aspect, the present invention relates to a host cell comprising the nucleic acid construct of the present invention.
[0040] In an eighth aspect, the present invention relates to a method for producing an activatable antibody or an antigen-binding fragment thereof according to the present invention, the method comprising the step of culturing a host cell according to the present invention under conditions favorable for the expression of the activatable antibody or antigen-binding fragment thereof.
[0041] In a ninth aspect, the present invention relates to the use of an antigenic peptide and an sdAb that specifically binds to the antigenic peptide as part of a mask in an activatable antibody or antigen-binding fragment thereof. BRIEF DESCRIPTION OF THE DRAWINGS
[0042] Figure 1A : Schematic representation of an antibody having a masking portion comprising a nanobody on the HC and a masking portion comprising an antigenic peptide on the LC, and vice versa.
[0043] Figure 1B : Schematic representation of high molecular weight (HMW) complex formation between masked antibodies during production.
[0044] Figure 2A to L: Dose-dependent binding of antibody S103 and antibodies S201 to 211 on HER2-positive SK-BR-3 cells when unmasked after masking and MMP2 treatment compared to unmasked (naked) trastuzumab binding as measured on a flow cytometer. Data show MFI (median fluorescence intensity) from one experiment.
[0045] Figure 3 Figure : Cytotoxicity induced by ADC-S105 on HER2-positive SK-BR-3 cells when masked and after MMP2 treatment (unmasking) compared to trastuzumab ADC (ADC-TRA41C) and isotype control rituximab ADC (ADC-RIT41C).
[0046] Figure 4 : Cytotoxicity induced by ADC ADC-S423 (masking moiety connected by L9 linker cleavable by MMP2 protease), ADC-S426 (masking moiety connected by L13 scrambled linker not cleavable by MMP2 protease), ADC-TRA41C and isotype control ADC-RIT41C on HER2 positive SK-BR-3 cells. As a control for the effect of MMP2 on ADC, ADC-TRA41C was also treated with MMP2 protease.
[0047] Figure 5: Dose-dependent binding of ADC-S423 and ADC-S426 on HER2-positive SK-BR-3 cells when they were masked and after MMP2 treatment as compared to ADC-TRA41C binding (+ / - MMP2 treatment) measured on a flow cytometer. Data show MFI (median fluorescence intensity) from one experiment.
[0048] Figure 6 : Dose-dependent binding of semi-masked antibodies S109 and S110 compared to trastuzumab on HER2-positive SK-BR-3 cells HER2- as measured on a flow cytometer. Data show MFI (median fluorescence intensity) from two experiments.
[0049] Fig. 7A : Dose-dependent binding of ADC-S424 and ADC-S425 on HER2-positive SK-BR-3 cells when masked and semi-unmasked after MMP2 treatment compared to ADC-TRA41C. Data show MFI (median fluorescence intensity) from one experiment.
[0050] Figure 7B : Cytotoxicity induced by semi-masked ADCs ADC-S109, ADC-S110 and ADC-S111 compared to that induced by ADC-TRA41C and isotype control ADC-RIT41C for HER2-positive SK-BR-3 cells.
[0051] Figure 7C : Cytotoxicity induced by semi-masked ADCs ADC-S424, ADC-S425 on HER2-positive SK-BR-3 cells, compared with that induced by ADC-TRA41C and isotype control ADC-RIT41C.
[0052] Fig. 8A To F: Target binding results for naked (unmasked), masked and masked-41C forms of mAbs targeting EpCAM (ING-1 and adecatumumab), CD137 (4-1BB) (urelumab) and TROP2 (sacituzumab), tissue factor (MORAb-066 and tisotumab) as well as naked (unmasked), masked CD137 (4-1BB) (urelumab), as measured using biolayer interferometry and corrected for differences in concentration (target binding / protein A binding*1000).
[0053] Fig. 9: High molecular weight percentage (HMW (%) - Y axis) relative to the average Drug-to-Antibody Ratio (DAR - X axis) for ADCs with different DARs generated by masking antibodies S103 and S211.
[0054] Figure 10: Results of testing a linker matrix of (18×18=324) linker pairs to attach masking to the tisofomab antibody. The masking moiety used in each case was a pair of masking moieties comprising a masking moiety with a BC2T SA mutant peptide attached to each LC of the anti-TF antibody tisofomab, and a masking moiety with a BC2-NbRS mutant nanobody attached to each HC of the anti-TF antibody tisofomab. For all antibodies produced, the amount of antibody produced (μg / mL) and HMW (%) (percentage of total antibody lost due to formation of high molecular weight complexes) were determined. The amount of antigen obtained for each antibody was Fig. 10A The HMW (%) of each antibody is shown in Fig. 10B The results of the bio-layer interferometry (BLI) binding experiment are shown in Fig. 10C Shown in FIG. 1 is binding to tissue factor / binding to protein A*1000. "ND" means the value was not determined.
[0055] Fig.11 : In vivo testing results of ADC 1 to 13 in the non-small cell lung cancer model LXFA629. The ADC dose was 2.4 mg / kg. The dose of unmasked (naked) teisomonab P41C and ADC (ADC13) was 2 mg / kg. The arrows show the time of randomization and dosing, N = 6 mice / group, data are expressed as mean ± SEM. Each data point represents the absolute tumor volume (mm) of all mice in each group 3 )’s average value.
[0056] Fig.12 : Absolute tumor volume (mm) of mice bearing TF-positive LXFA629 tumors in each group on the last day (day 42). 3 ), including the average. DETAILED DESCRIPTION
[0057] The present invention provides a novel antibody masking method, wherein an antigenic peptide and a single domain antibody (sdAb) that specifically binds to the antigenic peptide are used as part of the masking moiety.
[0058] The present invention provides a masked antibody or an antigen-binding fragment thereof, wherein the masked antibody or the antigen-binding fragment thereof comprises at least one light chain and heavy chain pair having a mask, wherein the mask comprises two peptide masking parts which specifically bind to each other and are connected to the N-termini of the light chain and the heavy chain of the pair, respectively, wherein at least one masking part is connected by a cleavable linker, one of the two masking parts comprises an antigenic peptide and the other comprises an sdAb which specifically binds to the antigenic peptide.
[0059] Therefore, the present invention provides novel uses of antigenic peptides and sdAbs that specifically bind to antigenic peptides as part of a masking moiety in a masked antibody or antigen-binding fragment.
[0060] The use of sdAb in / as an antibody masking moiety has several advantages. The masking effect is based on the specific affinity between the sdAb and its antigenic peptide (binding partner). By binding to each other, the two masking moieties shield the binding site of the antibody and block or at least significantly reduce the ability of the antibody to bind to its target antigen.
[0061] The present invention has a masking in which the bond between the masking moieties is independent of the formation of any -SS-bridge formed between the cysteine residues in the two masking moieties. This is an advantage, in particular, when the masked antibody is part of an ADC, since the linker-drug typically binds to (modified) cysteine residues in the antibody protein. Therefore, the masked antibodies or masked antigen-binding fragments thereof according to the present invention are particularly useful in ADCs. The masked antibodies according to the present invention also have the advantage that they can be produced with relatively high titers without losing most of the produced masked antibodies due to the formation of high molecular weight (HMW) complexes.
[0062] Thus, in a first aspect, the present invention relates to an activatable antibody or antigen-binding fragment thereof, comprising an antigen binding site comprising a heavy chain variable domain and a light chain variable domain capable of binding to a target protein, wherein the antigen binding site has a mask suitable for inhibiting binding of the antigen binding site to the target protein, the mask comprising two masking parts:
[0063] i. a masking moiety connected to the N-terminus of the variable heavy chain domain, and
[0064] ii. a masking moiety connected to the N-terminus of the variable light chain domain,
[0065] At least one of the masking moieties is connected by a cleavable linker, and one of the masking moieties comprises an antigenic peptide and the other masking moiety comprises a single domain antibody (sdAb) that specifically binds to the antigenic peptide.
[0066] The activatable antibodies according to the present invention can be used for treatment by administering them to a subject in need of treatment (eg, a cancer patient receiving immunotherapy with a therapeutic anti-cancer antibody or antibody drug conjugate (ADC)).
[0067] The activatable antibody or antigen-binding fragment thereof according to the present invention comprises an antigen binding site that is capable of binding to a target protein in a subject. In the activatable antibody, the antigen binding site has a mask that inhibits binding to the target protein. After the antibody is administered to the subject, the mask is removed in vivo, preferably at the target (disease, such as a tumor) site, wherein at least one cleavable linker is cleaved, after which the binding site of the antibody can bind to the target protein in the subject.
[0068] Antibody
[0069] The term "antibody" as used herein preferably refers to an antibody comprising two heavy chains and two light chains. The antibodies used according to the present invention may be of any isotype, such as IgA, IgE, IgG or IgM antibodies. Preferably, the antibody is an IgG antibody, more preferably an IgG 1 or IgG 2 Antibodies. The antibodies may be chimeric, humanized or human. Preferably, the antibodies are humanized or human. Even more preferably, the antibodies are humanized or human IgG antibodies, more preferably humanized or human IgG 1 Monoclonal antibodies. The antibodies may have a κ (kappa) or λ (lambda) light chain, preferably a κ (kappa) light chain, i.e., humanized or human IgG 1 -κ antibody.
[0070] The term "antigen-binding fragment" as used herein includes Fab, Fab', F(ab') 2 , Fv, scFv or reduced IgG (rIgG) fragment, or any fragment of an antibody comprising at least the variable binding regions of a light chain (LC) and a heavy chain (HC) pair.
[0071] A "humanized" form of a non-human (e.g., rodent) antibody is an antibody (e.g., a non-human-human chimeric antibody) comprising a minimal sequence derived from a non-human antibody. Various methods for humanizing non-human antibodies are known in the art. For example, the antigen-binding complementary determining regions (CDRs) in the variable domains (also referred to as variable regions or VRs) of the heavy chain (HC) and light chain (LC) are derived from antibodies from non-human species (usually mice, rats, or rabbits). These non-human CDRs can be combined with human framework regions (FRs, i.e., FR1, FR2, FR3, and FR4) of the HC and LC variable regions, in such a way that at least partially retains the functional properties of the antibody, such as binding affinity and specificity. The selected amino acids in the human FRs can be exchanged with the corresponding original non-human species amino acids to further improve antibody performance, such as improving binding affinity while maintaining low immunogenicity. Therefore, humanized variable regions are usually combined with human constant regions. An exemplary method for humanizing non-human antibodies is the method of Winter and colleagues (Jones et al, 1986, Nature, 321, 522-525; Riechmann et al, 1988, Nature, 332, 323-327; Verhoeynet al, 1988, Science, 239, 1534-1536). Alternatively, non-human antibodies can be humanized by modifying their amino acid sequences to increase the similarity to naturally occurring antibody variants in humans. For example, the selected amino acids of the original non-human species FR are replaced with their corresponding human amino acids to reduce immunogenicity while retaining the binding affinity of the antibody. For additional details, see Jones et al, 1986, Nature, 321, 522-525; Riechmann et al, 1988, Nature, 332, 323-327; and Presta, 1992, Curr. Op. Struct. Biol., 2, 593-596. See also the following review articles and references cited therein: Vaswani and Hamilton, 1998, Ann. Allergy, Asthma and Immunol., 1, 105-115; Harris, 1995, Biochem. Soc. Transactions, 23, 1035-1038; and Hurle and Gross, 1994, Curr. Op. Biotech., 5, 428-433.
[0072] CDRs can be determined using the method of Kabat (in Kabat, EA et al, 1991, Sequences of Proteins of Immunological Interest, 5th ed. Public Health Service, National Institutes of Health, Bethesda, MD, NIH publication no. 91-3242, pp. 662, 680, 689), Chothia (Chothia et al, 1989, Nature 342, 877-883), or IMGT (Lefranc, 1999, The Immunologist, 7, 132-136).
[0073] Typically, the antibody is a monospecific (i.e., specific for one antigen; such antigens may be shared between species, or have similar amino acid sequences between species) or a bispecific (i.e., specific for two different antigens of a species) antibody comprising at least one HC and LC variable region that binds to an antigenic target, preferably a membrane-bound antigenic target that may or may not be internalized. Preferably, in the case of an antibody drug conjugate, the antibody is internalized by the target cell after binding to the (antigen) target, after which the active effector molecule is released intracellularly.
[0074] Examples of therapeutic antibodies known in the art include: blinatumomab (CD19), rituximab (CD20), epratuzumab (CD22), iratumumab and brentuximab (CD30), vadastuximab (CD33), tetulumab (CD37), isatuximab (CD38), bivatuzumab (CD44), lorvotuzumab (CD56), vorsetuzumab (CD70), milatuzumab (CD74), polatuzumab (CD79), rovalpituzumab (DLL3) , futuximab (EGFR), oportuzumab (EPCAM), ING-1 (EpCAM), adelimumab (EpCAM), farletuzumab (FOLR1), glembatumumab (GPNMB), trastuzumab, pertuzumab and margetuximab (HER2), etaracizumab (integrin), anetumab (mesothelin), pankomab (MUC1), enfortumab (connexin-4), H8, A1, A3 (5T4), MORAb-66, tuxomab (TF), gosartumumab (TROP2) and usrulumab (TNFRSF9, CD137), etc.
[0075] Masking antibodies or antigen-binding fragments
[0076] A masked antibody or antigen-binding fragment thereof is an antibody or antigen-binding fragment with (removable) masking as defined above. The masking of the antibody or antigen-binding fragment consists of one or more masking parts, and blocks or at least significantly reduces the ability of the antibody or antigen-binding fragment to bind to its target antigen until the blocking effect of the masking is removed ("de-masking"), and the ability of the antibody or antigen-binding fragment to bind to its target antigen is fully restored to have the desired therapeutic effect. Masked antibodies or antigen-binding fragments are also referred to as activatable antibodies or antigen-binding fragments, or antibody or antigen-binding fragment-prodrugs, to indicate that they must be activated by removing the masking effect of the antibody masking, thereby showing their therapeutic effect. The terms "masked antibody" and "activatable antibody" are used interchangeably herein. According to the present invention, the masking parts used to produce at least one "mask" in the masked antibody or antigen-binding fragment appear in pairs; one masking part comprises an antigenic peptide, and the other comprises an sdAb that specifically binds to the antigenic peptide. According to the invention, one of the masking moieties of the pair is attached to the HCVR of at least one HC / LC binding pair of the masking antibody or antigen-binding fragment thereof, while the other is attached to the LCVR of at least one HC / LC binding pair of the masking antibody or antigen-binding fragment thereof. By binding to each other, the masking moieties form a mask that shields (at least one) binding site in the antibody or antigen-binding fragment thereof. Thus, an antibody or antigen-binding fragment thereof having two antigen-binding sites may carry two masks; one on each HCVR / LCVR pair.
[0077] Antibodies or antigen-binding fragments thereof can be masked to prevent or reduce off-target effects, since most target antigens are not completely specific to the target (disease) site. By shielding the antigen binding site in the masked antibody or antigen-binding fragment, masking reduces the ability of the antibody or antigen-binding fragment to bind to its target antigen until it reaches its target (disease) site, where the masking is removed. Reduced binding means that the ability of the masked antibody or antigen-binding fragment to bind to its target is significantly reduced by masking when compared to a naked (unmasked) or unmasked (after the mask has been removed from the antigen binding site) antibody or antigen-binding fragment.
[0078] Antibody or antigen binding fragment binding
[0079] Therapeutic antibodies or antigen binding fragments are selected for their selectivity and their affinity for the target antigen. Affinity is the strength of the binding interaction between a binding molecule and its ligand (e.g., between an antibody or antigen binding fragment and its target antigen). The target antigen may be an antigenic protein (the antigenic portion of a protein), called an "antigenic determinant" or "epitope," which may be presented on the surface of a target cell to which the antibody or antigen binding fragment binds. In the case of using a therapeutic antibody or antigen binding fragment in cancer treatment, the target cell may be a tumor cell.
[0080] Antibody or antigen binding fragment selectivity is a measure of how well an antibody or antigen binding fragment binds to its target antigen in a mixture of different proteins. In other words: if the epitope to which the antibody or antigen binding fragment binds is unique to one specific target antigen, then the selectivity is high because the antibody or antigen binding fragment binds only to the target antigen and does not cross-react with any off-target proteins present in the mixture.
[0081] Antibody or antigen binding fragment specificity refers to the ability of an antibody or antigen binding fragment to recognize a single epitope. Thus, a monoclonal antibody or antigen binding fragment that binds to a single epitope is specific. However, when the single epitope to which the antibody or antigen binding fragment binds occurs in multiple proteins, the antibody or antigen binding fragment may be specific (for the epitope to which it binds with high affinity) but not selective (because the epitope also occurs in non-target proteins). Therapeutic antibodies or antigen binding fragments preferably bind to their target antigens with high affinity and selectivity.
[0082] The antibody or antigen binding fragment that binds to the target antigen can be measured and compared by a variety of techniques, wherein the measured signal represented by a specific parameter reflects the degree of binding of the antibody or antigen binding fragment to its target antigen. In order to study the reduction in binding to the target antigen due to the masking of the antigen binding site of the antibody or antigen binding fragment, the binding signal of the "naked" antibody or antigen binding fragment (unmasked antibody or antigen binding fragment) can be compared with the binding signal of the same antibody or antigen binding fragment (masked antibody or antigen binding fragment) with masking under the same environment (e.g., at the same concentration). When compared with the naked antibody or antigen binding fragment, the difference in the obtained signal between the naked antibody or antigen binding fragment and the masked antibody or antigen binding fragment, i.e., the signal reduction of the masked antibody or antigen binding fragment, indicates that the masked antibody or antigen binding fragment shows a reduction in binding to the target antigen. In the same way, the recovery of binding can be determined by comparing the measured signal of the masked antibody or antigen binding fragment with the obtained binding signal of the antibody or antigen binding fragment in which the masking effect has been removed ("demasking" antibody or antigen binding fragment). To investigate the recovery of binding after unmasking, the signal of the unmasked antibody or antigen-binding fragment may also be compared to the signal obtained for the naked antibody or antigen-binding fragment.
[0083] One parameter that indicates the binding or affinity of an antibody or antigen-binding fragment is the dissociation constant (K D ). For antibodies or antigen-binding fragments, K D is the antibody or antigen-binding fragment dissociation rate (K off ) and the antibody or antigen-binding fragment association rate (K on), or the product of the concentration of unbound and the concentration of bound; [A]*[B] / [AB] at equilibrium, where the association and dissociation rates of the binding partners are the same. The higher the concentration of the binding partners, the more likely they are to bind to each other. D The binding partner with a high K D The binding partners will need to be in high concentrations to be able to bind to each other. D The smaller the value, the greater the binding affinity between the two binding partners. D Expressed as concentration in molar units. Most antibodies or antigen-binding fragments have low micromolar (10 -6 ) to nanomolar (10 -7 Up to 10 -9 ) in the range of K D High affinity antibodies or antigen-binding fragments are generally considered to be in the low nanomolar range (10 -9 ), while very high affinity antibodies or antigen-binding fragments are in the picomolar (10 -12 ) to the low nanomolar range. D Affected by many factors, such as buffer composition, pH and temperature. D ) can be measured, for example, using Surface Plasmon Resonance (SPR).
[0084] A technique suitable for indicative screening of antibodies or antigen-binding fragments that bind to (immobilized) target antigens is biolayer interferometry (BLI). BLI is an optical technique that measures the interference pattern of white light reflected from a sensor tip coated with a biolayer and from an internal reference layer; a sensor tip with an immobilized isolated target antigen can be contacted with a solution of an antibody or antigen-binding fragment having a predetermined concentration. When the antibody or antigen-binding fragment in the solution binds to the immobilized target antigen on the sensor tip, the optical thickness (number of binding molecules) of the biolayer on the sensor tip changes, resulting in a wavelength shift in the interference pattern of the reflected light measured in real time. The measured wavelength shift is a direct measurement of the optical thickness of the biolayer on the sensor tip (expressed in nm) and reflects the degree of binding of the antibody or antigen-binding fragment to the (immobilized) antigen. For example, BLI can be used to determine that the antibody is adequately masked; the binding of masked antibodies or antigen-binding fragments to their immobilized target antigen can be compared with the binding of unmasked or unmasked antibodies or antigen-binding fragments, and a first indication of the effectiveness of the masking can be obtained. In such a screen, when comparing the signals of naked (unmasked) and masked antibodies, a significant signal shift was observed for all masked antibodies according to the invention tested against a variety of target antigens.
[0085] The binding of an antibody or antigen binding fragment to a target cell that (over) expresses a target antigen recognized by the antibody or antigen binding fragment can be measured in vitro in a flow cytometer. Target cells expressing the target antigen can be contacted with different concentrations of the (masked) antibody, antigen binding fragment or ADC according to the invention and stained with a second fluorescently labeled antibody that binds to the first antibody or antigen binding fragment. The readout will provide the Median Fluorescent Intensity (MFI), which is also a measure of the binding of the antibody or antigen binding fragment to its target antigen expressed on the target cell. Dose-response curves (DRC) are generated by plotting different antibody or antigen binding fragment concentrations against the MFI, up to and exceeding the concentration at which maximum binding (saturation) occurs. From the dose-response curve, the EC 50 EC value (the concentration of antibody or antigen-binding fragment that gives a half-maximal response). 50 The shift (which can be expressed as the EC of the masked and naked (unmasked) antibody or antigen-binding fragment) 50 The ratio between the EC values of the unmasked antibody or antigen-binding fragment (the antibody or antigen-binding fragment from which the masking has been removed) is a measure of the reduction in target binding due to masking, provided that a full dose response curve of the masked antibody or antigen-binding fragment is available. 50 EC with naked (unmasked) antibody or antigen-binding fragment 50 By comparison, the extent of recovery of binding after unmasking can be determined. EC is typically determined by curve fitting using software such as GraphPad Prism (www.graphpad.com). 50 Using the same software tool, two EC 50 The ratio between the values (EC 50 offset or effectiveness offset).
[0086] The masked antibody or antigen-binding fragment will hardly bind to its target (which has been demonstrated to be effectively masked). Therefore, even at relatively high concentrations, saturated binding may not always be achieved in binding experiments, which can make EC 50 Determine the complexity.
[0087] EC 50 An alternative to determination is to compare the concentrations when minimum binding sets in. For masked antibodies according to the invention, the concentration at which minimum binding is observed is significantly higher than for unmasked or unmasked antibodies.
[0088] For example, for masked antibodies according to the invention (e.g., trastuzumab-based masked antibodies as shown in the experiments shown in the Examples), when tested for binding on antigen-positive cells by flow cytometry, minimal binding was observed only at very high concentrations of greater than 3 μg / mL and typically in the range of 10 to 30 μg / mL or higher, whereas minimal binding of naked or unmasked antibodies was already observed at 0.04 μg / mL.
[0089] Binding is preferably reduced by at least 10 fold, preferably 50 to 100 fold, or even 100 to 500 fold or greater than 500 fold.Since binding of the masked antibody to its target antigen expressed on the target cell is reduced, cytotoxicity will preferably be reduced by similar factors.
[0090] It is also possible to determine the reduction in cytotoxicity of the masked antibodies, antigen-binding fragments or ADCs according to the present invention to target cells compared to naked or unmasked antibodies, antigen-binding fragments or ADCs. When various masked ADCs according to the present invention were tested for their cytotoxicity, very significant IC values were observed compared to naked antibody-based ADCs. 50 When compared to the IC of the same ADC based on naked antibody 50 When compared to the IC 50 The shift was more than 30-fold shift, and even 500-fold shift. These findings indicate a very efficient masking of the binding regions of the masked antibodies according to the invention.
[0091] It is also important that, following unmasking (i.e., removal of the mask), binding to the target and thereby cytotoxicity is restored to levels sufficient for the unmasked antibody or antigen-binding fragment to have the desired therapeutic effect. Preferably, the unmasked antibody or antigen-binding fragment binds to the target antigen as effectively as the unmasked antibody or antigen-binding fragment. Thus, if the unmasked antibody or antigen-binding fragment binds to its target with an affinity in the high nanomolar range, then preferably, the unmasked antibody or antigen-binding fragment does the same. Preferably, the unmasked antibody binds to its target antigen exactly as the corresponding naked antibody binds to its target antigen. Some reduction in binding of the unmasked antibody when compared to the naked antibody is tolerable. Preferably, the binding grade of the unmasked antibody, expressed, for example, as EC 50 The EC between the unmasked and unmasked antibodies was 0.1477 and 0.1317, respectively. The EC between the unmasked and unmasked antibodies was 0.1477 and 0.1317, respectively. The EC between the unmasked and unmasked antibodies was 0.1477 and 0.1317, respectively. 50The shifts were generally in the range of 1 to 2, indicating restoration of binding after unmasking. The same was true for the cytotoxicity of the unmasked antibody; an almost complete restoration of the induction of cytotoxicity after unmasking was observed when compared to the naked antibody.
[0092] In the masked antibodies or antigen-binding fragments according to the present invention, the mask comprises a pair of peptide masking moieties having specific binding affinity for each other, wherein one masking moiety of the pair comprises an antigenic peptide and the other masking moiety comprises a single domain antibody (sdAb) that specifically binds to the antigenic peptide. The masking moiety is connected to the N-terminus of at least one HC and LC pair of the masked antibody or antigen-binding fragment. The masked antibodies or antigen-binding fragments according to the present invention are effectively masked and, once they reach their target site, they are effectively unmasked.
[0093] Masking part
[0094] The masking moiety is the actual (peptide) part that (reversibly) binds to the antibody or antigen-binding fragment thereof, which can form a mask when conjugated to the antibody or antigen-binding fragment. The masked antibody or antigen-binding fragment thereof according to the present invention has a mask comprising a set of two peptide masking moieties connected to the N-terminus of the heavy chain variable region (HCVR) and the light chain variable region (LCVR) that together form an antigen binding site. One of the two masking moieties of the masking moiety pair comprises an antigenic peptide, while the other comprises a single domain antibody (sdAb) that specifically binds to the antigenic peptide. At least one masking moiety is connected to the N-terminus of the HCVR and / or LCVR via a cleavable linker. By binding to each other, the masking moieties together form a mask, thereby shielding the antigen binding site of the masked antibody or antigen-binding fragment. When the antigen binding site of the antibody or antigen-binding fragment is shielded by the mask, the ability of the masked antibody or antigen-binding fragment to bind to its target antigen is significantly reduced. The antibody or antigen-binding fragment can be unmasked by cleaving at least one cleavable linker. Thus, masking of the antibody or antigen-binding fragment is reversible; the masking effect can be removed by cleaving at least one cleavable linker. Thus, demasking does not require breaking the bond between the sdAb and the antigenic peptide (the sdAb and the antigenic peptide together constitute the mask) (but once at least one linker is cleaved, the masking moieties may no longer remain bound to each other).
[0095] The antigenic peptide used in the antibody masking moiety of the present invention can be any suitable antigenic peptide as part of a sdAb / peptide pair. A peptide is selected that contains an epitope sequence (preferably a single epitope sequence) specifically recognized by the sdAb that is in the other masking moiety of the masking moiety pair that together forms a mask.
[0096] Preferably, the selected antigenic peptide sequence for masking part is relatively short (the peptide tag used in the labeling technology based on sdAb known in the art), so that antibody generation is not hindered and effective masking and demasking are allowed.Usually, the length of such antigenic peptide sequence is about 4 to 20 amino acid, preferably 8 to 15 amino acid.More preferably, the length of antigenic peptide sequence is 12 to 15 amino acid residues.Preferably, such peptide derives from human protein, so that the immunogenic possibility of antigenic peptide is minimized.
[0097] Single domain antibodies (sdAbs) are (recombinant) antigen-binding fragments with a unique structure consisting of a single heavy chain with only one variable domain. The first sdAbs developed were V chains of heavy chain antibodies derived from the Camelidae family (e.g., llama, camel, dromedary, Bactrian camel, alpaca, vicuna, and guanaco). H H fragment (variable domain) fragment. (S. Muyldermans et al., 2009, Veterinary Immunology and Immunopathology 128, 178-183). Heavy chain antibodies circulating in camelids contain only heavy chains and naturally do not contain light chains. sdAbs derived from these heavy chain-only antibodies of camelids are usually called "nanobodies" or "V H H”(Muyldermans S., 2013, annual 82:775-97. doi:10.1146 / annurev-biochem-063011-092449). The term "nanobody" was originally adopted by the Belgian company Ablynx because these sdAbs have nanomolecular size ("NANOBODY" is a registered trademark of Ablynx NV), and the term nanobody is often used as an equivalent of sdAbs, also covering sdAbs from other species. In addition to camelids, sdAbs are also found in cartilaginous fish (such as sharks). Sharks and other cartilaginous fish also produce heavy chain antibodies. These antibodies are called Ig new antigen receptors or IgNARs. IgNARs are composed of two identical heavy chains. The variable domains in these heavy chains are called V-NAR domains. IgNARs also have single variable domains. Unlike the V-NAR domains of camelid sdAbs, the V-NAR domains of the sdAbs of the camelid family are also called V-NAR domains. HCompared with the three CDRs in the H domain, the V-NAR domain contains two CDRs (Cheong et al., 2020, International journal of biological macromolecules, 147, 369-375). sdAb can also be obtained from common (human) IgG antibodies (Chen et al., 2009, Methods Mol Biol., 525: 81-99).
[0098] The sdAb used for the masking part according to the present invention can be humanized by methods known in the art (Vincke et al., 2009, The Journal of Biological Chemistry Vol. 284, No. 5, 3273-3284). The sdAb used for the masking part in the masking antibody or antigen-binding fragment according to the present invention is preferably a nanobody or a humanized form and / or variant thereof.
[0099] sdAbs can be selected and produced in a variety of ways known in the art. Methods for selecting nanobodies are reviewed in S. Muyldermans, 2021, The FEBS Journal, 288, 2084-2102; Hassanzadeh-Ghassabeh et al., 2013, Nanomedicine (Lond), 8(6): 1013-26.
[0100] Compared to common antibodies (150 to 160 kD), which are composed of two HC / LC pairs, sdAbs are small (12 to 15 kD). sdAbs are also smaller than antigen-binding fragments of common antibodies, such as the Fab fragment consisting of one (half) HC and LC, which has a molecular weight of about 50 kD. sdAbs are the smallest antibody fragments that still selectively bind to a specific target antigen.
[0101] sdAbs have been used for a variety of purposes, for example, for treatment (e.g., as a targeting moiety for cytotoxic drugs), but also for diagnostic and imaging applications, where nanobodies are linked to tracers (de Meyer et al., 2014, Trends in Biotechnology, Vol. 32, No. 5, 263-270; Bao et al., 2021, EJNMMI Res 11: 6, https: / / doi.org / 10.1186 / s13550-021-00750-5). For example, sdAbs against certain peptide tags have been developed for immunoprecipitation of labeled proteins. Such tags include tags developed by, for example, ChromoTek, Nanotag Biotechnologies, etc. Ren et al., 2020, Journal of Chromatography A, 1624, 461227 also describe peptide tag-specific antibodies.
[0102] In a preferred embodiment of the invention, the single polypeptide chain that comprises the variable domain of the entire antigen-binding capacity of a Nanobody preferably has an amino acid sequence and structure that can be considered to be composed of four framework regions or "FRs", which are referred to in the art and herein as "Framework Region 1" or "FR1", "Framework Region 2" or "FR2", "Framework Region 3" or "FR3", and "Framework Region 4" or "FR4", respectively; wherein the framework regions are interrupted by three complementarity determining regions or "CDRs", which are referred to in the art as "Complementarity Determining Region 1" or "CDR1", "Complementarity Determining Region 2" or "CDR2", and "Complementarity Determining Region 3" or "CDR3", respectively. These framework and complementarity determining regions are preferably operably linked in the order FR1-CDR1-FR2-CDR2-FR3-CDR3-FR4 (from amino terminus to carboxyl terminus).
[0103] An example of a peptide tag developed against which sdAbs are directed is a peptide derived from human β-catenin, such as the BC2T peptide PDRKAAVSHWQQ (SEQ ID NO: 1) representing a conserved epitope in human β-catenin, and nanobodies described therein in Traenkle et al., 2015, Molecular & Cellular Proteomics, 14 (3), 707-723 and EP3377632. This peptide sequence forms the basis for the following: The peptide tag is an inert 12 amino acid peptide tag having the amino acid sequence PDRVRAVSHWSS (SEQ ID NO: 44); and the Spot-Nanobody that specifically binds to the Spot-labeled protein is provided by ChromoTek.
[0104] sdAbs that recognize peptide tags of only 4 or 5 amino acids (AA) used as affinity tags are commercially available (e.g., based on the 4AA peptide tag EPEA). C-tag matrix, and a corresponding 13 kDa camelid antibody fragment (affinity ligand) consisting of a 13 kDa camelid antibody fragment with high affinity and selectivity for the 4-amino acid "C-tag" peptide tag EPEA (SEQ ID NO: 45) Biotin anti-C-tag conjugate, which was provided by ThermoFisher). Another tag sequence and corresponding sdAb is the "ALFA-tag", such as et al., 2019, Nature Communications, 10: 4403, https: / / doi.org / 10.1038 / s41467-019-12301-7. The amino acid sequence of the antigenic peptide of the minimum ALFA-tag is SRLEEELRRRLTE (SEQ ID NO: 46). This sequence is capable of forming a stable alpha helix. An sdAb specific for the ALFA-tag, namely NbALFA, was developed and binds to the peptide with low picomolar affinity. ALFA-tags and their variants as well as sdAbs are also mentioned in WO2020 / 053239 (Nanotag Biotechnologies).
[0105] Nanobodies generally have low immunogenicity due to their small size and high sequence identity with human IGHV3 family gene products, which means that they produce almost no anti-drug antibodies (ADA) (Ackaert C et al., 2021, Front. Immunol. 12: 632687, doi: 10.3389 / fimmu.2021.632687; Ren et al., supra; et al., 2020, BioDrugs, 34, 11-26; Rossotti M et al., 2021, FEBS J., doi: 10.1111 / febs.15809).
[0106] sdAbs are known to have relatively high affinities for their target antigenic peptides (low nanomolar to picomolar range), which is preferred in many known applications where binding partners are required even at very low concentrations (low K D ) should also be combined with each other. In the present invention, sub-nM K DNanobodies with a (picomolar) affinity for binding to ALFA are generally in the low picomolar (10 to 30 pM) range. For example, the KD of NbALFA for binding to its ALFA tag is D The value is about 26 pM. The affinity of BC2 Nb to its BC2T peptide tag is 1.4 nM (1400 pM).
[0107] The use of high affinity sdAb / antigenic peptide pairs results in efficient masking and unmasking of masked antibodies or antigen-binding fragments according to the invention. In masked antibodies or antigen-binding fragments according to the invention, the masking moieties are brought into close proximity to each other due to their respective N-terminal attachment to the light and heavy chains within the (masking) antibody. Due to their close proximity, the masked masking moieties can bind to each other, even if they have relatively low affinity for each other. The high affinity between Nanobodies and peptide tags, which are preferred for other applications, may therefore not be a prerequisite for their use as masking moieties in masked antibodies, antigen-binding fragments or ADCs according to the invention.
[0108] In the context of the present invention, it may be advantageous to use sdAbs with relatively low affinity for antigenic peptide sequences that are specifically recognized. The advantage of using sdAbs with relatively low affinity for antigenic peptides is that the ability of the masking moiety to form intermolecular bonds (i.e., bonds between masking moieties on the HC / LC of different antibodies or antigen-binding fragments, which can lead to HMW complex formation) can be reduced. The percentage of HMW-forming antibodies to the total amount of antibodies produced (HMW (%)) is preferably less than 10%.
[0109] At the same time, when conjugated to a HC and LC pair, their ability to form an intramolecular bond (i.e., a bond between the masking moieties on the heavy and light chain pair within the masked antibody or antigen-binding fragment) is maintained at a lower affinity due to the closer proximity of the two masking moieties.
[0110] Relatively low affinity means significantly (e.g., 2 to 100-fold and at least 2 to 10-fold) lower than the high (single) nanomolar to picomolar range of affinity commonly seen for sdAb epitope binding. Thus, the K for binding between Nb / peptide pairs used in the present invention for masking moieties is D The affinity can range from 10 pM (high affinity) to relatively low affinity, where K D Can be as high as 500nM (0.01nM to 500nM).
[0111] Preferably, the sdAb and antigenic peptide are selected such that the K of the affinity between the sdAb and the antigenic peptide is D The value is in the range of 0.1 to 500 nM, more preferably in the range of 1 to 100 nM.
[0112] In generating sdAb groups for certain epitope sequences, sdAbs with lower affinity for antigenic peptides can be selected. In an alternative, certain sdAbs originally selected for their high affinity for antigenic peptides can be modified (e.g., by introducing certain point mutations in the binding region of the sdAb to generate variant sdAbs with slightly altered amino acid sequences, thereby reducing their affinity for the peptide. Similarly, the antigenic peptide sequence recognized by the sdAb with high affinity can be changed to reduce the binding to the same sdAb.
[0113] As described in Traenkle et al., 2015, Molecular & Cellular Proteomics, 14(3), 707-723, good results were obtained with a β-catenin-based peptide (BC2T) and its nanobody (BC2)-based masking moiety. The BC2T peptide described by Traenkle et al. has an amino acid sequence PDRKAAVSHWQQ (SEQ ID NO: 1) representing a conserved linear epitope corresponding to amino acid residues 16 to 27 of β-catenin, while additional variants retaining this epitope are described in EP3377632, which discloses a peptide having the following sequence: X 1 X 2 RX 4 X 5 AX 7 SX 9 WX 11 X 12 , where X 1 Can be P or A, where X 2 It can be a conservative substitution of D or D, where X 4 Can be K or a conservative substitution of K or S, where X 5 It can be A or R, or a conservative substitution of A or R, wherein X 7 It can be a conservative substitution of V or V, where X 9 It can be a conservative substitution of H or H, where X 11 and X 12 Independently of each other, they can be Q or conservative substitutions of Q.
[0114] The term "conservative substitution" as used herein means that one amino acid is replaced by another amino acid, wherein the replacement results in a silent alteration. This means that an amino acid residue can be replaced by another amino acid of similar polarity that acts as a functional equivalent. The replacement of an amino acid can be selected from other members of the class to which the amino acid belongs (i.e., conservative substitution). For example, respectively, a polar amino acid can be replaced by another polar amino acid, and a positively charged or negatively charged amino acid can be replaced by another positively charged or negatively charged amino acid, etc. The categories of amino acids are, for example: non-polar (hydrophobic) amino acids, including alanine, leucine, isoleucine, valine, proline, phenylalanine, tryptophan and methionine; polar neutral amino acids, including glycine, serine, threonine, cysteine, tyrosine, asparagine and glutamine; positively charged (basic) amino acids, including arginine, lysine and histidine; negatively charged (acidic) amino acids, including aspartic acid and glutamic acid.
[0115] EP3377632 also discloses specific variant peptides showing fixed amino acids at positions 3 (R), 6 (A), 8 (S) and 10 (W) which are believed to be necessary for epitope binding, such as PVRSAALSQWSS (SEQ ID NO: 48), PDRVRAVSHWSS (SEQ ID NO: 49) and ADRVRAVSHWS S (SEQ ID NO: 50).
[0116] Traenkle et al. also described nanobodies that specifically bind to the BC2T peptide defined by their CDR3 sequences. Among their best performing nanobodies (BC1, BC2, BC6, BC9, and BC13), BC1 and BC2 bound to recombinant β-catenin with a K of 1.3, as measured by surface plasmon resonance spectroscopy. D In the low nanomolar (about 1.9 nM and about 3.1 nM) range. BC1 3 showed a slightly lower affinity of about 44 nM. For BC6 and BC9, the detected affinities were in the low micromolar range. The CDR3 sequence of the BC2 nanobody is: ARGCKRGRYEYDFW (SEQ ID NO: 5).
[0117] The BC2T peptide and its variants, as well as BC2-Nb (and its variants) that specifically bind to the BC2T or variant peptide sequence, can be used to mask moieties to produce masked antibodies, antigen-binding fragments and / or ADCs according to the present invention.
[0118] However, for prior art uses of peptide tags and sdAb sequences, pairs are selected for their high affinity binding, but very high affinity may not be required for use in the masking moiety in a masked antibody, antigen binding fragment or ADC according to the present invention. It has been found that good results are obtained by using antigen peptide / sdAb pairs with relatively low affinity (in the range of 1 to 100 nM) for each other. A variant of the BC2 nanobody is described in Braun et al., 2016, Sci Rep 6, 19211, https: / / doi.org / 10.1038 / srep19211. Braun et al. replaced Arg106 with serine (BC2-Nb R106S or "RS" mutant (SEQ ID NO: 8)) or glutamate (BC2-Nb R106E or "RE" mutant, (SEQ ID NO: 7)) to mutate the CDR3 region of the BC2 nanobody. In doing so, the nanobody-peptide conformation of the BC2 nanobody was disrupted ("headlock"), resulting in a high affinity binding to BC2-Nb (K D The binding affinity (K D The value was about 11 nM; the RE mutant showed a K of 12 nM D and the RS mutant showed a K of 9.7 nM D ), as measured using surface plasmon resonance spectroscopy (SPR)-based affinity measurements. K of RA mutants D Described in Ren et al., Journal of Chromatography A 1676 (2022) 463274. Ren et al. provided the same K value of 1.4 nM for wt BC2-Nb as Braun et al. D and compared it with the R106Q mutant (K D =282 nM), R106A mutant (K D =360.5 nM) and R106H mutant (K D =335.5 nM) for comparison.
[0119] The original BC2-Nb as well as RS and RE mutants and mutants in which arginine was replaced by alanine (BC2-Nb R106AFor example, when used as a masking moiety for the anti-HER2 antibody trastuzumab, the use of mutant Nb resulted in a significantly lower (5 to 10-fold lower) percentage of HMW complexes, while the ability of the masking moiety to shield the antibody binding site was retained until at least one linker was cleaved.
[0120] In an alternative or additional approach, in order to modify the sdAb sequence to adjust the affinity of the high affinity peptide / sdAb pair, the sequence of the antigenic peptide may be modified with the goal of reducing the affinity of the peptide-sdAb interaction. For example, in the case of the BC2T peptide, the serine at position 8 fixed in the molecular formula of the peptide sequence in EP3377632 may be replaced by an alanine or threonine residue, for example, in the peptide used in the masking moiety of the present invention. When S(8) is replaced by X 8 When replaced, this changes the general formula of the peptide to X 1 X 2 RX 4 X 5 AX 7 X 8 X 9 WX 11 X 12 (SEQ ID NO: 51), wherein X 1 Can be P or A, where X 2 It can be a conservative substitution of D or D, where X 4 Can be K or a conservative substitution of K or S, where X 5 It can be A or R or a conservative substitution of A or R, wherein X 7 It can be a conservative substitution of V or V, where X 8 Can be S, A, V or T, where X 9 It can be a conservative substitution of H or H, where X 11 and X 12 It can be Q or a conservative substitution of Q. In the variant disclosed in EP3377632, the fixed "R" at position 3 can additionally be replaced by "K", resulting in the following formula: X 1 X 2 X 3 X 4 X 5 AX 7 X 8 X 9 WX 11 X 12 (SEQ ID NO: 95), wherein X 1 , X 2 , X 4 , X 5 , X 7 , X 8, X 9 , X 11 and X 12 has the aforementioned meaning, and X 3 Can be R or K.
[0121] In preferred peptides according to the present invention, X 8 is A or T, most preferably A, resulting in, for example, a peptide having the sequence PDRKAAVAHWQQ (BC2TSA mutant, SEQ ID NO: 2) or PDRKAAVTHWQQ (BC2TST mutant, SEQ ID NO: 3), preferably the peptide PDRKAAVAHWQQ.
[0122] Variant peptides in which the S at position 8 is replaced by A or T at position 8, such as the BC2TST mutant or the BC2TSA mutant, can be used in pairs of masking moieties together with a masking moiety comprising BC2-Nb, or preferably, with a masking moiety comprising a variant of said BC2-Nb moiety having a mutation in the CDR3 region, such as the BC2RE, RS or RA mutant. Good results were obtained with pairs of masking moieties comprising a masking moiety with a BC2TSA mutant peptide and a masking moiety with a BC2-Nb RS mutant Nanobody.
[0123] WO2020 / 053239 also describes another sequence variant of the ALFA-tag of a nanobody that differs from the sdAb in its affinity. D is about 1 to 5 nM, or when the core structure is different, the K for binding to the same sdAb D An example of a modified ALFA-tagged Nb is Nb disclosed as SEQ ID NO: 134 in WO2020 / 053239. PE Mutants. et al., 2019, Nature Communications, 10:4403, https: / / doi.org / 10.1038 / s41467-019-12301-7, describe that this mutant was designed for protein elution. d When Nb is 26pM) and the tag peptide is bound, PE The mutant (“PE” for peptide elution) binds with lower affinity (K d is 11 nM), such as Supplementary Figure 6As shown in FIG. 1 . Using a protease-cleavable linker to connect the Nb in the masking portion to the antibody PE Mutant and ALFA-tagged peptides, antibodies can be efficiently masked and unmasked.
[0124] In an alternative or in addition to using low affinity sdAb / antigenic peptide pairs as masks to prevent HMW complex formation, prior art methods for removing or reducing HMW complex formation during protein production, purification and storage may also be used.
[0125] Connectors
[0126] At least one masking moiety (antigenic peptide or sdAb or both) may be linked to the antibody via a cleavable linker, preferably a conditionally cleavable peptide linker that is cleaved only at the target site (site where the antibody should bind to its target antigen and the masking should be removed, e.g., in or at a tumor site). Preferably, the cleavable peptide linker is a protease cleavable peptide linker. Preferably, both masking moieties are linked to the antibody via a cleavable linker.
[0127] When only one linker is cleaved, the binding capacity of the antibody or antigen binding fragment may have been fully restored so that the antibody has its desired therapeutic effect. Preferably, all linkers connecting the masking moiety to the antibody or antigen binding fragment are cleaved, and after cleavage, the mask is completely detached from the antibody.
[0128] Peptide linkers for connecting different parts of the fusion protein are known in the art. Such linkers can be flexible linkers, rigid linkers, cleavable linkers, or combinations (Chen et al., 2013, Adv Drug Deliv Rev., 65(10), 1357-1369). The cleavable linker for masking antibodies or antigen-binding fragments according to the present invention is preferably a peptide linker comprising one or more cleavage sites recognized by one or more proteases.
[0129] When the masked antibody is used for cancer treatment, the cleavable peptide linker should preferably contain at least one cleavage site recognized by at least one tumor-specific protease. Some examples of such proteases known in the art are proteinase (MT-SP1, tumor-associated type II transmembrane serine protease), cathepsin B, urokinase-type plasminogen activator (uPA) and / or metalloproteinase (MMP), such as MMP2, MMP9 and MMP14, which are overexpressed in almost all types of cancer (Li Yanan et al., 2021, Acta Pharmaceutica Sinica B, 11 (8), 2220-2242). Linkers comprising tumor-specific protease cleavage sites are known in the art, also in the context of antibody masking technology, for example WO2018 / 107125 (Seattle Genetics), WO2009 / 025846, WO2010 / 081173 and WO2016 / 118629 (Cytomx Therapeutics LLC) and Choi et al., 2012, Theranostics, 2(2), 156-178.
[0130] The cleavage site is the (minimum) amino acid sequence recognized by the protease, and the actual cleavage will be carried out within this site. An example of a cleavage site known in the art is the MMP cleavage site PLGLAG, SEQ ID NO: 13, which can be cleaved by MMP2 and MMP9 (Jiang et al., 2004, PNAS, 101 (51), 17867-17872; Trang et al., 2019, Nature biotechnology, 37 (7), 761-765.). Other known cleavage sites are, for example: the cleavage site LSGRSDNH (SEQ ID NO: 14) recognized by urokinase-type plasminogen activator (uPA) (EP2385955, Cytomx Therapeutics LLC), and the cleavage site of MMP14 with sequence ISSGLL (SEQ ID NO: 15) (WO2016 / 118629, Cytomx Therapeutics LLC). Other cleavage sites recognized by proteases are described in WO2017 / 162587 (Hoffmann-La Roche AG), such as RQARVVNG (SEQ ID NO: 53) and PMAKK (SEQ ID NO: 54). Other cleavage sites: PLGVR (SEQ ID NO: 70) and IPVSLR (SEQ ID NO: 52) or IPVSLRSG (SEQ ID75) are MMP2 cleavable sites mentioned, for example, in WO2018107125 (Seattle Genetics Inc.) and Lin et al. supra. Preferred cleavage sites for masking antibodies or antigen-binding fragments according to the present invention are PLGLAG, IPVSLR (SG), LSGRSDNH and / or PMAKK.
[0131] Preferably, both masking moieties are connected to a linker comprising at least one protease cleavage site. Linkers cleaved by multiple proteases can be used to increase activation at target sites, such as tumors, where at least one masking moiety should be removed to restore the binding of the unmasked antibody. Such a linker may comprise a cleavage site recognized by multiple proteases and / or different cleavage sites each recognized by a different protease. An example of a cleavable linker sequence comprising multiple cleavage sites recognized by different proteases is, for example, "2001" described in WO2016 / 118629 (Cytomx Therapeutics LLC). The "2001" sequence has the AA sequence ISSGLLSGRSDNH (SEQ ID NO: 16), and combines the ISSGLL cleavage site for MMP14 with the LSGRSDNH cleavage site recognized by proteases and uPA. The cleavage sites can be combined in different orders (WO2016 / 118629).
[0132] The two linkers in the masking pair of masking moieties forming the mask may comprise the same (identical) or different (non-identical) cleavage sites. A preferred combination of linkers for conjugating the masking pair of masking moieties to the HC / LC pair of the antibody or antigen-binding fragment is a linker combination in which both linkers comprise a cleavage site selected from the group consisting of PLGLAG, LSGRSDNH, IPVSLR (SG) and PMAKK.
[0133] Preferred linker combinations include the following combinations: a linker comprising a cleavage site recognized by a protease for conjugating a masking moiety with an SdAb to an antibody, preferably to the LC VR, and a linker comprising a cleavage site recognized by a metalloprotease, such as MMP2, to conjugate a masking moiety with an antigenic peptide to an antibody or antigen-binding fragment, preferably to the HC VR of the same HC / LC binding pair. The masking moiety conjugated to the LCVR of at least one HC / LC chain pair preferably comprises a BC2-Nb RS mutant nanobody and is conjugated to the LCVR via a linker comprising a cleavage site recognized by a metalloprotease, preferably PLGLAG, while the masking moiety conjugated to the HCVR of the HC / LC chain pair comprises a BC2T SA mutant peptide and is conjugated to the HC via a linker comprising a cleavage site recognized by a protease, preferably LSGRSDNH. In an alternative, the cleavage site recognized by a metalloprotease may be ISSGLL or IPVSLR (SG), while an alternative to the protease cleavage site may be PMAKK.
[0134] In addition to the actual protease cleavage site, the linker used to connect the masking moiety to the HC and / or LC of the masked antibody or antigen-binding fragment according to the invention may contain additional spacer sequences, such as serine and / or glycine-rich stretches, mainly to increase the flexibility of the linker, for example to provide sufficient movement for the masking moieties to adopt the appropriate orientation to bind to each other and to optimize the degree of shielding of the actual mask. By varying the spacer sequence (e.g. in length) and the position of the cleavage site within the linker, the linker can be optimized for use with a specific antibody.
[0135] Commonly used flexible linkers are linkers containing a glycine-rich region, such as -GGS- or -(GGGGS) n -, (Chen et al., supra). In the linker for masking antibodies according to the present invention, such a glycine ( / serine)-rich spacer sequence can be connected to the actual protease cleavage site at one or both ends of the cleavage site sequence within the cleavable linker. The separated cleavage recognition sequences can be directly connected, or they can be separated (and / or flanked) by G / S-rich sequences, and the separated cleavage sites can be placed in different orders (for example, ISSGLLSSGGSGGS.LSGRSDNH (SEQ ID NO: 17), or LSGRSDNHGGSGGSISSGLLSS (SEQ ID NO: 18) mentioned in WO2016 / 118629).
[0136] Thus, preferred linkers comprise one or more proteolytic cleavage sites flanked by one or more spacer sequences. Most preferred are linkers with no or only short flanking sequences between the N-terminus of the HCVR or LCVR of the antibody or antigen-binding fragment sequence and the (first) cleavage site. Depending on the position at which the linker is cleaved, a portion of the linker sequence may of course remain attached to the N-terminus of the HC and / or LC of the unmasked antibody after cleavage of the linker. When the cleavage site is close to the N-terminus of the HC and / or LC sequence, after the linker is cleaved, no or only a few amino acids will remain attached to the HC and / or LC. If there is a spacer sequence between the N-terminus of the HC and / or LC and the cleavage site, such a sequence is preferably short, preferably 0 to 10 or 1 to 5 amino acids in length, for example 3 amino acids in length, such as the amino acid sequence -GGS-. Additional spacer sequences may be included at the N'-terminus of the cleavage site sequence, such as -GGS- or -GGGGS- spacer sequences. The resulting linker may, for example, have the sequence -GGGGS PLGLAGGGS-(SEQ ID NO: 11), wherein the MMP cleavage site is underlined. The resulting linker can be of any suitable length, depending on the length and number of cleavage recognition sites used and the length of the flexible linker placed before and / or after the cleavage site sequence. Suitable linkers can generally contain 2 to 30 amino acids, preferably 5 to 25 amino acids.
[0137] Although antibodies or AFs have the same overall conformation, each corresponding antibody or AF has different antigen binding regions. Antigen binding regions may be different in sequence, and therefore different in charge distribution and 3D conformation. Although the present invention provides an ingenious and unified method for masking antibodies or AFs, good results can be obtained by using the same masking moiety on each antibody or AF in principle, but adjusting the length of the joint can lead to, for example, higher yields during production. For this effect, joint groups of different lengths can be prepared and tested in different combinations of pairs as the masking moiety is connected to LC and HC. When testing different combinations, the joint group of best performance can be selected. "Best performance" can be defined based on a variety of measurement parameters. For example, when compared with other joint pairs tested, the expression level is high and the HMW (%) produced is very low while still maintaining the joint pair of masking and demasking efficiency.
[0138] For example, the linker can be selected from the group shown in Table 18 (as tested for a specific antibody, tisomonab, in Example 10). In the group shown in Table 18, linkers L14 to L25 all contain the protease cleavage site LSGRSDNH but are of different lengths, while linkers L26, L27, L9, L29, L7, and L30 all contain the PLGLAG protease cleavage site and are also of different lengths. This linker group covers 324 possible combinations and allows testing of linker combinations with the same or different cleavage sites for two masking parts, while the lengths of the linkers can also be the same or different. In Example 10, which is an example of testing all 324 linker combinations, it is illustrated that most can be produced at an expression level greater than 100 μg / mL. In most cases, the HMW (%) is less than 10%, and for 198 linker combinations, the HMW (%) is less than 5%, and for 39 even less than 3%.
[0139] A preferred subset of linker combinations for generating these 324 antibodies is the subset of linker combinations used to generate ADCs 1 to 12 shown in Table 22 of Example 11.
[0140] For any antibody or antigen-binding fragment to be masked, this operation can be repeated with all or only a subset of the possible combinations selected from Table 18. The present invention encompasses masked antibodies or antigen-binding fragments thereof, wherein the masking moiety is conjugated to at least one HC / LC pair of the antibody or antigen-binding fragment using any suitable linker combination selected from the linkers shown in Table 18.
[0141] A spacer sequence such as QGQSGQG (Qg) (SEQ ID NO: 9) or QVQLVES (Qv) (SEQ ID NO: 10) may also be included in the expressed protein sequence, for example, at the N-terminus of the antigenic peptide shielding the moiety (while the linker is attached to the C-terminus of the peptide).
[0142] In the NC sequence, when the peptide is linked to the HC, the construct may thus have the following conformation: Qg-antigenic peptide-linker-HC.
[0143] Qg spacer sequences are known in the art (e.g., Cytomx WO2016 / 179285).
[0144] The Qv sequence is based on a conserved sequence present in the framework 1 region of the heavy chain variable domain (VH) sequence in many human antibodies. By introducing a spacer sequence such as a Qg or Qv sequence before the amino acid sequence of the antigenic peptide (at the N-terminus), the production titer of the masked antibody according to the present invention can be increased to 10 times, regardless of whether the masked portion with the antigenic peptide is connected to the N-terminus of the HC or LC. Good results were obtained with a masked antibody in which the Qg or Qv sequence was located before the BC2T peptide sequence (variant).
[0145] In the masked antibody according to the present invention, at least one of the two masking moieties may be connected to the antibody or antigen binding fragment via a cleavable linker. When the binding capacity of the masked antibody or antigen binding fragment can be fully restored by cleaving the bond between one masking moiety and the antibody or antigen binding fragment, it is sufficient to connect only one masking moiety with a cleavable linker. When only one of the masking moieties is connected to the HC or LC of the antibody or antigen binding fragment via a cleavable linker, and the linker is cleaved, the masking will be removed from the antigen binding site of the antibody or antigen binding fragment (its binding site will be "unmasked"), although in fact the masking moiety may still be connected to the antibody or antigen binding fragment. When only one masking moiety is connected via a cleavable linker, the masking moiety is preferably a masking moiety comprising an sdAb.
[0146] Due to the masking of the small size used in the present invention, the actual binding site of the antibody or Fab can be used for binding to its target antigen, and the masking part no longer prevents the antibody or Fab from binding to its target, even when only one joint is cut. When one of the masking parts, i.e., sdAb or antigenic peptide, is combined by a cleavable (peptide) joint, another masking part can also be connected by a peptide joint, or, in an alternative, can be put together with the antibody or Fab. For example, a fusion protein comprising an antibody or Fab fused with an sdAb can be produced, and the antigenic peptide used as the second masking part can be put together with the fusion protein. In an alternative, sdAb and the antigenic peptide are both included in the (fusion) protein sequence produced by recombinant production, and one or both are connected by a (cleavable) peptide joint that is a part of the expressed fusion protein.
[0147] In the masked antibody or antigen-binding fragment according to the present invention, the antigenic peptide and sdAb can be linked to the N-terminus of at least one of the light chain and heavy chain pairs of the antibody or antigen-binding fragment. In a bivalent antibody or antibody fragment (e.g., F(ab') 2 fragment), preferably all (two) HC / LC pairs are masked.
[0148] In the masked antibody or antigen-binding fragment according to the invention, the masking moiety comprising the antigenic peptide may be linked to the N-terminus of the LC, while the sdAb may be linked to the N-terminus of the HC, or vice versa.
[0149] From a production perspective, the use of masked antibodies may be advantageous, wherein preferably the masking moiety with the sdAb is linked to the N-terminus of the LC and the antigenic peptide is linked to the N-terminus of the HC. In the experiments performed, these antibodies were produced with higher titers (and lower HMW (%)).
[0150] The masked antibodies or antigen-binding fragments according to the present invention are particularly suitable for use in antibody-drug conjugates because the masking does not interfere with the conjugation of the masked antibody or antigen-binding fragment to the linker-drug. The targeting antibody or antigen-binding fragment that can be used in the antibody-drug conjugate for cancer treatment can be a tumor-targeted antibody or antigen-binding fragment that selectively binds to a tumor-specific or tumor-associated antigen. Tumor-specific antigens only appear on tumor cells, while tumor-associated antigens are antigens that are expressed (e.g., overexpressed) at higher levels in cancer cells when compared to normal (healthy) cells.
[0151] The antigen target (also referred to herein as "target protein") to which the masked antibody or antigen-binding fragment according to the present invention binds can be selected from: annexin A1, B7H3, B7H4, BCMA, CA6, CA9, CA15-3, CA19-9, CA27-29, CA125, CA242 (cancer antigen 242), CAIX, CCR2, CCR5, CD2, CD3, CD19, CD20, CD22, CD24, CD28, CD30 (tumor necrosis factor 8), CD33, CD37, CD38 (cyclic ADP ribose hydrolase), CD40, CD44, CD47 (integrin-associated protein), CD56 (neural cell adhesion molecule), CD70, CD71, CD73, CD74, CD79, CD11 15 (colony stimulating factor 1 receptor), CD123 (interleukin-3 receptor), CD138 (syndecan 1), CD203c (ENPP3), CD303, CD333, CDCP1, CEA, CEACAM, claudin 4, claudin 7, CLCA-1 (C-type lectin-like molecule-1), CLL1, c-MET (hepatocyte growth factor receptor), Cripto, CTLA4, DLL3, EGFL, EGFR, EPCAM, EphA2, EphB3, ETBR (endothelin type B receptor), FAP, FcRL5 (Fc receptor-like protein 5, CD307), FGFR3, FOLR1 (folate receptor α), Frβ, GCC (guanylate cyclase C), GD2, GITR, GLOBO H, GPA33, GPC3, GPNMB, HER2, p95HER2, HER3, HMW-MAA (high molecular weight melanoma-associated antigen), integrin α (e.g., αvβ3 and αvβ5), IGF1R, TM4SF1 (L6), Lewis A-like carbohydrate, Lewis X, Lewis Y (CD174), LGR5, LIV1, mesothelin (MSLN), MN (CA9), MUC1, MUC16, NaPi2b, connexin-4, Notch3, PD-1, PD-L1, PSMA, PTK7, SLC44A4, STEAP-1, SIRP-α, 5T4 (or TPBG, trophoblast glycoprotein), TF (tissue factor, thromboplastin, CD142), TF-Ag, Tag72, TNFα, TNFR, TROP2 (tumor-associated calcium signal transducer 2), uPAR, VEGFR and VLA.
[0152] If applicable, the masked antibody or antigen-binding fragment thereof may comprise: (1) an engineered constant region, i.e., one or more mutations may be introduced to, for example, increase half-life, provide a linker-drug attachment site, and / or increase or decrease effector function; and / or (2) an engineered variable region, i.e., one or more mutations may be introduced to, for example, provide a linker-drug attachment site. The antibody or antigen-binding fragment thereof may be recombinant, synthetic, or produced by other known suitable methods. Mutations that may reduce Fc-mediated effector function of an antibody or antigen-binding fragment are, for example, those described in Leabman et al., Mabs, 2013, 5(6):896-903.
[0153] ADC
[0154] The present invention also provides an antibody-drug conjugate based on a masked antibody or antigen-binding fragment according to the present invention. The ADC according to the present invention comprises a masked antibody or antigen-binding fragment according to the present invention to which one or more linker-drug compounds have been conjugated. Methods for producing ADCs are known in the art. The masked antibody or antigen-binding fragment according to the present invention can be used in producing ADCs according to conventional methods.
[0155] To synthesize a conjugate according to the invention, one or more linker-drug compounds may be conjugated to a masked antibody or antigen-binding fragment according to the invention. A "drug" in an ADC is a pharmacologically active moiety; for example, a cytotoxic moiety or an immunomodulatory moiety or any moiety that exerts a desired pharmacological effect, directly or indirectly, on or via a cell to which the antibody or antigen-binding fragment in the ADC may bind.
[0156] Suitable linkers-drugs for ADC are known in the art. For example, WO2011 / 133039 and Elgersma et al., 2015, Molecular Pharmaceutics, 12, 1813-1835 disclose ADCs based on linkers-drugs based on duocarmycin (e.g., SYD985 based on trastuzumab). Elgersma et al. disclose the synthesis and use of linkers-drugs based on seco-DUBA (linker-drug 2 or LD2 (SYD980)).
[0157] The linker-drug compound can be conjugated through reactive natural amino acid residues present in the appropriate polypeptide, such as lysine or cysteine, or through the N-terminus or C-terminus. Alternatively, natural or non-natural reactive amino acid residues can be genetically engineered into the appropriate polypeptide, or reactive groups can be introduced by post-translational modification.
[0158] ADC according to the present invention can be produced by conjugating a linker-drug compound to a masked antibody or antigen-binding fragment thereof, for example, through the lysine ε-amino group of the antibody or antigen-binding fragment, preferably using an intermediate (e.g., activated ester) containing an amine reactive group to conjugate. Such methods are known to be used to produce conventional ADCs. Alternatively, methods and conditions known in the art can be used, see, for example, Doronina et al, 2006, Bioconjugate Chem. 17, 114-124, to conjugate a linker to produce an ADC by conjugating a free thiol of a cysteine side chain produced by reduction of an interchain disulfide bond. The preparation method involves partially reducing the interchain disulfide exposed to the solvent, followed by modifying the resulting thiol with a linker containing a Michael acceptor, such as a linker containing maleimide, α-haloacetamide or ester. The cysteine connection strategy results in a maximum of two linkers-drugs containing a linker for each reduced disulfide.
[0159] Preferred antibodies according to the present invention are human IgG types. Most human IgG molecules have four disulfide bonds exposed to the solvent, which is equivalent to an integer range of zero to eight connected linking moieties per antibody. The exact number of linker-drug molecules per antibody (or antigen-binding fragment) is determined by the degree of disulfide reduction and the number of molar equivalents of linkers-drugs containing linkers in the subsequent conjugation reaction. Complete reduction of all four disulfide bonds produces a homogeneous construct with eight linker moieties per antibody, while partial reduction generally results in a heterogeneous mixture of zero, two, four, six or eight linking moieties per antibody.
[0160] In a preferred embodiment, the present invention relates to an ADC wherein the linker-drug compound is conjugated to the masked antibody or antigen-binding fragment thereof via a cysteine residue of the antibody or antigen-binding fragment.
[0161] Because antibodies contain many lysine residues and cysteine disulfide bonds, conventional conjugation often produces heterogeneous mixtures, which pose challenges for analytical characterization and preparation. In addition, the individual components of these mixtures exhibit different physicochemical properties and pharmacology for their pharmacokinetic, efficacy and safety profiles, hindering a rational approach to optimize this model.
[0162] In order to improve the homogeneity of the conjugate, the masked antibody or antigen binding fragment used to produce the masked ADC according to the present invention can be modified to allow site-specific conjugation of the linker-drug molecule. CR Behrens and B. Liu, mAbs, 2014, 6 (1), 1-8, WO 2005 / 084390 and WO 2006 / 034488 comprehensively review the methods of conjugating site-specific drugs to antibodies (or antigen binding fragments). Site-specific ADCs are preferably produced by conjugating the linker-drug compound to the antibody or its antigen binding fragment through the side chain of the modified cysteine residue at the appropriate position of the mutant antibody or antigen binding fragment. The modified cysteine is usually blocked by other thiols such as cysteine or glutathione to form a disulfide. These blocked residues need to be uncapped before the linker-drug connection can occur. Attachment of the linker-drug to the engineered residues was achieved by: (1) reduction of both the native interchain disulfides and the mutant disulfides followed by oxidative degradation using a mild oxidizing agent (e.g., CuSO 4 or dehydroascorbic acid) followed by reoxidation of the native interchain cysteine and standard conjugation of the uncapped engineered cysteine to the linker-drug; or (2) by using a mild reducing agent that reduces the mutant disulfide at a higher rate than the interchain disulfide bonds and standard conjugation of the uncapped engineered cysteine to the linker-drug. Suitable methods for site-specific conjugation of linkers to drugs can be found, for example, in WO2015 / 177360, which describes reduction and reoxidation processes, WO2017 / 137628, which describes methods using mild reducing agents, and WO2018 / 215427, which describes methods for conjugating reduced interchain cysteines to unblocked engineered cysteines, and in Coumans et al., 2020, Bioconjugate Chem 31(9), 2136-2146, https: / / doi.org / 10.1021 / acs.bioconjchem.0c00337.
[0163] Pharmaceutical composition
[0164] In another aspect, the present invention provides a composition comprising a masked antibody, masked antigen binding fragment or masked ADC according to the present invention, preferably wherein the composition is a pharmaceutical composition, more preferably further comprising one or more pharmaceutically acceptable excipients. Such a composition is hereinafter referred to as a composition according to the present invention. The composition may be, for example, a liquid preparation, a lyophilized preparation or in the form of, for example, a capsule or tablet.
[0165] Typically, the pharmaceutical composition according to the present invention takes the form of a lyophilized cake (lyophilized powder), which needs to be (aqueous) dissolved (i.e., reconstituted) before intravenous infusion; or takes the form of a frozen (aqueous) solution, which needs to be thawed before use. Therefore, in some preferred embodiments, the present invention provides a lyophilized composition comprising a masked antibody, a masked antigen-binding fragment or a masked ADC according to the present invention, preferably wherein the composition is a pharmaceutical composition, more preferably further comprising one or more pharmaceutically acceptable excipients. In other preferred embodiments, the present invention provides a frozen composition comprising water and a masked antibody, a masked antigen-binding fragment or a masked ADC according to the present invention, preferably wherein the composition is a pharmaceutical composition, more preferably further comprising one or more pharmaceutically acceptable excipients. In this context, the frozen solution is preferably at atmospheric pressure, and the frozen solution is preferably obtained by freezing the liquid composition according to the present invention at a temperature below 0°C. Suitable pharmaceutically acceptable excipients for inclusion in pharmaceutical compositions according to the invention (before freeze drying) include buffer solutions (e.g., citrates, amino acids such as histidine, or salts comprising succinate in water), lyoprotectants (e.g., sucrose, trehalose), tonicity modifiers (e.g., chloride salts, such as sodium chloride), surfactants (e.g., polysorbates), and fillers (e.g., mannitol, glycine). Excipients for freeze-dried protein formulations are selected for their ability to prevent protein denaturation during the freeze-drying process and during storage.
[0166] Medical Uses
[0167] In another aspect, the present invention provides a masked antibody or antigen-binding fragment thereof or a masked ADC according to the present invention, or a composition according to the present invention, for use as a medicament, preferably for the treatment of cancer, an autoimmune disease or an infectious disease.
[0168] The masked antibodies, masked antigen binding fragments, masked ADCs or compositions according to the present invention are collectively referred to below as products for use according to the present invention. In one embodiment, the products for use according to the present invention are used to treat solid tumors or hematological malignancies. In a second embodiment, the products for use according to the present invention are used to treat autoimmune diseases. In a third embodiment, the products for use according to the present invention are used to treat infectious diseases, such as bacterial, viral, parasitic or other infections.
[0169] In the context of the present invention, cancer is preferably a tumor expressing an antigen directed to the product used according to the present invention. Such a tumor can be a solid tumor or a hematological malignancy. Some examples of tumors or hematological malignancies that can be treated with the product used according to the present invention as defined above may include, but are not limited to: breast cancer, brain cancer (e.g., glioblastoma), head and neck cancer, thyroid cancer, parotid cancer, adrenal cancer (e.g., neuroblastoma, paraganglioma or pheochromocytoma), bone cancer (e.g., osteosarcoma), soft tissue sarcoma (soft tissue sarcoma, STS), eye cancer (e.g., uveal melanoma), esophageal cancer, gastric cancer, small intestine cancer, colorectal cancer, urothelial cell cancer (e.g., bladder, penis, ureter or kidney cancer), ovarian cancer, uterine cancer, vaginal cancer, vulvar cancer and cervical cancer, lung cancer (especially non-small cell lung cancer (non-smallcell lung cancer, NSCLC) and small cell lung cancer (small-cell lung cancer, NSCLC)). cancer (SCLC), melanoma, mesothelioma (especially malignant pleural and abdominal mesothelioma), liver cancer (e.g., hepatocellular carcinoma), pancreatic cancer, skin cancer (e.g., basal cell carcinoma, squamous cell carcinoma, or dermatofibrosarcoma protuberans), testicular cancer, prostate cancer, acute myeloid leukemia (AML), chronic myeloid leukemia (CML), chronic lymphatic leukemia (CLL), acute lymphoblastic leukemia (ALL), myelodysplastic syndrome (MDS), blastic plasmacytoid dendritic cell neoplasia (BPDCN), Hodgkin's lymphoma, non-Hodgkin's lymphoma lymphoma (NHL) (including follicular lymphoma (FL), CNS lymphoma and diffuse large B-cell lymphoma (DLBCL)), light chain amyloidosis, plasma cell leukemia, and multiple myeloma (MM).
[0170] In the context of the present invention, an autoimmune disease is preferably an autoimmune disease associated with an antigen to which the product to be used according to the invention is directed. An autoimmune disease refers to a condition caused by an abnormal immune response to normal body cells and tissues. There are a wide variety of autoimmune diseases of at least 80 types. Some diseases are organ-specific and limited to affecting certain tissues, while other diseases are similar to systemic inflammatory diseases affecting many tissues throughout the body. The appearance and severity of these signs and symptoms depends on the site and type of inflammatory response occurring and may fluctuate over time. Some examples of autoimmune diseases that may be treated with the products used according to the invention as defined above may include, but are not limited to: rheumatoid arthritis; juvenile dermatomyositis; psoriasis; psoriatic arthritis; lupus; sarcoidosis; Crohn's disease; eczema; nephritis; uveitis; polymyositis; neuritis, including Guillain-Barre syndrome; encephalitis; arachnoiditis; systemic sclerosis; autoimmune mediated musculoskeletal and connective tissue diseases; neuromuscular degenerative diseases, including Alzheimer's disease, multiple sclerosis (MS), amyotrophic lateral sclerosis (ALS), neuromyelitis optica and large, middle size, small vessel Kawasaki disease and Henoch-Schonlein vasculitis; cold-warm agglutinin disease; autoimmune hemolytic anemia; type 1 diabetes mellitus; Hashimoto's thyroiditis; Graves' disease; Graves' ophthalmopathy; adrenalitis; hypophysitis; pemphigus vulgaris; Addison's disease; ankylosing spondylitis; Behcet's syndrome; celiac disease; Goodpasture's syndrome; myasthenia gravis; sarcoidosis; scleroderma; primary sclerosing cholangitis, epidermolysis bullosa acquisita, and bullous pemphigoid.
[0171] In the context of the present invention, infectious diseases are preferably infectious diseases associated with the antigens to which the products used according to the present invention are directed. Such infectious diseases may be bacterial infections, viral infections, parasitic infections or other infections. Some examples of infectious diseases that may be treated with the products used according to the present invention as defined above may include, but are not limited to: malaria; toxoplasmosis; pneumocystis jirovecii melioidosis; shigellosis; listeria; cyclospora; mycobacterium leprae; tuberculosis; and infection prevention in immunocompromised individuals (e.g., HIV-positive individuals, individuals in immunosuppressive therapy, or individuals with congenital defects such as cystic fibrosis or benign proliferative diseases (e.g., mola hydatidosa or endometriosis)).
[0172] The products used according to the invention as described herein can be used to prepare a medicament as described herein. The products used according to the invention as described herein are preferably used in a method of treatment, wherein the product used is administered to a subject in a therapeutically effective amount, preferably to a subject in need thereof. Therefore, alternatively, or in combination with any other embodiment, in one embodiment, the present invention relates to the use of the product used according to the invention in the preparation of a medicament for the treatment of cancer, autoimmune disease or infectious disease, particularly for the treatment of cancer. For illustrative, non-limiting cancers or other diseases to be treated according to the present invention: see above.
[0173] Alternatively, or in combination with any further embodiment, in one embodiment, the invention relates to a method for treating cancer, an autoimmune disease or an infectious disease, in particular cancer, comprising administering to a subject in need of said treatment a therapeutically effective amount of a product for use according to the invention. For illustrative, non-limiting cancers or other diseases to be treated according to the invention: see above.
[0174] The products used according to the present invention are used for administration to a subject. The products used according to the present invention can be used in the above-described therapeutic methods of administering an effective amount of the composition to a subject in need thereof. The term "subject" as used herein refers to all animals classified as mammals, and includes but is not limited to primates and humans. The subject is preferably a human. The expression "therapeutically effective amount" means an amount sufficient to achieve a desired response or improve a symptom or sign. The therapeutically effective amount for a particular subject may vary according to factors such as the condition being treated, the overall health of the subject, the method, route and dosage of administration, and the severity of the side effects.
[0175] Combination
[0176] In other embodiments, the invention provides a product for use according to the invention, wherein the use is in combination with one or more additional therapeutic agents. The product for use according to the invention may be used simultaneously or sequentially with one or more additional therapeutic agents. Suitable chemotherapeutic agents include alkylating agents such as nitrogen mustards, hydroxyureas, nitrosoureas, tetrazines (e.g., temozolomide), and aziridines (e.g., mitomycin); drugs that interfere with the DNA damage response, such as PARP inhibitors, ATR and ATM inhibitors, CHK1 and CHK2 inhibitors, DNA-PK inhibitors, and WEE1 inhibitors; antimetabolites such as antifolates (e.g., pemetrexed), fluoropyrimidines (e.g., gemcitabine), deoxynucleoside analogs, and thiopurines; antimicrotubule agents such as vinca alkaloids and taxanes; topoisomerase I and II inhibitors; cytotoxic antibiotics such as anthracyclines and bleomycin; hypomethylating agents such as decitabine and azacitidine; histone deacetylase inhibitors; all-trans retinoic acid; and arsenic trioxide. Suitable radiotherapeutic agents include radioisotopes such as 131 I-metaiodobenzylguanidine (MIBG), as sodium phosphate 32 P. 223 Ra chloride, 89 Sr chloride and 153 Sm diamine tetramethylene phosphonate / ester (EDTMP). Suitable agents for use as hormone therapeutic agents include hormone synthesis inhibitors, such as aromatase inhibitors and GnRH analogs; hormone receptor antagonists, such as selective estrogen receptor modulators (e.g., tamoxifen and fulvestrant) and anti-androgens, such as bicalutamide, enzalutamide and flutamide; CYP17A1 inhibitors, such as abiraterone; and somatostatin analogs.
[0177] Targeted therapeutics are therapeutics that interfere with specific proteins involved in tumorigenesis and proliferation, and can be small molecule drugs; proteins, such as therapeutic antibodies; peptides and peptide derivatives; or protein-small molecule hybrids, such as ADCs. Some examples of targeted small molecule drugs include TLR ligands, mTor inhibitors, such as everolimus, temsirolimus, and rapamycin; kinase inhibitors, such as imatinib, dasatinib, and nilotinib; VEGF inhibitors, such as sorafenib and regorafenib; EGFR / HER2 inhibitors, such as gefitinib, lapatinib, and erlotinib; and CDK4 / 6 inhibitors, such as palbociclib, ribociclib, and abemaciclib. Some examples of peptide or peptide-derivative targeted therapeutics include proteasome inhibitors, such as bortezomib and carfilzomib.
[0178] Suitable anti-inflammatory drugs include D-penicillamine, azathioprine and 6-mercaptopurine, cyclosporine, anti-TNF biologics (e.g., infliximab, etanercept, adalimumab, golimumab, certolizumab or certolizumab pegol), leflunomide, abatacept, tocilizumab, anakinra, nakinra, ustekinumab, rituximab, daratumumab, ofatumumab, obinutuzumab, secukinumab, apremilast, acitretin, and JAK inhibitors (for example, tofacitinib, baricitinib, or upadacitinib).
[0179] Immunotherapeutic agents include agents that induce, enhance, or suppress immune responses, such as cytokines (IL-2 and IFN-α); immunomodulatory imide drugs, such as thalidomide, lenalidomide, pomalidomide, or imiquimod; therapeutic cancer vaccines, such as talimogenelaherparepvec; cell-based immunotherapeutics, such as dendritic cell vaccines, adoptive T cells, or chimeric antigen receptor-modified T cells; and therapeutic antibodies that can trigger antibody-dependent cell-mediated cytotoxicity (ADCC), antibody-dependent cellular phagocytosis (ADCP), or complement-dependent cytotoxicity (CDC) through their Fc regions when bound to membrane-bound ligands on cells.
[0180] In the context of the present invention, treatment is preferably prevention, reversal, cure, amelioration and / or delay of cancer, autoimmune disease or infectious disease. This may mean that the severity of at least one symptom of the cancer, autoimmune disease or infectious disease has been reduced, and / or at least a parameter associated with the cancer, autoimmune disease or infectious disease has been improved.
[0181] In the context of the present invention, the subject may be alive and / or may be considered free of disease. Alternatively, the disease or condition may have been stopped or delayed. In the context of the present invention, the improvement in quality of life and the observed pain relief may mean that the subject may need less pain relief medication than at the beginning of treatment. In this context, "less" may mean 5% less, 10% less, 20% less, 30% less, 40% less, 50% less, 60% less, 70% less, 80% less, 90% less. The subject may no longer need any pain relief medication. And compared to the quality of life and observed pain relief at the beginning of treatment of the subject, this improvement in quality of life and observed pain relief can be observed, detected or assessed in the subject after treatment for at least one week, two weeks, three weeks, four weeks, one month, two months, three months, four months, five months, six months or more.
[0182] Nucleic acid construct
[0183] In another aspect, the present invention relates to a nucleic acid construct comprising:
[0184] A nucleotide sequence encoding a heavy chain variable domain, a protease cleavable peptide linker and a masking moiety; and / or a nucleotide sequence encoding a light chain variable domain, a cleavable peptide linker and a masking moiety;
[0185] Wherein said nucleotide sequence is operably linked to an expression control sequence for expression in a host cell, preferably in a mammalian host cell. Both the heavy chain and the light chain can be expressed by the same expression vector (different expression cassettes in), or by different expression vectors.
[0186] Preferably, the expression control sequence comprises a promoter and optionally other regulatory elements such as, for example, a terminator, an enhancer, a polyadenylation signal, a secretion signal sequence and the like. Such nucleic acid constructs are particularly useful for producing an activatable antibody or antigen-binding fragment thereof of the invention using recombinant techniques, wherein the nucleotide sequence encoding the activatable antibody is expressed in a suitable host cell, such as described in Ausubel et al., "Current Protocols in Molecular Biology", Greene Publishing and Wiley-Interscience, New York (1987) and in Sambrook and Russell (2001) "Molecular Cloning: A Laboratory Manual (3rd Edition), Cold Spring Harbor Laboratory, Cold Spring Harbor Laboratory Press, New York). As used herein, the term "operably linked" refers to the connection of polynucleotide elements in a functional relationship. A nucleic acid is "operably linked" when it is placed in a functional relationship with another nucleic acid sequence. For example, if a promoter or enhancer affects the transcription of a coding sequence, the promoter or enhancer is operably linked to a coding sequence. Operably linked means that the DNA sequences being linked are generally contiguous and, where necessary to connect two protein coding regions, contiguous and in reading frame.
[0187] In a preferred embodiment, the nucleotide sequence defined above is codon-optimized for the host cell in which the activatable antibody or antigen-binding fragment thereof according to the present invention is produced.
[0188] In a preferred embodiment, the nucleic acid construct is a vector that is compatible with the host cell.
[0189] Host cells
[0190] In another aspect, the present invention relates to a host cell comprising a nucleic acid construct as defined above. Preferably, the host cell is a host cell for producing an activatable antibody or antigen-binding fragment thereof of the present invention. The host cell may be any host cell capable of producing an antigen-binding protein of the present invention, including, for example, a prokaryotic host cell, such as, for example, Escherichia coli (E. coli); or (cultured) mammalian, plant, insect, fungal or yeast host cells, including, for example, CHO cells, BHK cells, human cell lines (including HeLa, COS, HEK293 and PER.C6), Sf9 cells and Sf+ cells. However, the preferred host cell for producing an activatable antibody or antigen-binding fragment thereof of the present invention is a mammalian cell, more preferably a HEK293 or CHO cell.
[0191] preparation
[0192] The masked antibody or antigen-binding fragment according to the present invention can be prepared by recombinant DNA techniques known in the art for producing recombinant antibodies or antigen-binding fragments. In essence, such a method involves the production of a DNA construct (e.g., an expression vector) comprising the following: a DNA coding sequence of an antibody or antigen-binding fragment, and a coding sequence of a masking portion and optionally a peptide linker (including optionally a spacer sequence), all of which are in a suitable order and orientation and have conventional expression control elements known in the art. Such a construct can be introduced into a suitable host (e.g., bacteria, yeast or mammalian cells) for the production of masked antibodies or antigen-binding fragments in cell culture in a suitable culture medium. Most monoclonal antibodies or antigen-binding fragments produced for use as biotherapeutics are produced in mammalian cell cultures such as Chinese hamster ovary (CHO) cells that stably express the antibody or antigen-binding fragment. The antibody or antigen-binding fragment can be harvested and purified from the cell culture. Therefore, another aspect of the present invention relates to a method for producing an activatable antibody or antigen-binding fragment thereof according to the present invention, the method comprising the step of culturing a host cell as defined above under conditions that are conducive to the expression of the activatable antibody or antigen-binding fragment thereof. The method preferably comprises the following steps: (a) culturing a host cell as defined above under conditions that are favorable for the expression of an activatable antibody or antigen-binding fragment thereof; and optionally (b) purifying the activatable antibody or antigen-binding fragment thereof from at least one of the host cell and the culture medium. Suitable conditions may include the use of a suitable culture medium, the presence of a suitable food and / or a source of suitable nutrients, a suitable temperature, and optionally the presence of a suitable inducing factor or compound (e.g., when the nucleotide sequence of the present invention is under the control of an inducible promoter); all suitable conditions can be selected by a skilled person. Under such conditions, the amino acid sequence of the present invention may be expressed in a constitutive manner, transiently, or only upon appropriate induction. Then, the activatable antibody or antigen-binding fragment thereof of the present invention may be isolated from the host cell / host organism and / or from the culture medium in which the host cell or host organism is cultured using protein separation and / or purification techniques known per se, such as (preparative) chromatography and / or electrophoresis techniques, differential precipitation techniques, affinity techniques (e.g., using a specific cleavable amino acid sequence fused to the amino acid sequence of the present invention) and / or preparative immunological techniques (i.e., using antibodies against the antigen-binding protein to be isolated).
[0193] use
[0194] In another aspect, the present invention relates to the use of an antigenic peptide and an sdAb that specifically binds to the antigenic peptide as part of a mask in an activatable antibody or antigen-binding fragment thereof according to the present invention.
[0195] In this document and its claims, the verb "comprise" and its conjugations are used in their non-restrictive sense, meaning that the items following the word are included, but items not specifically mentioned are not excluded. Furthermore, unless the context clearly requires the presence of one and only one element, reference to an element by an indefinite article does not exclude the possibility that more than one element is present. Thus, a noun not qualified by a quantifier generally means "at least one".
[0196] When used with a numerical value (eg, about 10), the word "about" or "approximately" preferably means that the value may be 1% of the given value, greater or less than that value.
[0197] Whenever a parameter of a substance is discussed in the context of the present invention, it is assumed that the parameter is determined, measured or displayed under physiological conditions unless otherwise stated. Physiological conditions are known to those skilled in the art and include an aqueous solvent system, atmospheric pressure, a pH value of 6 to 8, a temperature of room temperature (RT) to about 37°C (about 20°C to about 40°C), and appropriate concentrations of buffer salts or other components.
[0198] All patent and literature references cited in this specification are hereby incorporated by reference in their entirety.
[0199] The following examples are provided for illustrative purposes only and are not intended to limit the scope of the present invention in any way.
[0200] Example
[0201] Example 1: Transient expression, purification and HMW (%) determination of (masked) antibodies (Materials and methods)
[0202] Design of masked antibody chains
[0203] The masked antibody chains are assembled by linking the HAVT20 leader sequence (Boel et al., 2000, J. of Immunol. Meth., 239, 153-166, (SEQ ID NO: 43) to the N-terminus of a masking moiety comprising a BC2 nanobody (or a variant thereof) or a BC2T peptide (or a variant thereof), followed by a linker comprising a matrix metalloproteinase 2 (MMP2) linker sequence L9 (SEQ ID NO: 11) or L13 (SEQ ID NO: 12) (a scrambled form of L9 that cannot be cleaved by MMP2) substrate sequence, optionally flanked by spacer sequences (Qg: SEQ ID NO: 9 or Qv: SEQ ID NO: 10), and finally followed by the anti-HER2 antibody trastuzumab (HC: SEQ ID NO: 21, HC-P41 C SEQ ID NO: 22, LC: SEQ ID NO: 23), the anti-EpCAM antibody ING-1 (HC: SEQ ID NO: 24, HC-P41 C SEQ ID NO: 25), the anti-EpCAM antibody ING-2 (HC: SEQ ID NO: 26, HC-P41 C SEQ ID NO: 27), the anti-EpCAM antibody ING-3 (HC: SEQ ID NO: 28, HC-P41 C SEQ ID NO: 29), the anti-EpCAM antibody ING-4 (HC: SEQ ID NO: 30, HC-P41 C SEQ ID NO: 31), the anti-EpCAM antibody ING-5 (HC: SEQ ID NO: 32, HC-P41 C SEQ ID NO: 33), the anti-EpCAM antibody ING-6 (HC: SEQ ID NO: 34, HC-P41 C SEQ ID NO: 35), the anti-EpCAM antibody ING-7 (HC: SEQ ID NO: SEQ ID NO: 25, LC: SEQ ID NO: 26) or adelimumab (HC: SEQ ID NO: 27, HC-P41C SEQ ID NO: 28, LC: SEQ ID NO: 29), anti-TROP2 antibody gosatuzumab (HC: SEQ ID NO: 30, HC-P41 C SEQ ID NO: 31, LC: SEQ ID NO: 32), anti-TF antibody MORAb-066 (HC: SEQ ID NO: 33, HC-P41C SEQ ID NO: 34, LC: SEQ ID NO: 35) or tesorumab (HC: SEQ ID NO: 36, HC-P41C SEQ ID NO: 37, LC: SEQ ID NO: 38), anti-CD137 antibody urerulumab (HC: SEQ ID NO: 39, LC: SEQ ID NO: 40), or anti-CTLA4 antibody tremelimumab (HC: SEQ ID NO: 41, LC: SEQ ID NO: 42). NO: 42).
[0204] The BC2-Nb sequence is shown in SEQ ID NO: 4, while its CDR3 sequence is described in SEQ ID NO: 5 and variant CDR3 sequences are described in SEQ ID NOs: 6 to 8. The BC2T peptide sequence is described in SEQ ID NO: 1 and the sequences of the BC2T peptide variants are described in SEQ ID NOs: 2 to 3. Optionally, the masking portion comprising the BC2T peptide comprises a 7 amino acid residue spacer sequence QGQSGQG (Qg, SEQ ID NO: 9) or QVQLVES (Qv, SEQ ID NO: 10) in front of the BC2T sequence (at the N-terminus), enabling efficient cleavage of the signal peptide and enhancing expression of the antibody chain. Cleavage of the leader sequence corresponds to the use of the SignalP program (https: / / services.healthtech.dtu.dk / service.php?SignalP,Bendtsen,Jannick et al. "Improved prediction of signal peptides: SignalP 3.0." Journal of molecular biology 340.4 (2004): 783-795.) predicted cleavage site.
[0205] In order to enable site-specific conjugation to be used for the development of antibody-drug conjugates, by replacing the residue at the 41st position according to Kabat numbering with a cysteine residue, a single cysteine residue is introduced into the heavy chain variable domain. The resulting masked heavy chain and light chain amino acid sequences are back-translated into cDNA sequences and codon optimized for expression in human cells (Homo sapiens). Heavy chain and light chain cDNA constructs are chemically synthesized by commercial sources (Geneart, Thermo Fisher Scientific) and obtained from commercial sources (Geneart, Thermo Fisher Scientific).
[0206] Expression vector construction
[0207] For the expression of antibody chains, a mammalian expression vector pcDNA3.4-TOPO (Invitrogen, Thermo Fisher Scientific) containing a CMV:BGHpA expression cassette was used. Heavy chain (HC) or light chain (LC) genes were synthesized by synthetic oligonucleotide (oligo) assemblies and / or PCR products and cloned into pcDNA3.4-TOPO, respectively. After transformation into E. coli K12 DH10B T1R and amplification, large-scale production of antibody chain expression vectors for transfection was performed using the EndoFree plasmid Maxi kit according to the manufacturer's instructions (Qiagen).
[0208] Transient expression in mammalian cells
[0209] Commercially available Expi293F cells (Gibco, Thermo Fisher Scientific) were co-transfected with the antibody chain expression vectors prepared as described above using FectoPRO transfection reagent (Polyplus-transfection) according to the manufacturer's instructions as follows: one day before transfection, cells were plated at 1×10 6 Viable cells (viable cell, VC) / mL were inoculated in 270mL Expi293 expression medium; One day after inoculation, 240μg antibody chain expression vector (1: 1 ratio) was combined with 240μL FectoPRO transfection agent and added to cells. Cells were cultured according to the manufacturer's instructions (Gibco, Thermo Fisher Scientific). As described above, small-scale batches were obtained by linear scalability and thus using equal volume ratios. 6 days after transfection, cell culture supernatants were harvested by centrifugation at 4,000g for 15 minutes, and the clarified harvest was filtered through a MF75 filter (Nalgene). For smaller batches (3mL), the filtration step was omitted. According to the manufacturer's instructions (Sartorius), the antibody concentration was determined by ForteBio Octet QK384 using protein A biosensor and trastuzumab as a calibrant.
[0210] Purification of (masked) antibodies, and determination of titer and HMW
[0211] Use the following operation to purify (masking) antibody. The harvest obtained by expression comprising the (masking) antibody produced in Expi293 cells is captured on protein A resin (column or plate), and is adjusted to pH 5.5. to 6.0 or rebuffered to pH 6.0 buffer after elution. Step is carried out as follows: protein A step (MabSelect SuRe, Cytiva) starts with 0.1M NaOH to resin (column or plate) cleaning (sanitization) 15 minutes, then rinsed with pure water (PurifiedWater, PW). Then balance resin with PBS of pH 7.4. The harvest comprising (masking) antibody is loaded on resin with the ratio of ≤25mg (masking) antibody / mL MabSelect SuRe. Then resin is washed with balancing buffer (PBS pH7.4), then washed with 25mM NaAc pH 5.0. Then 25mM acetate elution (masking) antibody with pH 3.0. The concentration of the purified (masked) antibody is measured using the Lambert-Beer law using a theoretical extinction coefficient based on the amino acid sequence of the (masked) antibody.After elution, the (masked) antibody is adjusted to pH 5.5. to 6.0 or rebuffered to a pH 6.0 buffer.
[0212] For purity of (masked) antibodies and soluble protein aggregation analysis, Size Exclusion Chromatography (SEC) with UV detection was used. 400 mM NaCl, 150 mM PO 4 The (masked) antibody was diluted to a concentration of 1 mg / mL using buffer pH 6.2 with 15% 2-propanol as the mobile phase and injected onto a TSKgel UP-SW3000 column with a TSKgel UPSW direct-connect guard column inline. The percentage of the main peak (monomer) and high molecular weight (HMW) species (e.g., soluble aggregates) was determined based on the peak area relative to the total peak area of all protein-related peaks.
[0213] For antibodies purified using a column, the integrity of the (masked) antibody was checked by SDS-PAGE.
[0214] Example 2: Generation of multiple (semi-)masked antibodies
[0215] 2A: Masked trastuzumab antibody variants (Group 1: masked antibodies S101 to S112)
[0216] To investigate whether different variants of masked trastuzumab antibodies could be produced with acceptable titers, variants were generated that differed in how the masking moiety was attached to the antibody. The masked antibodies were based on the anti-HER2 antibody trastuzumab (Ab) or trastuzumab with a modified HC sequence in which the proline amino acid at position 41 was replaced by cysteine (41C). All antibodies carried a masking moiety comprising a BCT2 peptide (SEQ ID NO: 1) or BC2-Nb (SEQ ID NO: 4).
[0217] The cleavable sequence comprises the MMP2 cleavable sequence PLGLAG (SEQ ID NO: 13) flanked by GS-rich regions at both ends, resulting in the following sequence: GGGGSPLGLAGGGS (referred to herein as "L9" and shown in SEQ ID NO: 11).
[0218] The antibodies differ in whether the masking moiety with BC2-Nb is attached to the HC and the masking moiety with the BCT2 peptide is attached to the LC, or vice versa.
[0219] In Table 1, the production titers of various masked antibodies are shown. The "HC" column in Table 1 shows various masking moieties at the N-terminal site of the peptide (Nanobody BC2-Nb or peptide BCT2, preceded by a spacer sequence (Qg or Qv)) and a linker (L9) on the heavy chain of the prepared masked antibodies, while the "LC" column shows various masking moieties and linkers on the light chain, which are combined with the heavy chain to form a masked antibody. Antibodies in which both HC and LC contain the same BC2 nanobody masking moiety can also be produced with high titers.
[0220] It has been found that, in general, the production of antibodies with masking moieties comprising nanobodies on the LC (antibodies S103, S104, S107, S108 and S110) leads to higher titers and forms a lower percentage of HMW complexes. In Table 1, the percentage of HMW complexes in the products of masked antibodies with a variety of masking moiety combinations is shown. For comparison, some "half-masked" antibodies were also produced, in which only the HC or LC comprised a masking moiety. Antibodies in which both the HC and LC comprised the same masking moiety were also included. Obviously, half-masked antibodies and antibodies with double Nb masking do form less HMW complexes.
[0221]
[0222] Figure 1A A schematic diagram of an antibody is provided in which the Nb is on the HC and the peptide is on the LC, and vice versa, while Figure 1B A schematic diagram of the formation of HMW complexes between masked antibodies is shown in FIG.
[0223] 2B: Comparison of yield (titer (μg / mL)) and HMW (%) after production of mutant and wild-type BC2T / BC2-Nb masked trastuzumab
[0224] It was hypothesized that HMW formation may be related to the strength (affinity) of the bond between the masking moieties on the LC and HC.To test this hypothesis, a panel of antibodies (panel 2) was generated using masking moieties based on the BC2-Nb and BC2T peptides.
[0225] The antibody used was trastuzumab and the masking moiety was attached to the N-terminus of the HC and LC pair using the L9 linker sequence also described in Example 1 for attachment of the (mutated) BC2T peptide and the (mutated) BC2-Nb.
[0226] Under the concept that the affinity of the bond between the nanobody and the peptide will be reduced and the aggregation leading to high molecular weight (HMW) complexes can be reduced, mutations were introduced in the amino acid sequence of BC2-Nb and / or BC2 peptide to generate BC2 / BC2T variants. The mutations tested are shown in Table 2. The mutations in Nb are in the CDR3 of Nb. The CDR3 of BC2-Nb is shown in SEQ ID NO: 5. As shown in Table 2, the mutant CDR3 sequences of the BC2-Nb variants are shown in SEQ ID NOs 6 to 8.
[0227] Table 2: Mutated BC2T peptides and BC2-Nb variants used in the experiments
[0228] BC2T peptide variants BC2-Nb variant BC2T-S8A (SA, SEQ ID NO: 2) <![CDATA[BC2 N b-R106A(RA,SEQ ID NO:6)]]> BC2T-S8T (ST, SEQ ID NO: 3) <![CDATA[BC2 N b-R106E(RE,SEQ ID NO:7)]]> <![CDATA[BC2 N b-R1 06S(RS,SEQ ID NO:8)]]>
[0229] Each masking moiety with BC2T peptide or variants thereof was combined with each corresponding masking moiety with BC2-Nb or variants thereof, resulting in 3×4=12 combinations. All combinations with Nb or peptide masking moieties on HC or LC were tested, resulting in a total of 12×2=24 different masked antibodies.
[0230] The results (titer (μg / mL) and HMW (%) after generation) for all 24 masked antibodies with combinations of masking moieties tested are shown in Table 3.
[0231]
[0232] It is evident from Table 3 that the production of antibodies in which the masking moiety with Nb is on the LC (S201 to S211) resulted in higher production titers compared to the production of antibodies in which Nb is on the HC (S212 to S222).
[0233] It is also apparent that mutations in the Nanobody sequence and / or peptide sequence lead to a lower percentage of HMW complex formation (probably because the binding affinity between the masking moieties is reduced due to the mutation). See the example results for antibodies S209, S205, S201, S203, S211 and S207. This effect can be seen for antibodies where the Nb is on the LC as well as for antibodies where the Nb is on the HC (especially for the BC2-Nb R106A and R106S mutants, such as S214 and S220).
[0234] 2C: Generation of masked antibodies against multiple targets (Group 3)
[0235] To demonstrate that masked antibodies according to the invention can be prepared based on antibodies against a wide variety of targets, different masked antibodies (Group 3) were generated. Group 3 includes mAbs targeting EpCAM, TROP2, Tissue Factor (TF), CD137 and CTLA4. This group includes unmasked (naked) antibodies as well as masked antibodies and masked forms of antibodies carrying a 41C mutation in their HC (prepared for subsequent use in site-directed linkage of linker-drug molecules to produce masked antibody-drug conjugates).
[0236] All masked antibodies in Group 3 carry the following masking moieties: BC2-Nb is linked to the N-terminus of the LC of each antibody with a linker sequence L9, whereas the masking moiety on the N-terminus of the HC of each antibody carries a BC2T peptide preceded by a Qg sequence and linked to the HC via the L9 sequence.
[0237] Results of antibody production
[0238] 30 mL batches were produced for all masked antibodies.
[0239] HMW (%) was relatively low for all wt mAbs (0.6 to 2.7) and for both masked wt and 41C tesomab (4.1 and 3.7, respectively).
[0240] Masked ING-1 mAb targeting EpCAM showed the highest HMW(%) (65.5). Masked wt and 41C-adelimumab (also targeting EpCAM) also showed relatively high HMW(%) (33.7 and 30.9, respectively), as shown in Table 4.
[0241] 2D: Production yield (mg) and HMW (%) of semi-masked antibodies against different targets (EpCAM, TROP2, TF, CD137 and CTLA4)
[0242] As demonstrated by the data shown in Table 5, semi-masked antibodies can also be generated.
[0243] Antibodies against various targets are semi-masked, which means that the masking part comprising BC2-Nb is linked to the LC of the antibody using the L9 linker sequence, or the masking part comprising the BC2T antigen peptide is linked to the HC of the antibody. When linked to the HC, the Qg spacer sequence is included in front of the peptide sequence.
[0244] For several antibodies, two variants of the masked HC were generated: one with the wild-type sequence of the antibody HC, and one with a 41C HC mutation (the 41C mutation introduces a cysteine that can be used for site-specific bonding of the linker-drug molecule when the antibody is used in an ADC).
[0245]
[0246]
[0247] Example 3: Synthesis of antibody-drug conjugates based on (masked or semi-masked) antibodies.
[0248] The masked antibody-based ADC in which the duocarmycin-based linker-drug LD2 (Elgersma et al., 2015, supra) was conjugated in a site-specific manner to the HC41 cysteine introduced in the antibody HC sequence was synthesized according to the following scheme as disclosed in Coumans et al., 2020, supra:
[0249] The antibody solution (10 to 15 mg / mL, pH 5, 100 mM histidine) modified by cysteine was treated with 2-(diphenylphosphino)benzenesulfonic acid (diPPBS, 16 to 32 equivalents, 10 mM in water), and the resulting mixture was incubated for 16 to 24 hours at RT. Excess diPPBS was removed by centrifugal concentrator (Vivaspin filter, 30 kDa cutoff, PES) using 4.2 mM histidine, 50 mM trehalose, pH 6 or by carbon filtration. The pH of the resulting antibody solution was raised to about 7.4 using Tris (1 M in water, pH 8), or maintained at pH 6, after which N, N-dimethylacetamide (DMA) was added, followed by a linker-drug solution (10 mM in DMA). The final concentration of DMA was 5% to 10%. The resulting mixture was incubated in dark at RT for 2 to 3 hours, or incubated overnight in the case of pH 6. To remove excess linker-drug, activated charcoal was added and the mixture was incubated at RT for at least 0.5 h. The charcoal was removed using a 0.2 μm PES filter and the resulting ADC was formulated in 4.2 mM histidine, 50 mM trehalose, pH 6 using a Vivaspin centrifugal concentrator (30 kDa cutoff, PES).
[0250] By using the above-mentioned site-directed conjugation method, the ADCs shown in Table 6 were prepared, all of which were prepared by site-directed conjugation, carrying linker-drug 2 (LD2) with a drug to antibody ratio (DAR) of 2.
[0251]
[0252] Example 4: Biolayer Interferometry (BLI) Experiments to Compare Target Binding of Masked and Naked (Unmasked) Antibodies
[0253] Materials and methods
[0254] Enzymatic demasking of masked antibodies / ADCs by protease treatment
[0255] A 24 mM APMA (p-aminophenylmercuric acetate, Sigma-Aldrich, A-9563) stock solution in DMSO was prepared. rhMMP2 (R&D systems, 902-MP, 0.1 mg / mL) was thawed to RT. 10 μL of enzyme was activated with 0.44 μL of fresh APMA solution at 37°C for 1 hour. After incubation, 329 μL of TCNB (50 mM TRIS, 10 mM CaCl 2 The activated enzyme solution was diluted to 3.0 ng / μL in 5% (w / v) BRIJ-35 (pH 7.5) buffer.
[0256] The masked mAb or ADC was diluted with TCNB buffer, after which the activated enzyme was added. Final conditions: 0.4 mg / mL substrate and a substrate to enzyme molar ratio of about 300. The mixture was incubated at 37°C with gentle shaking. After 2 hours, the mixture was cooled to RT and a sample was taken for SDS-page analysis, after which it was stored at -80°C until further use.
[0257] Biolayer Interferometry (BLI)
[0258] The target protein of biotinylation is fixed on SAX biosensor tip by biotin-streptavidin interaction, or is fixed on Ni-NTA for fixing the target protein through His labelling.By immersing the SAX / Ni-NTA sensor tip fixing in these analytes, the masking antibody is tested with 50 times of dilutions.In order to correct the concentration difference between the analyte stock solution, the protein A of the same sample is combined and evaluated.After 10 seconds of association, the binding level of masking antibody (analyte) and SAX or Ni-NTA tip through fixing target and protein A sensor tip is measured, and is used for analysis, because it is in the linear part of binding curve and therefore proportional to analyte concentration.
[0259] Material
[0260] 384-well slanted bottom microplate, Forté bio, catalog number 18-5076
[0261] High Precision Streptavidin (SAX) Dip and Read Sensor Tip, Forte Bio, 18-5118
[0262] Protein A sensor tip, Forte Bio, 18-5012
[0263] HBS-EP+ buffer, GE Healthcare, 10×: BR-1006-69, #30964
[0264] Biocytin in HBS-EP+, 1 mg / mL (Example 4B)
[0265] Ni-NTA biosensor, Forte Bio, 18-5102 (Example 4B)
[0266] HBS-P+, GE Healthcare (Example 4B)
[0267] Example 4A (Group 2 Antibodies)
[0268] • Biotinylated human HER2 (Acro Biosystems, Cat. No. HE2-H822R) as target protein was immobilized on the SAX biosensor tip.
[0269] The masking antibodies and their concentrations are shown in Table 7.
[0270] Table 7: Masking antibodies tested for binding to HER2 and respective concentrations
[0271]
[0272] Example 4B (Group 3 Antibodies)
[0273] The following reagents were immobilized on the sensor tip as target proteins:
[0274] Biotinylated antigens EpCAM, TROP2 and CD-137 (4-1BB) immobilized on the SAX biosensor tip:
[0275] Biotinylated human TROP2 / TACSTD2 protein, His Avitag, catalog number TR2-H82E5, 200 μg / mL
[0276] Biotinylated human EpCAM / TROP1 protein, Avitag His tag, catalog number EPM-H82E8, 200 μg / mL
[0277] Biotinylated human CTLA4 protein, His Avitag, catalog number CT4-H82E1, 200 μg / mL
[0278] Biotinylated human 4-1BB / TNFRSF9 protein, His Avitag, catalog number 41B-H82E6, 200 μg / mL
[0279] Human Coagulation Factor III / Tissue Factor Protein, His-Tag, Catalog No. TF3-H52H5, 600 μg / mL (His-Tagged Tissue Factor immobilized on Ni-NTA biosensor tip)
[0280] • Group 3 Nanobody-peptide-masked trastuzumab constructs tested are shown in Table 10 at the respective indicated concentrations.
[0281] Table 10: Group 3 antibodies tested and their respective concentrations
[0282]
[0283] Example 4A: Reduced binding of masked Group 2 antibodies to HER2 when compared to trastuzumab
[0284] In order to test the effect of the mutations introduced in the masking part of the masking antibodies of group 2 on their masking efficiency, the ability of antibodies S201 to S222 to bind to HER2 (the target antigen of the masking trastuzumab antibody) was tested. Their binding was compared with two masking antibodies (S101 and S103) from group 1, and wt trastuzumab (TRA) was used as a positive control (at a concentration of 21 mg / mL), and anti-PSMA SYD1030 (HC S41C) (aPSMA) was used as a negative control (at a concentration of 1 mg / mL). The experiment was performed as described in "Materials and Methods" above.
[0285] result:
[0286] Tables 8 and 9 show the binding levels of the results of the potential binding of antibodies to HER2 fixed on the SAX sensor tip and to the protein A sensor tip after 130 seconds when compared with the baseline. The results are shown as wavelength shifts, representing the optical thickness of the biolayer on the sensor tip in nm. The 130 second reporting point corresponds to an association of 10 seconds, because the baseline was generated in the first 120 seconds. The last column shows the quotient of HER2 / concentration, and this can be used to compare the binding between different nanobody-peptide trastuzumab constructs.
[0287] It was found that all masked trastuzumab constructs showed significantly reduced binding to HER2 compared to unmasked trastuzumab, indicating that modifications of the Nanobodies and / or peptides used in the masking portion of the masked antibodies in Group 2 do not negatively affect their ability to mask the binding of the masked trastuzumab to its target HER2. Together with the data on the titer and HMW (%) produced in Table 3, these data indicate that masked antibodies that can be efficiently produced (shown by high titers and relatively low HMW (%)) still have favorable masking characteristics (meaning that the binding of the masked antibodies to their targets is significantly reduced when compared to the positive control trastuzumab).
[0288] Table 8: Binding of modified masked trastuzumab antibodies S201 to S211 (Nanobodies on the light chain variable domain) compared to Group 1 antibodies S101 and S103 and unmasked trastuzumab.
[0289] N / A = Not Applicable
[0290]
[0291] Table 9: Binding of modified masked trastuzumab antibodies S212 to S222 (Nb on HC) compared to Group 1 antibodies S101 and S103 and unmasked trastuzumab. N / A = Not applicable
[0292] Example 4B: Reduced binding of masked Group 3 antibodies to their respective targets when compared to naked (unmasked) antibodies
[0293] In order to study the masking efficiency of masked antibodies against multiple targets, the ability of group 3 antibodies S300 to S317 to bind to their respective targets was tested using biolayer interferometry (BLI). As described in Example 2C, group 3 includes mAbs targeting EpCAM (ING-1 and adelimumab), TROP2 (gosatuzumab), TF (MORAb-066 and tesomonab) and CD137 (urerulumab). Group 3 includes naked (unmasked) antibodies as well as masked wt antibodies and masked antibodies carrying P41C mutations in their HC (prepared for subsequent use in site-directed linkage of linker-drug molecules to produce antibody-drug conjugates based on masked antibodies).
[0294] Experiments were performed as described above in “Materials and Methods”.
[0295] result
[0296] Tables 11A to F show the binding levels shown as wavelength shifts in nm, indicating potential binding of the antibody to the target immobilized on the SAX sensor tip or Ni-NTA sensor tip and to the Protein A sensor tip after 70 seconds. The reporting point at 70 seconds corresponds to 10 seconds of association, since the baseline is generated in the first 60 seconds. The last column shows the quotient of target binding / concentration, and it can be used to compare binding between different Nanobody-peptibody constructs.
[0297] Fig. 8A Also shown in Figures F to F are the results for binding of naked, masked, masked (P41C), and MMP2-treated forms of each antibody tested to its respective target, corrected for differences in concentration (target binding / protein A binding*1000).
[0298] It was observed that the masked antibodies all showed significantly reduced binding to the target compared to the wild-type antibody.
[0299] Masked antibodies were also treated with MMP2 to analyze whether unmasking occurs in the presence of MMP2. For all antibodies, the binding levels of wild-type antibodies were similar compared to the binding levels of masked antibodies treated with MMP2, thus proving the concept of MMP2-cleavable nanobody-masked constructs.
[0300] Table 11A: Anti-EpCAM ING-1. N / A = Not applicable
[0301]
[0302] Table 11B: Anti-EpCAM adelimumab. N / A = Not applicable
[0303]
[0304] Table 11C: Anti-TROP2 Gosartuzumab. N / A = Not applicable
[0305]
[0306] Table 11D: Anti-CD137 (4-1BB) Urecizumab. N / A = Not applicable
[0307]
[0308] Table 11E: Anti-Tissue Factor (TF) MORAb-066. N / A = Not Applicable
[0309]
[0310] Table 11F: Anti-Tissue Factor (TF) Tesomab. N / A = Not Applicable
[0311]
[0312] Example 5: Cell Binding and Cytotoxicity of Masked, Semi-masked and Unmasked Antibodies
[0313] Materials and methods
[0314] Determine cell binding
[0315] Cellular binding of masked, semi-masked, naked (unmasked) and unmasked (MMP2 treated according to the protocol described in Example 4) antibodies and ADCs was studied in the human HER2 positive tumor cell line SK-BR-3. Cells at approximately 90% confluence were used for the analysis, detached with trypsin-Versene (EDTA) (Lonza) for 5 to 10 minutes, washed and in ice-cold FACS buffer (PBS 1×, 0.1% v / w BSA, 0.02% v / v sodium azide (NaN 3 )) is adjusted to 1×10 6 The concentration of cells / mL.
[0316] Staining was performed in 96-well round-bottom microtiter plates at 4°C using ice-cold reagents / solutions to prevent modulation and internalization of surface antigens. 100,000 cells / well were added to a 96-well plate (100 μL / well) and centrifuged at 300×g for 3 minutes. The supernatant was discarded and the cells were stained with 50 μl of each antibody or ADC for 30 minutes. Serial dilutions were performed in ice-cold FACS buffer. The cells were washed three times by centrifugation at 300×g for 3 minutes and resuspended in 50 μl of a 6000-fold diluted second F(ab') 2 Goat anti-human IgG (Fc fragment specific) antibody APC conjugate (Jackson ImmunoResearch).
[0317] After incubation on ice for 30 minutes, cells were washed again and resuspended in 150 μL ice-cold FACS buffer for FACS analysis. To analyze the data, fluorescence intensity was determined by flow cytometry (BD Biosciences) and expressed as median fluorescence intensity (MFI-median).
[0318] Dose-response binding curves were fitted by nonlinear regression with variable slope (four parameters) in GraphPad Prism version 9. EC was calculated in GraphPad Prism 50 value.
[0319] Determination of in vitro cytotoxicity
[0320] The in vitro cytotoxicity of masked, unmasked, demasked and / or semi-masked anti-HER2 ADCs was determined in the human HER2-positive tumor cell line SK-BR-3 and the human HER2-negative tumor cell line Jurkat NucLightRed. Cells were plated in 96-well plates (90 μL / well) in complete growth medium and incubated at 37°C, 5% CO 2 Incubate at the following cell densities: 6500 SK-BR-3 and 3000 Jurkat NucLight Red, 90 μl per well.
[0321] After overnight incubation, 10 μL of ADC was added. ADC was serially diluted in culture medium. The luminescent cell viability assay (Promega Corporation, USA) was used to assess cell viability according to the manufacturer's instructions.
[0322] Percent survival was calculated by dividing the luminescence measured for each ADC concentration by the mean of untreated cells (growth medium only) and multiplying by 100.
[0323] IC was calculated for each dose-response curve using log(inhibitor) versus response with variable slope (four parameters) fit available in GraphPad Prism version 9. 50 value.
[0324] Example 5A: Cellular binding of masked and unmasked Group 2 (S201 to S211) antibodies to SK-BR-3 cells
[0325] Cell binding experiments were performed with antibodies from Group 2. Group 2 antibodies are shown in Table 3 and include antibodies with mutations in the BC2-Nb and / or BC2T peptide sequences. Unmasking with MMP2 and cell binding were performed as described (M&M, supra).
[0326] The aim of this experiment was to investigate whether there was restoration of target binding after demasking of the masked antibody when compared to target binding of naked (unmasked) trastuzumab. In parallel, the effect of different mutations in the de BC2-Nb and / or BCT2 peptide sequences on the masking efficiency of the set 2 antibodies (S201 to S211) was investigated and compared to S103 (S103 was masked with a masking moiety similar to the SET2 antibody but without mutations in the BC2-Nb and BC2T peptide sequences).
[0327] The ECs for cellular HER2 target binding on SK-BR-3 cells due to unmasking of masking antibodies with MMP2 when compared to binding of naked (unmasked) trastuzumab are shown in Table 12 50Since the dose response curve showed no top (saturation binding was not reached), a reliable EC for target binding of the masked Ab could not be determined. 50 Table 12 shows the binding characteristics of all unmasked Group 2 antibodies and unmasked S103, expressed as binding EC 50 , which means the antibody concentration at which 50% of total binding occurs on HER2-positive SK-BR-3 cells and relative to the EC of naked (unmasked) trastuzumab 50 Offset.
[0328] Figures 2A to 2L show the dose-dependent binding of antibodies S103 and S201 to S211 on HER2-positive SK-BR-3 cells when masked and after MMP2 treatment (unmasking) compared to naked (unmasked) trastuzumab binding as measured on a flow cytometer. The data show the MFI (median fluorescence intensity) from one experiment.
[0329] The dose response curves shown in Figures 2A to 2L and the results shown in Table 12 indicate that the antibodies were effectively unmasked following MMP2 treatment, resulting in almost complete restoration of HER2 target binding for most of the unmasked antibodies (when compared to naked (unmasked) trastuzumab).
[0330] Table 12: EC of cellular HER2 target binding on SK-BR-3 cells due to unmasking of masking antibodies with MMP2 50 Offset
[0331] Reference Number Ab <![CDATA[EC 50 ]]> <![CDATA[EC 50 Shift (unmasked Ab, when compared to wt)]]> Naked (unmasked) trastuzumab 0.49 na Demasking S103 0.64 1.3 Demasking S201 0.64 1.3 Demasking S202 2.49 5.1 Demasking S203 1.29 2.6 Demasking S204 0.34 0.7 Unmasking S205 0.77 1.6 Demasking S206 0.69 1.4 Unmasking S207 0.48 1.0 Demasking S208 0.42 0.9 Demasking S209 0.83 1.7 Demasking S210 0.58 1.2 Unmasking S211 0.55 1.1
[0332] Because the masked antibody barely bound to its target, even at very high concentrations, saturation binding was not achieved for the masked antibody (resulting in an incomplete dose-response curve, as shown in Figures 2A to 2L), and a reliable EC could not be determined. 50 value.
[0333] The differences in binding were apparent when comparing the concentrations at which minimal binding was observed for masked, naked (unmasked), and unmasked (MMP2-treated) antibodies. While minimal binding was observed for either unmasked or unmasked antibodies at concentrations as low as 0.04 μg / mL, masked antibodies showed minimal binding only at concentrations ranging from 3.7 μg / mL (S103 and S204) to as high as 33 μg / mL (S210). For most masked antibodies tested, minimal binding was only observed at concentrations as high as 11 μg / mL. If the estimated EC values were determined manually by the calculator, the results were consistent with those of the previous study. 50 , the values obtained for the binding of the masked antibody were >300 μg / mL.
[0334] Example 5B: Masked / unmasked ADC-S105 induces cytotoxicity on HER2-positive SK-BR-3 cells.
[0335] As shown in Table 6, ADC-S105 is a masked trastuzumab (P41C)-based ADC based on the masked antibody S105 mentioned in Table 1, which carries a masking moiety comprising a BC2 nanobody on its heavy chain (BC2-Nb-L9-HC (P41C)) and a masking moiety comprising a BC2T peptide preceded by a Qg sequence on its LC (Qg-BC2T-L9-LC). Both masking moieties are connected to the antibody via an MMP2-cleavable linker L9.
[0336] The cytotoxicity of masked and unmasked (MMP2 treated) ADC-S105 was tested on HER2 positive SK-BR-3 cells to investigate whether the ADC was effectively masked and whether the reduction in cell viability after unmasking ADC-S105 with MMP2 protease was comparable to that of naked (unmasked trastuzumab 41C based) ADC. Therefore, for comparison, unmasked ADCs (ADC-TRA41C and ADC-RIT41C, as shown in Table 6) based on trastuzumab and rituximab (as isotype controls) were also tested.
[0337] After 6 days of treatment, CellTiter-Glo TM (CTG) luminescence assay kit to measure cell viability.
[0338] Figure 3 Dose-response curves related to cytotoxicity induced by ADC-S105 when masked and after MMP2 treatment (unmasked) are shown, compared to ADC-TRA41C and isotype control ADC-RIT41C. Data show the percentage of survival for both experiments. Table 13 shows the calculated IC determined using CTG readout after 6 days of treatment of HER2-positive SK-BR-3 cells with masked ADC-S105 and its unmasked (MMP2-treated) ADC. 50 IC induced by masking concept 50 The potency shift is relative to ADC-TRA41C. Masked ADC-S105 showed a 73-fold IC relative to ADC-TRA41C. 50 When ADC-S105 was unmasked after MMP2 protease treatment, almost complete restoration of cytotoxicity induction was observed (unmasked ADC-S105 showed a 2-fold IC relative to ADC-TRA41C). 50 offset).
[0339] Table 13: IC determined by CTG readout after 6 days of treatment of HER2-positive SK-BR-3 cells with masked ADC-S105 and its unmasked ADC 50 .
[0340] Compound <![CDATA[IC 50 (μg / mL)]]> Effectiveness deviation ADC-TRA41C 0.016 n / a ADC-S105 Masking 1.17 73 (masked vs naked) ADC-S105 Demasking 0.030 2 (unmasked vs. bare)
[0341] Example 6: Cellular binding of ADC-S423 and ADC-S426 (non-cleavable linker L13) treated with masking / MMP2 to HER2-positive SK-BR-3 cells compared to ADC-TRA41C
[0342] Cellular binding of masked and MMP2 treated ADCs ADC-S423 and ADC-S426 to HER2 positive SK-BR-3 cells was compared to ADC-TRA41C binding (+ / - MMP2 treatment). Dose dependent binding on HER2 positive SK-BR-3 cells was measured on a flow cytometer.
[0343] To investigate whether the cleavage of the L9 linker connecting the masking moiety to the S423 antibody in ADC-S423 is MMP2-dependent, ADC-S426 was also tested. In the masked S426 antibody, the same masking moiety as used in S423 is linked to the antibody via a scrambled form of the L9 linker (L13: GGGGSLALGPGGGS, SEQ ID NO: 12). L13 no longer contains the MMP2 cleavable site present in L9. (Trang et al. Nature biotechnology 37.7 (2019): 761-765, Jiang et al. Proceedings of the National Academy of Sciences 101.51 (2004): 17867-17872).
[0344] Both ADCs are shown in Table 6. For comparison, a naked (unmasked) ADC based on trastuzumab (ADC-TRA41C) was also subjected to MMP2 treatment.
[0345] The results are Figure 5 , demonstrating dose-dependent binding of ADC-S423 and ADC-S426 on HER2-positive SK-BR-3 cells after masking and MMP2 treatment (unmasking) compared to ADC-TRA41C binding (+ / -MMP2 treatment). Data show MFI (median fluorescence intensity) from one experiment.
[0346] Masked ADC-S423 showed 540-fold EC relative to ADC-TRA41C 50ADC-S426 showed a 137-fold higher EC50 relative to ADC-TRA41C. 50 , and HER2 target binding was not restored after MMP2 protease treatment.
[0347] The results show that for both masked ADCs (ADC-S423 and ADC-S426), trastuzumab-mediated binding on HER2-positive SK-BR-3 cells was effectively masked. After unmasking by MMP2 treatment, binding of ADC-S423 was restored. The data also show that linker L13 is insensitive to MMP2, as binding was not restored after MMP2 treatment.
[0348] Example 7: Cytotoxicity of Masked / MMP2-treated ADC-S423 and ADC-S426 on HER2-positive SK-BR-3 cells
[0349] To investigate the effect of masked ADC-S423 on cell viability after unmasking with MMP2 protease, the cytotoxicity of masked and unmasked (MMP2 treated) ADC-S423 and ADC-S426 was tested on HER2 positive SK-BR-3 cells and compared with naked (unmasked) trastuzumab based ADC-TRA41C and isotype control ADC-RIT41C. TM (CTG) luminescence assay kit to measure cell viability.
[0350] The results are Figure 4 , where cytotoxicity (shown as % cell survival) was induced on HER2-positive SK-BR-3 cells by different concentrations of ADCs ADC-S423, ADC-S426, ADC-TRA41C and isotype control ADC-RIT41C.
[0351] The resulting IC values were determined using CTG readout after 6 days of treatment of HER2-positive SK-BR-3 cells with masked ADC-S423 and ADC-S426 and the respective MMP2-treated ADCs. 50 Shown in Table 14.
[0352] Masked ADC-S423 shows 36 times IC relative to ADC-TRA41C 50 However, when unmasked after MMP2 protease treatment, an almost complete restoration of the induction of cytotoxicity was observed (1.3-fold IC of unmasked ADC-S423 relative to ADC-TRA41C). 50 offset).
[0353] The scrambled linker in (ADC-)L026S426 is not cleavable by MMP2, as can be seen by the comparable 30-fold and 27-fold shifts of masked and MMP2-treated ADC-S426 relative to ADC-TRA41C.
[0354] Table 14: Cellular binding to HER2-positive SK-BR-3 cells (EC 50 ) and cytotoxicity (IC 50 )
[0355]
[0356] *DRC for ADC-S426 does not show top because the highest concentration tested in the binding assay was 100 μg / mL.
[0357] Example 8: Cell binding to HER2-positive SK-BR-3 cells by half-masked Ab (A) and half-demasked ADC (B), and cytotoxicity of half-(de)masked ADC (C) to HER2-positive SK-BR-3 cells.
[0358] To investigate whether the half-masked Ab or ADC is still able to bind to its target and induce cell death as effectively as the original (unmasked) Ab or ADC, the cell binding and cytotoxicity of the half-masked Ab and ADC were tested on HER2-positive SK-BR-3 cells. "Half-masking" in this experiment means that the masking moiety containing the Nb or antigenic peptide is conjugated to the HC or LC of the "half-masked" antibody. 8A: Cell binding of half-masked Ab to HER2-positive SK-BR-3 cells
[0359] Cell binding of the semi-masked antibodies S109 and S110 was compared to that of trastuzumab. Dose-dependent binding of the antibodies on HER2-positive SK-BR-3 cells was measured on a flow cytometer. Fig. 7A The data shown show the MFI (median fluorescence intensity) from two experiments. Table 15 shows the binding characteristics of the semi-masked antibodies S109 and S110, expressed as binding EC on HER2 positive SK-BR-3 cells. 50 and EC relative to trastuzumab 50 Binding of the half-masked Ab to the HER2 target was still possible, but the potency was reduced by more than 4-fold compared to trastuzumab.
[0360] Table 15: Cellular binding of semi-masked antibodies S109 and S110 compared to trastuzumab on HER2-positive SK-BR-3 cells
[0361]
[0362] *Incomplete DRC
[0363] 8B: Cellular binding of (semi-unmasked) ADC-S424 and ADC-S425 on HER2-positive SK-BR-3 cells compared to ADC-TRA41C.
[0364] ADC ADC-S424 and ADC-S425 have complete masking, with one masking part connected to a non-cleavable linker (detailed in Table 16). In ADC-S424, the BC2T antigen peptide is connected to the HC with a non-cleavable linker, while in ADC-S425, Nb BC2 is connected to the LC with a non-cleavable linker. ADC ADC-S424 and ADC-S425 are fully masked before MMP2 treatment and "half-unmasked" after MMP2 treatment. Cell binding experiments on HER2-positive SK-BR-3 cells were performed as described above for semi-masked antibodies S109 and S110. MMP2 treatment of ADC-S424 and ADC-S425 was performed as described above.
[0365] For comparison, a trastuzumab-based naked (unmasked) ADC (ADC-TRA41C) and an isotype control (ADC-RIT41C) were used.
[0366] The results are Figure 7B , demonstrating dose-dependent binding of ADC-S424 and ADC-S425 on HER2-positive SK-BR-3 cells after masking and MMP2 treatment (semi-unmasking) compared to ADC-TRA41C and ADC-RIT41C binding. Data show MFI (median fluorescence intensity) from one experiment.
[0367] Masked ADC-S424 and ADC-S425 barely bind to cells, as shown by very flat dose response curves. Clearly, after MMP-2 treatment, with only one masking moiety removed, cell binding of both semi-unmasked ADC-S424 and ADC-S425 is largely restored, but not fully restored to the level of naked ADC-TRA41C. The results are shown in Table 16.
[0368] Table 16: Cellular binding of semi-(de)masked ADCs ADC-S424 and ADC-S425 on HER2-positive SK-BR-3 cells compared to ADC-TRA41C and isotype control ADC-RIT41C
[0369]
[0370] 8C: Cytotoxicity of semi-(de)masked ADCs against HER2-positive SK-BR-3 cells compared to ADC-TRA41C and isotype control ADC-RIT41C.
[0371] The cytotoxicity induced by the semi-masked ADCs ADC-S109, ADC-S110, ADC-S111 (detailed in Table 6) as well as ADC-S424 and ADC-S425 on HER2-positive SK-BR-3 cells was compared with ADC-TRA41C and the isotype control ADC-RIT41C.
[0372] ADCs ADC-S109, ADC-S110 and ADC-S111 are based on antibodies with only one masking moiety (comprising Nb BC2). These ADCs are therefore "semi-masked". ADCs ADC-S109, ADC-S110 and ADC-S111 are not treated with MMP2.
[0373] The cytotoxicity of ADC-S424 and ADC-S425 was determined before and after MMP2 treatment. After MMP2 treatment, the masking moiety attached via a non-cleavable linker will remain attached to the antibody in ADC-S424 and ADC-S425 (and the ADC is "half-unmasked").
[0374] The results are Figure 7B (for ADC-S109, ADC-S110, and ADC-S111) and Figure 7C (for ADC-S424 and ADC-S425). After 6 days of treatment, the cells were treated with CellTiter-Glo TM Cell viability was measured using the (CTG) luminescence assay kit. Data show the percentage of survival from two experiments.
[0375] Table 17 shows the IC values determined using CTG readout after 6 days of treatment of HER2-positive SK-BR-3 cells with semi-masked ADCs ADC-S109, ADC-S110, ADC-S111, ADC-S424, ADC-S425, and ADC-TRA41C (ADCs shown in Table 6). 50 IC induced by the concept of half-masking 50 Potency shifts are relative to ADC-TRA41C.
[0376] Table 17: Cytotoxicity of semi-masked ADC-S109, ADC-S110 and ADC-S111 on HER2-positive SK-BR-3 cells
[0377]
[0378] After target binding, the half-masked ADCs ADC-S109, ADC-S110, ADC-S111 induced efficient cell death but with a 2-fold decrease in potency compared to ADC-TRA41C, indicating that cleavage of one linker is sufficient to achieve almost similar cell killing potency as the unmasked ADC. No differences in binding or cytotoxic potency were observed that correlated with differences in masking of the half-masked Abs and ADCs ADC-S109, ADC-S110, ADC-S111.
[0379] Table 18 shows the IC determined using CTG readout after 6 days of treatment of HER2-positive SK-BR-3 cells with ADCs ADC-S424, ADC-S425, and ADC-TRA41C (ADCs shown in Table 6). 50 IC induced by the concept of half-masking 50 The effectiveness offset is relative to the corresponding masked ADC. (ADC-TRA41C has no top in the DRC and therefore the IC 50 ).
[0380] Table 18: IC values of ADC-S424, ADC-S425 and ADC-TRA41C on HER2-positive SK-BR-3 cells before and after MMP treatment 50 .
[0381]
[0382] Semi-unmasked ADCs ADC-S424, ADC-S425 also induced efficient cell death after MMP2 treatment when compared to their fully masked (before MMP2 treatment) counterparts, but with 11-fold reduced potency compared to ADC-TRA41C.
[0383] Example 9: HMW measurements in the generation of masked ADCs with different DARs
[0384] In order to study the effect of mutations in the BC2-Nb sequence and the BC2T peptide sequence on HMW (%) in the generation of ADCs with different drug-to-antibody ratios (DARs), ADCs with different DARs were synthesized based on masked antibodies S103 and S211. S103 carries a mask based on BC2-Nb and BCT2 peptides, while S211 carries a mask in which mutations were introduced in BC2-Nb as well as BCT2 peptides. S103 is detailed in Table 1, while S211 is detailed in Table 3. For convenience, both antibodies are again described in Table 19:
[0385] Table 19: Details of antibodies S103 and S211.
[0386]
[0387] ADCs based on masked antibodies S103 and S211 carrying linker-drug LD2 (Elgersma et al., supra) with different DARs were synthesized according to the following procedure:
[0388] The solution of masked antibody was mixed with different aliquots of buffer (4.2 mM histidine, 50 mM trehalose, pH 6), 1 M TRIS*HCl (pH 8), 25 mM EDTA solution, and 2 mM TCEP solution. Different aliquots of TCEP were used to obtain a range of DARs after conjugation (different amounts of TCEP yield different amounts of thiols available for conjugation, and the amount of thiols determines the amount of conjugated linker-drug).
[0389] Final conditions were 2 mg / mL antibody, 10 mM TRIS, 1 mM EDTA, TCEP (1 to 1.6 molar equivalents relative to antibody). N,N-dimethylacetamide (DMA) was added followed by the linker-drug solution (10 mM in DMA, 3 molar equivalents of linker-drug relative to TCEP). The final concentration of DMA was 5%.
[0390] The resulting mixture was incubated at RT in the dark for 3 hours, after which a sample was taken for analysis by size exclusion chromatography (SEC) for high molecular weight according to the following method:
[0391] For soluble protein aggregation analysis of ADC, a SEC method with UV detection was used. The (masked) antibody was diluted to a concentration of 1 mg / mL and injected on a Waters Acquity UPLC BEH200 SEC using 100 mM PO 4 Buffer pH 7.5 with 10% 2-propanol was used as mobile phase. The percentage of the main peak (monomer) and high molecular weight (HMW) species (eg, soluble aggregates) was determined based on the peak area relative to the total peak area of all protein related peaks.
[0392] The mean DAR of the ADC was determined by HIC chromatography as described in Coumans et al, 2020, supra.
[0393] Fig. 9 The high molecular weight relative to the average DAR of the tested ADC is shown in . The tested range of DAR is shown on the X-axis, while the percentage of HMW in the obtained product is shown on the Y-axis.
[0394] It is evident that the HMW (%) of ADC-S211 is lower in the DAR range tested when compared to ADC-S103.
[0395] Example 10: Linker selection (for masked anti-TF antibodies)
[0396] Using the linkers as exemplified in Table 18 in all possible combinations, 324 masked antibody groups were generated using a method similar to that described in Example 1. The masking moiety used in each case was a pair of masking moieties comprising a masking moiety with a BC2T SA mutant peptide linked to each LC of the anti-TF antibody Tisomab, and a masking moiety with a BC2-Nb RS mutant nanobody linked to each HC of the anti-TF antibody Tisomab. In each antibody, the sequence of the antigenic peptide was preceded by a Qg sequence.
[0397] In the following joint group in Table 20, joints 14 to 25 all contain the protease cleavage site LSGRSDNH, but of different lengths, while joints L26, L27, L9, L29, L7 and L30 all contain the PLGLAG protease cleavage site and are likewise of different lengths. The joint group shown in the above table allows the testing of combinations of joints that have the same cleavage site for both masking moieties despite the different joint lengths. The group also allows combinations of joints with different cleavage sites and / or lengths, and of course also encompasses combinations using exactly the same joint for both masking moieties. By testing all possible combinations (18×18=324 combinations), the best joint pair can be selected, for example as used to optimize expression levels and / or HMW (%).
[0398] For all antibodies produced, the amount of antibody produced (μg / mL), HMW (%) (percentage of total antibody lost due to formation of high molecular weight complexes) were determined.
[0399] Table 20: Sets of adapters used in each possible combination
[0400] Linker Configuration Connectors SEQ ID NO: LSGRSDNH-Ab L14 14 GGLSGRSDNH-Ab L15 55 SGGLSGRSDNH-Ab L16 56 GSGGLSGRSDNH-Ab L17 57 GGLSGRSDNHGGS-Ab L18 58 SGGLSGRSDNHGGS-Ab L19 59 GSGGLSGRSDNHGGS-Ab L20 60 GGSGGLSGRSDNHGGS-Ab L21 61 GGGSGGLSGRSDNHGGS-Ab L22 62 GGGGSGGLSGRSDNHGGS-Ab L23 63 GGGGSGGSGGLSGRSDNHGGS-Ab L24 64 GGGSSGGSGGSGGLSGRSDNHGGS-Ab L25 65 GGGSGGSPLGLAGGGS-Ab L26 66 GGSGGSPLGLAGGGS-Ab L27 67 GGGGSPLGLAGGGS-Ab L9 11 GGGSPLGLAGGGS-Ab L29 68 GGSPLGLAGGGS-Ab L7 19 GGPLGLAGGGS-Ab L30 69
[0401] All antibodies were also tested for binding to their immobilized target tissue factor antigen in BLI essentially as described in Example 4.
[0402] The amount of antigen obtained by each antibody is Fig. 10A Shown in.
[0403] The HMW (%) of each antibody was Fig. 10B Shown in.
[0404] exist Fig. 10C In the Figure, the results of the BLI binding experiments are shown as binding to tissue factor / binding to protein A*1000.
[0405] The results of the control antibody S312 (naked teisomab) were also Fig. 10A to C are provided in the table shown.
[0406] By setting a predetermined cutoff value for each parameter measured and combining the results of the three plots, the choice of linker combination for that particular antibody can be determined.
[0407] It is obvious that most antibodies can be produced at expression levels greater than 100 μg / mL (264 out of 324). For the vast majority of antibodies, the HMW (%) is less than 10%, while for 86 linker combinations, the HMW (%) is less than 5%, and for 13 even less than 3%.
[0408] When the binding in the BLI experiments was compared with that of the parental antibody, a significant reduction in binding was observed for all masked antibodies tested; the binding parameter for the parental antibody tesomab was approximately 750, whereas that for all masked antibodies was less than 150, and for almost 50% even less than 40.
[0409] When the three parameters were combined (expression level >100 μg / mL, HMW (%) less than 4%, and BLI binding parameter less than 40), 52 antibodies still passed the test.
[0410] When the standard of HMW (%) was set to less than 3%, there were still 14 antibodies that met all the standards. 10 of these 14 antibodies could be produced at expression levels greater than 200 μg / mL.
[0411] Of course, further comparisons of subsets of antibodies selected as described above can be performed, for example in cell binding assays and / or by using cytotoxicity measurements.
[0412] This experiment shows that good results can be obtained with different linker lengths, and with combinations of linkers with the same or different protease cleavage sites.
[0413] Example 11: ADCs of 12 selected antibodies (from Example 10)
[0414] Of the masked antibodies tested in Example 10, 12 were selected for conjugation with DUBA to generate masked antibody drug conjugates (masked ADCs).
[0415] Selected antibodies and their linkers are shown in Table 21.
[0416] As described in Example 3, ADCs were prepared based on these antibodies by subjecting them to typical selective reduction and DUBA conjugation conditions.
[0417] For all masked antibodies, only a minor increase in HMW (%) was observed, and the final formulated form contained 3% to 8% HMW species with a DAR of 1.9. The HMW of the non-masked ADC was 1.4% and the DAR was also 1.9.
[0418] Table 21: Selected conjugates based on masked 41C teisomab antibody
[0419] ADC LC-Connector HC-Connector Protease combination DAR HMW(%) ADC 1 L14 L26 Mat / MMP 1.85 3.0 ADC2 L14 L9 Mat / MMP 1.88 3.0 ADC3 L1 8 L20 Mat / Mat 1.87 6.3 ADC4 L21 L20 Mat / Mat 1.85 6.2 ADC5 L20 L9 Mat / MMP 1.87 5.2 ADC6 L29 L1 5 MMP / Mat 1.86 6.9 ADC7 L19 L24 Mat / Mat 1.86 5.5 ADC8 L30 L27 MMP / MMP 1.84 5.6 ADC9 L9 L30 MMP / MMP 1.88 6.6 ADC10 L29 L30 MMP / MMP 1.87 7.1 ADC 11 L19 L9 Mat / MMP 1.88 8.1 ADC12 L14 L30 Mat / MMP 1.85 5.1 ADC13 na na na 1.85 1.4
[0420] Example 12: In vitro characterization of 12 selected masked anti-TF ADCs.
[0421] The 12 ADCs shown in Table 21 were tested in vitro.
[0422] The cytotoxicity of the 12 ADCs was tested essentially as described in the Materials and Methods in Example 5. Because the antibody used in ADC1-ADC13 was the anti-tissue factor (TF) antibody tesomonab, the cytotoxicity against cell lines expressing TF (rather than the HER2-positive tumor cell lines used in Example 5) was determined. Three cell lines were used to study cytotoxicity: HCT-116, BxPC3, and FaDu, whose antigen (TF) densities were about 30.000, about 200.000, and about 400.000, respectively, classified as low, "medium," and high. Antibodies S312 (naked tesomonab) and a second F(ab') conjugated to APC were calibrated with a human IgG calibration kit (Biocytex, Marseille, France) according to the manufacturer's protocol. 2 Goat anti-human IgG (Fc fragment specific) antibody was used to determine TF antigen expression on the surface of human tumor cell lines.
[0423] There appears to be a correlation between antigen expression levels and masked ADC activity:
[0424] In the low-expressing HCT116 cell line, the masked ADC had a curve that overlapped with the isotype (non-binding) control ADC, meaning that at low levels of target density (e.g., healthy tissue), little to no target-mediated activity would be expected.
[0425] In the medium and high expressers BxPC3 and FaDu, the masked ADCs showed higher activity than the isotype control (as evidenced by smaller IC values between the masked ADCs and the unmasked / semi-unmasked ADCs). 50 This means that at higher antigen concentrations (e.g. in tumors), masked ADCs can already start to exert anti-tumor effects (or at least show tumor accumulation) without unmasking.
[0426] Semi-demasked ADCs have IC equivalent to fully demasked ADCs50 value.
[0427] All ADCs (masked vs unmasked) for TF-negative BT474 and IncuCyte TM Jurkat NucLight TM Red lacked cytotoxicity (data not shown).
[0428] Differences in activity and masking efficiency could be observed between the individual masked ADCs (the linker used to attach the masking moiety to the antibody was different from one another), but all ADCs were effectively masked.
[0429] Unmasking and partial unmasking resulted in regaining the cell-killing ability of the ADC.
[0430] The results of the cytotoxicity test are shown in Table 22 as IC 50 Offset.
[0431] For each ADC (ADC1 to 12), the IC is provided compared to the dose response curve of the unmasked ADC13. 50 Offset (column 1). This value provides an indication of the efficiency of the masking.
[0432] In the next column, the IC of the unmasked ADC (protease treated) when compared to the unmasked (naked) ADC13 is shown. 50 The values here are very small, indicating that the unmasked ADC binds to the target antigen in the same way as the naked ADC (which carries no masking to begin with).
[0433] The next two columns provide the results of treating ADC1 to 12 with only one protease (MMP2 or protease). For ADCs constructed with linker pairs in which one linker comprises a MMP cleavage site and the other linker comprises a protease cleavage site (ADC1, ADC2, ADC5, ADC6, ADC11 and ADC12), cleavage with one or the other protease results in a semi-(de)masked ADC, since only one linker is cleaved and thus only one masking moiety is disconnected from the antibody. The IC values of only these semi-(de)masked ADCs when compared to the dose response curve of the unmasked ADC13 are shown in Table 22. 50 Offset.
[0434] As can be seen in Table 22, half (de)masking results in a dose response comparable to that of full demasking and also comparable to that of unmasked ADC13 (given by IC 50 This clearly demonstrates that binding to the target antigen has been restored after cleavage of only one of the linkers masking the ADC.
[0435] It is evident from the table that (semi-) unmasking of the masked ADCs results in a shift in the dose response curve, indicating that all masked ADCs are effectively masked and effectively unmasked.
[0436] Table 22: Cytotoxicity measurement results showing IC for masked, semi(de)masked and unmasked ADCs 50 Offset.
[0437]
[0438] Example 13: In vivo testing of masked ADC
[0439] Xenotransplantation studies were performed at AAALAC (Association for Assessment and Accreditation of Laboratory Animal Care) accredited CROs using well-established protocols developed and maintained by these CROs.
[0440] Masked anti-TF ADCs 1 to 13 were tested in vivo in the patient-derived xenograft (PDX) model non-small cell lung cancer model LXFA629. TF expression in this model was assessed using immunohistochemistry (IHC). Tumor tissues were fixed in formalin, embedded in paraffin, and used for TF IHC using a rabbit monoclonal anti-TF antibody from Cell Signaling Technologies reference number 97438. IHC staining showed TF 2+ expression levels.
[0441] LXFA629 tumors were cut into fragments (3 to 4 mm side length) and implanted unilaterally SC in the flank of immunodeficient CES1c KO SCID mice. 3 When the target range was within 500 μg / mL, mice were randomly assigned to treatment groups (n=6 / group) for comparable median and mean group tumor volumes. The next day, mice were dosed with a single IV dose of 2 or 2.4 mg / kg ADC in a buffer containing 4.2 mM histidine, 50 mM trehalose, pH 6.0 (buffer SOL092) via the tail vein.
[0442] The antitumor activity of ADC was evaluated by measuring tumor size twice a week until day 42. The length and width of the tumor were measured with calipers, and the tumor volume was estimated by the following formula (Simpson-Herren et al., 1970):
[0443]
[0444] Individual mice were sacrificed when their tumors reached 2000 mm3, and data from this time point are no longer shown for all groups.
[0445] All masked ADCs showed potency in the same range as that of the unmasked anti-TF ADC.
[0446] The results are Fig.11 Each data point represents the absolute tumor volume (mm) of all mice in each group. 3 )’s average value.
[0447] exist Fig.12 The absolute tumor volumes of individual mice in each group on the last day of the experiment (day 42) are shown in . The lines represent the mean values.
[0448] Example 14: Use of alternative Nb / peptide tags in masking moieties
[0449] The masking part is designed based on the "ALFA tag" described in WO2020053239.
[0450] To design the masking moiety, a mutant with lower affinity for the ALFA tag peptide was selected. PE The sequence of is disclosed in WO2020053239 as SEQ ID NO: 134 (SEQ ID NO: 91). The second masking moiety is based on a peptide sequence comprising an ALFA tag (SEQ ID NO: 92).
[0451] Based on the use of these sequences in the masking moiety, the anti-HER2 antibody trastuzumab was masked, whereby the peptide tag was attached to the N-terminus of the HC using a "Qg" spacer sequence and a linker containing the PLGLAG cleavage site, while the Nb PE Attached to LC. A set of linkers (L12, L9, L7, L34 and L5) was used in all possible combinations, resulting in a set of 25 masked antibodies. The linkers used are shown in Table 23.
[0452] Table 23: Linkers for ALFA-tag based MMs
[0453] Connector (-Ab) Connector No. SEQ ID NO: GGGGS GGGGSPLGLAGGGS-Ab L12 93 GGGGSPLGLAGGGS-Ab L9 11 GGSPLGLAGGGS-Ab L7 1 9 PLGLAGGGS-Ab L34 94 PL GLAG-Ab L5 13
[0454] Masked antibodies were produced and purified. After purification, the expression levels and percentages of HMW complexes were determined as described in Example 1. The expression levels of all produced masked antibodies were greater than 100 μg / mL, and the expression levels of most antibodies were greater than 200 μg / mL. For all joint combinations tested, HMW (%) was all very low (about 1%). The expression levels are shown in Table 24, and the HMW (%) is shown in Table 25.
[0455] Table 24: Expression levels (μg / mL) of 25 masked trastuzumab antibodies, where masking is based on the ALFA tag (Nb PE ) and ALFA-tagged peptide (Tag ALFA )
[0456]
[0457] Table 25: HMW (%) complexes of 25 masked trastuzumab antibodies, where masking is based on the ALFA tag (Nb PE ) and ALFA-tagged peptide (Tag ALFA )
[0458]
[0459] Binding of masked antibodies to immobilized HER2 protein was measured using the BLI method essentially as described in Example 4. BLI measurements showed that the binding of masked antibodies to immobilized HER2 protein was significantly reduced by using ALFA-based tags (Nb PE ) and ALFA-tagged peptide (Tag ALFA ) masking part to mask the trastuzumab antibody, which can achieve effective shielding.
[0460] The BLI binding results for all 25 antibodies are shown in Table 25. Although all antibodies were effectively masked (binding to HER-2 was significantly reduced), it appears that masking with shorter linkers (L5, L34, and L7) was most effective. For comparison; the BLI value obtained for unmasked (naked) trastuzumab was 331.2.
[0461] Table 25: BLI binding results of 25 masked trastuzumab antibodies, where masking is based on the ALFA tag (Nb PE ) and ALFA-tagged peptide (Tag ALFA )
[0462]
[0463] Sequences in the sequence listing
[0464]
[0465]
[0466]
[0467]
[0468]
[0469]
[0470]
[0471]
[0472]
[0473]
[0474]
[0475]
[0476]
Claims
1. An activatable antibody or antigen-binding fragment thereof, comprising an antigen-binding site, the antigen-binding site comprising a heavy chain variable domain and a light chain variable domain capable of binding to a target protein, wherein the antigen-binding site has a mask suitable for inhibiting binding of the antigen-binding site to the target protein, the mask comprising two masking parts: i. a masking moiety connected to the N-terminus of the variable heavy chain domain, and ii. a masking moiety connected to the N-terminus of the variable light chain domain, At least one of the masking moieties is connected by a cleavable linker, and one of the masking moieties comprises an antigenic peptide and the other masking moiety comprises a single domain antibody (sdAb) that specifically binds to the antigenic peptide.
2. The activatable antibody or antigen-binding fragment thereof according to claim 1, wherein all masking moieties are connected via cleavable linkers.
3. The activatable antibody or antigen-binding fragment thereof according to claim 1 or 2, wherein the cleavable linker is a protease-cleavable peptide linker.
4. The activatable antibody or antigen-binding fragment thereof according to any one of the preceding claims, wherein the sdAb is a nanobody.
5. The activatable antibody or antigen-binding fragment thereof according to any one of the preceding claims, comprising two antigen-binding sites, wherein both antigen-binding sites are masked.
6. The activatable antibody or antigen-binding fragment thereof according to any one of the preceding claims, wherein the antigenic peptide has a length of 4 to 20 amino acid residues.
7. The activatable antibody or antigen-binding fragment thereof according to any one of the preceding claims, wherein the antigenic peptide has a length of 12 to 20 amino acid residues and comprises or consists of the amino acid sequence defined in SEQ ID NO:
51.
8. The activatable antibody or antigen-binding fragment thereof according to any one of the preceding claims, wherein the antigenic peptide comprises the amino acid sequence defined in SEQ ID NO: 1, SEQ ID NO: 2 (SA mutant) or SEQ ID NO: 3 (ST mutant).
9. An activatable antibody or antigen-binding fragment thereof according to any one of claims 4 to 8, wherein the Nanobody comprises the amino acid sequence defined in SEQ ID NO: 4, a humanized form thereof, or a variant of SEQ ID NO: 4, or a humanized form of said variant having an amino acid sequence defined in SEQ ID NO: 76 (RA mutant), SEQ ID NO: 77 (RE mutant) or SEQ ID NO: 78 (RS mutant).
10. An activatable antibody or antigen-binding fragment thereof according to claim 9, wherein the antigenic peptide comprises or consists of a peptide having an amino acid sequence as defined in SEQ ID NO: 2 (SA mutant), and the Nanobody comprises an amino acid sequence as defined in SEQ ID NO: 78 (RS mutant) or a humanized form thereof.
11. The activatable antibody or antigen-binding fragment thereof according to any one of the preceding claims, wherein the masking portion comprising the sdAb is linked to the N-terminus of the light chain variable domain, and the masking portion comprising the antigenic peptide is linked to the N-terminus of the heavy chain variable domain.
12. The activatable antibody or antigen-binding fragment thereof according to any one of the preceding claims, wherein the cleavable linker comprises one or more cleavage sites recognized by one or more tumor-specific proteases.
13. The activatable antibody or antigen-binding fragment thereof according to any one of the preceding claims, wherein at least one cleavable linker comprises a cleavage site recognized by a protease or a cleavage site recognized by a metalloprotease.
14. An antibody-drug conjugate (ADC), comprising the activatable antibody or antigen-binding fragment thereof according to any one of claims 1 to 13, and a linker-drug.
15. A pharmaceutical composition comprising the activatable antibody or antigen-binding fragment thereof according to any one of claims 1 to 13, or the ADC according to claim 14, and a pharmaceutically acceptable excipient.
16. The activatable antibody or antigen-binding fragment thereof according to any one of claims 1 to 13, the antibody-drug conjugate according to claim 14, or the pharmaceutical composition according to claim 15, for use as a medicament.
17. The activatable antibody or antigen-binding fragment thereof according to any one of claims 1 to 13, the antibody-drug conjugate according to claim 14, or the pharmaceutical composition according to claim 15, for use in treating cancer, autoimmune diseases or infectious diseases, preferably for treating cancer.
18. A nucleic acid construct comprising: a nucleotide sequence encoding a heavy chain variable domain, a protease cleavable peptide linker and a masking moiety; and / or a nucleotide sequence encoding a light chain variable domain, a cleavable peptide linker, and a masking moiety; The nucleotide sequence is operably linked to an expression control sequence for expression in a host cell, preferably a mammalian host cell.
19. A host cell comprising the nucleic acid construct according to claim 18.
20. A method for producing an activatable antibody or an antigen-binding fragment thereof according to any one of claims 1 to 13, the method comprising the step of culturing the host cell according to claim 19 under conditions that are favorable for expression of the activatable antibody or the antigen-binding fragment thereof.
Citation Information
Patent Citations
Modified antibody compositions, methods of making and using thereof
EP2385955A2
Epitope tag and method for detection, capture and / or purification of tagged polypeptides
EP3377632A1
Development of masked therapeutic antibodies to limit off-target effects; application to anti-EGFR antibodies
US8895702B2
Therapeutic antibodies with reduced side effects
WO2004009638A1
Partially loaded antibodies and methods of their conjugation
WO2005084390A2