Bispecificity checkpoint inhibitor antibodies

Abstract

To provide bispecific heterodimeric antibodies that bind to two different checkpoint cell surface receptors.SOLUTION: Disclosed are antibodies and the like that bind to PD-1 and CTLA-4, PD-1 and TIM-3, PD-1 and LAG-3, PD-1 and TIGIT, PD-1 and BTLA, CTLA-4 and TIM-3, CTLA-4 and LAG-3, CTLA-4 and TIGIT, CTLA-4 and BTLA, TIM-3 and LAG-3, TIM-3 and TIGIT, TIM-3 and BTLA, LAG-3 and TIGIT, LAG-3 and BTLA, and TIGIT and BTLA.SELECTED DRAWING: Figure 1A-E

Description

[Technical Field]

[0001] Cross-reference of related applications This application claims priority to U.S. Provisional Patent Application No. 62 / 350,145 filed June 14, 2016, No. 62 / 353,511 filed June 22, 2016, and No. 62 / 420,500 filed November 10, 2016, the contents of which are expressly and completely incorporated herein by reference in their entirety. Sequence List This application includes a sequence listing, which is filed electronically in ASCII format and is incorporated herein by reference in its entirety. A copy of this ASCII file, created on 9 June 2017, is named 067461_5191WO_SL.txt and has a size of 32,442,145 kilobytes. [Background technology]

[0002] Checkpoint receptors such as CTLA-4, PD-1 (programmed cell death 1), TIM-3 (T cell immunoglobulin and mucin domain 3), LAG-3 (lymphocyte activation gene 3), and TIGIT (T cell immune receptor with Ig and ITIM domains) inhibit the activation, proliferation, and / or effector activity of T cells and other cell types. Guided by the hypothesis that checkpoint receptors suppress the endogenous T cell response against tumor cells, preclinical and clinical studies of anti-CTLA4 and anti-PD1 antibodies, including nivolumab, pembrolizumab, ipilimumab, and tremelimumab, have indeed shown that in a small percentage of patients with various malignancies, checkpoint blockade results in a superior antitumor response, stimulating endogenous T cells to attack tumor cells and leading to long-term cancer remission. Unfortunately, only a fraction of patients respond to these therapies, and the response rate is typically in the range of 10-30%, depending on the indication and other factors, and sometimes higher for each monotherapy. Therapeutic combinations of these agonists, such as ipilimumab plus nivolumab, can yield even higher response rates, sometimes close to 60%. Preclinical studies have shown additional synergistic effects between anti-PD-1 antibodies and / or anti-CTLA-4 antibodies and blockade of recently identified checkpoint receptors, including LAG-3, TIM-3, BTLA, and TIGIT. While the potential of multiple checkpoint blockade is extremely promising, combination therapies with such agonists require a heavy financial burden. Furthermore, the autoimmune toxicity of combination therapies, such as nivolumab plus ipilimumab, is significantly increased compared to monotherapy, leading many patients to discontinue treatment.

[0003] Several studies have investigated tumor-infiltrating lymphocytes (TILs) (Ahmadzadeh et al.) TILs (TILs) have generally expressed multiple checkpoint receptors, as shown by al., Blood 114:1537 (2009), Matsuzaki et al., PNAS 107(17):7875-7880 (2010), Fourcade et al., Cancer Res. 72(4):887-896 (2012), and Gros et al., J. Clinical Invest. 124(5):2246 (2014). Furthermore, TILs expressing multiple checkpoints are likely to be the most tumor-responsive. In contrast, peripheral non-tumor-responsive T cells are more likely to express a single checkpoint. Checkpoint blockade using monospecific full-length antibodies is likely indiscriminate in desuppressing tumor-responsive TILs against mono-expressing T cells that are thought to contribute to autoimmune toxicity.

[0004] Therefore, the present invention relates to bispecific antibodies that bind to two different checkpoint inhibitor proteins. [Overview of the Initiative]

[0005] The present invention provides a bispecific heterodimer antibody that binds to two different checkpoint cell surface receptors, such as human PD-1, human CTLA-4, human TIM-3, human LAG-3, and human TIGIT. Thus, in some embodiments, a suitable bispecific antibody binds to PD-1 and CTLA-4, PD-1 and TIM-3, PD-1 and LAG-3, PD-1 and TIGIT, PD-1 and BTLA, CTLA-4 and TIM-3, CTLA-4 and LAG-3, CTLA-4 and TIGIT, CTLA-4 and BTLA, TIM-3 and LAG-3, TIM-3 and TIGIT, TIM-3 and BTLA, LAG-3 and TIGIT, LAG-3 and BTLA, and TIGIT and BTLA.

[0006] In one embodiment, the present invention provides a bottle opener type comprising a) an scFv having a variable heavy domain and a variable light domain linked using a charged scFv linker (in some embodiments, the +H sequence in Figure 7 is preferred), an Fc domain including the asymmetric mutant S364K / E357Q and the cleavage mutant E233P / L234V / L235A / G236del / S267K, and a first monomer (sometimes referred to as the "scFv heavy chain") comprising an Fv that binds to a checkpoint receptor outlined herein. The invention comprises a) a "scFv monomer" referred to as a second monomer ("Fab monomer" or "heavy chain") which includes a variable heavy domain that, together with an Fc domain having the asymmetric mutant L368D / K370S, the pI mutant N208D / Q295E / N384D / Q418E / N421D, the cleavage mutant E233P / L234V / L235A / G236del / S267K, and a variable light domain, makes up an Fv that binds to a second checkpoint receptor outlined herein, and a) a light chain. In this particular embodiment, preferred monomer Fv pairs include (Fab listed first, followed by scFv) PD-1 and CTLA-4, CTLA-4 and PD-1, PD-1 and TIM-3, TIM-3 and PD-1, PD-1 and LAG-3, LAG-3XPD1, PD-1 and TIGIT, TIGIT and PD-1, PD-1 and BTLA, BTLA and PD-1, CTLA-4 and TIM-3, TIM-3 and CTLA-4, CTLA-4 and LAG-3, LAG-3 and CTL Examples include A-4, CTLA-4 and TIGIT, TIGIT and CTLA-4, CTLA-4 and BTLA, BTLA and CTLA-4, TIM-3 and LAG-3, LAG-3 and TIM-3, TIM-3 and TIGIT, TIGIT and TIM-3, TIM-3 and BTLA, BTLA and TIM-3, LAG-3 and TIGIT, TIGIT and LAG-3, LAG-3 and BTLA, BTLA and LAG-3, BTLA and TIGIT, and TIGIT and BTLA.

[0007] Other aspects of the present invention are provided herein. [Brief explanation of the drawing]

[0008] [Figure 1A-I] Several types of the present invention are shown. The first type is the “bottle opener” type, which has first and second anti-antigen binding domains. In addition, all types of mAb-Fv, mAb-scFv, central-scFv, central-Fv, single-arm central-scFv, 1-scFv-mAb, scFv-mAb, and double scFv types are shown. For all of the scFv domains shown, they can be either N-to-C-terminal variable heavy-(any linker)-variable light, or vice versa. In addition, for single-arm scFv-mAb, the scFv can be bound to either the N-terminus of the heavy chain monomer or the N-terminus of the light chain. [Figure 2] (Figures 2A, 2B, 2C, and 2D) To facilitate the development of antigen-binding domains that bind to both human and cynomolgus monkey sequences for easier clinical development, antigen sequences of several antigens used in the present invention are shown, often including both human and cynomolgus monkey sequences. [Figure 3A-3F] A useful pair of heterodimerized mutants (including asymmetric and pI mutants) is shown. In Figure 3E, there are mutants for which the corresponding "monomer 2" mutant does not exist. These are pI mutants that can be used alone in either monomer or included in the Fab side of the bottle opener, and can be used, for example, in a second monomer where a suitable charged scFv linker utilizes scFv as the second antigen-binding domain. A suitable charged linker is shown in Figure 7. [Figure 4] A list of isoelectron-distributed mutant antibody constant regions and their respective substitutions is shown. pI_(-) indicates a lower pI mutant, while pI_(+) indicates a higher pI mutant. These can be optionally and independently combined with other dimerized mutants of the present invention (and with other mutant types as outlined herein). [Figure 5] This exhibits useful cleavage mutants (sometimes called "knockout" or "KO" mutants) that cleave the FcγR bond. Generally, cleavage mutants are found in both monomers, but in some cases, they may only be present in one monomer. [Figure 6] Two particularly useful embodiments of the present invention are shown, which can be used in either the type shown in Figure 1A or Figure 1F. For the type shown in Figure 1A, the “non-Fv” components of this embodiment are shown in Figure 37A, but the other types (and those in Figure 38) can be used in a similar manner. [Figure 7] Several charged scFv linkers useful in increasing or decreasing the pI of heterodimer antibodies utilizing one or more scFv as components are described herein. (+H) positive linkers are particularly useful herein, especially in conjunction with the anti-CD3vl and vh sequences shown herein. A single charge, single prior art scFv linker is referred to as "Whitlow," after Whitlow et al., Protein Engineering 6(8):989-995 (1993). It should be noted that this linker is used in scFv for reducing aggregation and enhancing stability against proteolysis. [Figure 8] A list of manipulated heterodimer asymmetric Fc mutants is shown, along with heterodimer yield (determined by HPLC-CIEX) and thermal stability (determined by DSC). Undetermined thermal stability is indicated as "nd". [Figure 9A-E]Several selected PD-1 ABDs are shown, along with additional anti-PD-1 ABDs listed as sequence numbers 6209-11464, 11465-17134, 33003-33072, 33073-35394, and 36127-36146. CDRs are underlined, scFv linkers are double-underlined (in the sequence, the scFv linker is a positively charged scFv(GKPGS)4 linker (sequence number 37755), but as will be understood by those skilled in the art, this linker may be replaced by other linkers, including uncharged or negatively charged linkers, some of which are shown in Figure 7), and diagonal lines indicate the boundary(s) of variable domains. In addition, the naming convention illustrates the orientation of scFv from N to C-terminus. That is, “H1.279_L1.194” indicates that the orientation is vh-scFv linker-vl (N to C-terminal, with an optional domain linker on one or both sides depending on the type used), but these sequences can also be used in the opposite orientation (N to C-terminal) vl-linker-vh. Similarly, “L1.194_H1.279” indicates that the orientation is vl-scFv linker-vh (N to C-terminal, similarly with an optional domain linker), and the opposite orientation is also included in the present invention. As is true for all sequences described herein that contain CDRs, the precise identification of the CDR location may vary slightly depending on the numbering used, as shown in Table 1, and therefore, not only the underlined CDRs but also CDRs contained within vh and vl domains using other numbering systems are included herein. Furthermore, with respect to all sequences in the figure, these vh and vl sequences can be used in either the scFv type or the Fab type. [Figure 10A-PP]Several CTLA-4 ABDs are shown, along with additional anti-CTLA-4 ABDs listed as SEQ ID NOs. 21-2918, 2919-6208, 36739-36818, and 35395-35416. CDRs are underlined, scFv linkers are double underlined (in the sequence, the scFv linker is the positively charged scFv(GKPGS)4 linker (SEQ ID NO. 37755), but as will be understood by those skilled in the art, this linker may be replaced by other linkers, including uncharged or negatively charged linkers, some of which are shown in Figure 7), and diagonal lines indicate the boundary(s) of variable domains. As described above, the naming convention illustrates the orientation of scFv from N to C-terminus. In the sequences listed in this figure, they are all oriented as vh-scFv linker-vl (N to C-terminal), but these sequences can also be used in the opposite orientation (N to C-terminal)vl-linker-vh. In addition, some of the sequences of sequence numbers 21-2918, 2919-6208, 36739-36818, and 35395-35416 are in the opposite orientation. As is true for all sequences described herein that contain CDRs, the precise identification of the CDR location may vary slightly depending on the numbering used, as shown in Table 1, and therefore, not only the underlined CDRs but also CDRs contained within vh and vl domains using other numbering systems are included herein. Furthermore, for all sequences in the figure, these vh and vl sequences can be used in either the scFv or Fab type. In particular, many of the figures include both scFv and Fab type XENP identifiers. See Figure 10A, for example, which shows that XENP19235 is a molecule using the Fab type and XENP19769 is an scFv molecule. [Figure 11A-N]Several LAG-3 ABDs are shown, along with additional anti-LAG-3 ABDs listed as SEQ ID NOs. 17135-20764, 36819-36962, 35417-35606, 25194-32793, and 32794-33002. CDRs are underlined, scFv linkers are double underlined (in the sequence, the scFv linker is a positively charged scFv(GKPGS)4 linker, but as will be understood by those skilled in the art, this linker may be replaced by other linkers, including uncharged or negatively charged linkers, some of which are shown in Figure 7), and diagonal lines indicate the boundary(s) of a variable domain. As described above, the naming convention illustrates the orientation of scFv from N to C-terminus. In the sequences listed in this figure, they are all oriented as vh-scFv linker-vl (N to C-terminal), but these sequences can also be used in the opposite orientation (N to C-terminal)vl-linker-vh. In addition, some of the sequences of sequence numbers 17135-20764, 36819-36962, 35417-35606, 25194-32793, and 32794-33002 are in the opposite orientation. As is true for all sequences described herein that contain CDRs, the precise identification of the CDR location may vary slightly depending on the numbering used, as shown in Table 1, and therefore, not only the underlined CDRs but also CDRs contained within vh and vl domains using other numbering systems are included herein. Furthermore, for all sequences in the figure, these vh and vl sequences can be used in either the scFv or Fab type. [Figure 12A-C]Several BTLA ABDs are shown, along with additional anti-BTLA ABDs listed as sequence numbers 20885-21503 and 36707-36738. CDRs are underlined, scFv linkers are double underlined (in the sequence, the scFv linker is a positively charged scFv(GKPGS)4 linker, but as will be understood by those skilled in the art, this linker may be replaced by other linkers, including uncharged or negatively charged linkers, some of which are shown in Figure 7), and diagonal lines indicate the boundary(s) of a variable domain. As described above, the naming convention illustrates the orientation of scFv from N to C-terminus. In the sequences listed in this figure, they are all oriented as vh-scFv linker-vl(N to C-terminus), but these sequences can also be used with the opposite orientation (N to C-terminus)vl-linker-vh. In addition, some of the sequences of sequence numbers 20885-21503 and 36707-36738 are in the opposite orientation. As is true for all sequences described herein that contain CDRs, the precise identification of the CDR location may vary slightly depending on the numbering used, as shown in Table 1, and therefore, not only the underlined CDRs but also CDRs contained within vh and vl domains using other numbering systems are included herein. Furthermore, with respect to all sequences in the figure, these vh and vl sequences may be used in either the scFv or Fab type. [Figure 13A-I]Several TIM-3 ABDs are shown, along with additional anti-TIM-3 ABDs listed as sequence numbers 20765-20884, 37587-37698, and 36347-36706. CDRs are underlined, scFv linkers are double underlined (in the sequences, the scFv linker is a positively charged scFv(GKPGS)4 linker, but as will be understood by those skilled in the art, this linker may be replaced by other linkers, including uncharged or negatively charged linkers, some of which are shown in Figure 7), and diagonal lines indicate the boundary(s) of a variable domain. As described above, the naming convention illustrates the orientation of scFv from N to C-terminus. In the sequences listed in this figure, they are all oriented as vh-scFv linker-vl(N to C-terminus), but these sequences can also be used with the opposite orientation (N to C-terminus)vl-linker-vh. In addition, some of the sequences of sequence numbers 30765-20884, 37587-37698, and 36347-36706 are in the opposite orientation. As is true for all sequences described herein that contain CDRs, the precise identification of the CDR location may vary slightly depending on the numbering used, as shown in Table 1, and therefore, not only the underlined CDRs but also CDRs contained within vh and vl domains using other numbering systems are included herein. Furthermore, with respect to all sequences in the figure, these vh and vl sequences may be used in either the scFv or Fab type. [Figure 14A-I]The amino acid sequences of specific anti-CTLA-4 X anti-PD-1 antibodies (Fab-scFv-Fc) are shown. The antibodies are named using a chain designation (Fab-Fc heavy chain, scFv-Fc heavy chain, or light chain), separated by a dash, first the Fab variable region, then the scFv variable region. CDRs are underlined, and slashes indicate the boundary(s) of the variable region. The scFv domain has a different orientation (N to C-terminus) of either vh-linker-vl or vl-linker-vh, as shown, but this can be reversed. In addition, each sequence outlined herein that results in a longer half-life in serum may contain or exclude the M428L / N434S variant in one, or preferably both, of the Fc domains. [Figure 15A-K] The amino acid sequences of specific anti-LAG-3X anti-PD-1 Fab-scFv-Fc bispecific antibodies are shown. The antibodies are named using the chain designation (Fab-Fc heavy chain, scFv-Fc heavy chain, or light chain), separated by a dash, first the Fab variable region, then the scFv variable region. CDRs are underlined, and slashes indicate the boundary(s) of the variable region. The scFv domain has an orientation (N to C-terminal) vl-linker-vh, which can be reversed. In addition, each sequence outlined herein that results in a longer half-life in serum may contain or exclude the M428L / N434S variant in one, or preferably both, of the Fc domains. [Figure 16] The amino acid sequences of specific anti-BTLA X anti-PD-1 Fab-scFv-Fc bispecific antibodies are shown. The antibodies are named using the chain designation (Fab-Fc heavy chain, scFv-Fc heavy chain, or light chain), separated by a dash, first the Fab variable region, then the scFv variable region. CDRs are underlined, and slashes indicate the boundary(s) of the variable region. The scFv domain has an orientation (N to C-terminal) vl-linker-vh, which can be reversed. In addition, each sequence outlined herein that results in a longer half-life in serum may contain or exclude the M428L / N434S variant in one, or preferably both, of the Fc domains. [Figure 17]The amino acid sequences of specific anti-LAG-3 X anti-CTLA-4 Fab-scFv-Fc bispecific antibodies are shown. The antibodies are named first by the Fab variable region, then the scFv variable region, separated by a dash, and subsequently using the chain designations (Fab-Fc heavy chain, scFv-Fc heavy chain, or light chain). CDRs are underlined and slashes indicate the variable region boundaries (if plural). The scFv domain has the orientation (N to C terminus) vh-linker-vl, which can be reversed. In addition, each sequence outlined herein that results in a longer half-life in serum can include, or exclude, the M428L / N434S variant in one, or preferably both, Fc domains. [Figure 18] Results of several anti-LAG-3 hybridoma screenings are shown. 1 μg of human LAG-3-hIg in 10 μL was mixed with 50 μL of hybridoma supernatant (2-fold diluted, 8 times in RPMI medium containing 10% FBS) for 20 minutes at room temperature. 40 μL of Daudi or Ramos cells (which endogenously express MHC-II) were added and incubated for 30 minutes at 4°C. The cells were then washed and incubated for 30 minutes with anti-human Fc-Alexa647 secondary antibody. The cells were then washed and analyzed for Alexa647 by FACS. [Figure 19A-B] Cytokine release assays (A: IL-2, B: IFNγ) after SEB stimulation of human PBMCs and treatment with anti-CTLA-4X anti-PD-1 bispecific antibodies are shown. [Figure 20A-C] CD45+ events and CD8+ events on day 14 are shown after human PBMCs were transplanted into NSG mice on day 0 and subsequently dosed with the test article shown on day 1. [Figure 21A-B] T cell binding in a SEB-stimulated PBMC assay by a chimeric antibody generated from an anti-TIM-3 hybridoma is shown. [Figure 22]Several anti-TIM-3 antigen-binding domain technical data from three experiments are shown. These include the XENP code for the bivalent embodiment, derivative clones, designation of the vh and vl manipulated domains, and the KD binding constant, association constant, and dissociation constant for human TIM-3 as measured by Octet. [Figure 23A-N] Several anti-PD-1 antigen-binding domain technical data are shown. These include the XENP codes for the bivalent and scFv embodiments, the designation of the vh and vl manipulated domains, the scFv orientation (N to C-terminus), the KD binding constant to human PD-1 as measured by Octet, and the Tm of the scFv. [Figure 24A-G] The results of several anti-CTLA-4 Fab screenings are shown. These include the XENP codes for the Fab and scFv embodiments, the designation of the vh and vl manipulated domains, the KD binding constants to human and cynoCTLA-4 as measured by Octet, and the Tm of scFv and Fab. In addition, the number of sequence 9-mers that exactly matched at least one human VH or VL germline is shown as a measure of the humanity of the variable regions of both Fab and scFv. [Figure 25] The mixed lymphocyte response shows enhanced IL-2 release with nivolumab (anti-PD-1 monoclonal antibody, marketed as Opdivo®), ipilimumab alone (anti-CTLA-4 monoclonal antibody, marketed as Yervoy®), a prototype anti-CTLA-4x anti-PD-1 bispecificity based on nivolumab and ipilimumab arms, and a "single-arm" combination control. [Figure 26] The study shows mixed lymphocyte responses that demonstrate enhanced IL-2 release with anti-CTLA-4x anti-PD-1 bispecific antibodies having mutant CTLA-4 Fab arms and mutant anti-PD-1 scFv arms, as well as with nivolumab alone, ipilimumab alone, and a prototype anti-CTLA-4x anti-PD-1 bispecific antibody based on nivolumab and ipilimumab arms as a control. [Figure 27]This study demonstrates that the anti-CTLA-4x anti-PD-1 bispecific compound enhances engraftment (measured by human CD45 counting) in human PBMC-transplanted NSG mice. The enhancement is greater than that observed with nivolumab (XENP16432) alone (dashed line). [Figure 28] This study demonstrates a correlation between body weight and CD45 cell count in graft-versus-host disease, suggesting that CD45 cell levels predict the disease. [Figure 29] Figure 27 shows the correlation between CD45 cell count and IFNγ release in the studies presented. [Figure 30] This study demonstrates that the anti-CTLA-4x anti-PD-1 bispecific compound enhances engraftment (measured by human CD45 counting) in human PBMC-transplanted NSG mice. The enhancement is greater than that observed with nivolumab (XENP16432) alone (dashed line). [Figure 31] Figure 30 shows the correlation between CD45 cell count and IFNγ release in the study shown. [Figure 32] Figures 27 and 30 show a comparison of study-to-study efficacy of the test item, demonstrating consistent superiority of the anti-PD-1x anti-CTLA-4 bispecific checkpoint antibody compared to nivolumab alone. [Figure 33A-B] The results of mixed lymphocyte reactions to evaluate anti-CTLA-4 x anti-PD-1, anti-LAG-3 x anti-PD-1, and anti-LAG-3 x anti-CTLA-4 bispecific compounds are shown. Analyte levels were standardized to levels induced by nivolumab monotherapy (values ​​greater than 1 represent enhancement to nivolumab). [Figure 34] The SEB reaction for evaluating the anti-LAG-3x anti-CTLA-4 bispecificity is shown. While the anti-LAG-3x anti-CTLA-4 bispecificity itself enhances the IL-2 response compared to the control, it is inferior to nivolumab alone. However, the anti-LAG-3x anti-CTLA-4 bispecificity combined with nivolumab yields a significantly higher IL-2 response than either drug alone. [Figure 35]The anti-CTLA-4 x anti-PD-1, anti-LAG-3 x anti-PD-1, anti-BTLA x anti-PD-1, and anti-LAG-3 x anti-CTLA-4 bispecific compounds enhance engraftment (measured by human CD45 counting) in human PBMC-transplanted NSG mice. This enhancement is greater than that seen with nivolumab (XENP16432) alone. Furthermore, the anti-LAG-3 x anti-CTLA-4 bispecific compounds, when combined with nivolumab, result in the highest engraftment levels. [Figure 36A-B] This study demonstrates that the anti-BTLAx anti-PD-1 bispecific compound requires disruption of the HVEM / BTLA interaction to exhibit de-repressive activity comparable to nivolumab. [Figure 37A-E]The following shows sequences of several useful bottle opener-type skeletons based on human IgG1 that do not contain the Fv sequence (e.g., scFv, as well as vh and vl on the Fab side). Bottle opener skeleton 1 is based on human IgG1 (356E / 358M allotype) and includes the N208D / Q295E / N384D / Q418E / N421D pI mutant on the Fab side and the S364K / E357Q:L368D / K370S asymmetric mutant of the E233P / L234V / L235A / G236del / S267K cleavage mutant on both strands. The bottle opener skeleton 2 is based on human IgG1 (356E / 358M allotype) and contains different asymmetric mutants: the N208D / Q295E / N384D / Q418E / N421D pI mutant on the Fab side and the E233P / L234V / L235A / G236del / S267K cleavage mutant on both strands. The bottle opener skeleton 3 is based on human IgG1 (356E / 358M allotype) and contains different asymmetric mutants: the N208D / Q295E / N384D / Q418E / N421D pI mutant on the Fab side and the E233P / L234V / L235A / G236del / S267K cleavage mutant on both strands. The bottle opener skeleton 4 is based on human IgG1 (356E / 358M allotype) and contains different asymmetric mutants of the N208D / Q295E / N384D / Q418E / N421D pI mutant on the Fab side and E233P / L234V / L235A / G236del / S267K cleavage mutant on both strands. The bottle opener skeleton 5 is based on human IgG1 (356D / 358L allotype) and contains the N208D / Q295E / N384D / Q418E / N421D pI mutant on the Fab side and S364K / E357Q:L368D / K370S asymmetric mutant of the E233P / L234V / L235A / G236del / S267K cleavage mutant on both strands. The bottle opener skeleton 6 is based on human IgG1 (356E / 358M allotype) and contains the N208D / Q295E / N384D / Q418E / N421D pI variant on the Fab side, the S364K / E357Q:L368D / K370S asymmetric variant of the E233P / L234V / L235A / G236del / S267K cleavage variant on both strands, and the N297A variant on both strands. The bottle opener skeleton 7 is identical to 6 except that the mutation is N297S.Alternative versions of the bottle opener skeletons 6 and 7 can exclude the bi-strand cleavage mutants E233P / L234V / L235A / G236del / S267K. The skeleton 8, based on human IgG4, includes the N208D / Q295E / N384D / Q418E / N421D pI mutant on the Fab side and the S364K / E357Q:L368D / K370S asymmetric mutant of the E233P / L234V / L235A / G236del / S267K cleavage mutant on both strands, as well as the S228P (EU numbered, which is S241P in Kabat) mutant on both strands (which cleaves the Fab arm exchange as known in the art). Alternative versions of the bottle opener skeleton 8 can exclude the E233P / L234V / L235A / G236del / S267K cleavage mutants on both strands. Skeleton 9, based on human IgG2, includes the S364K / E357Q:L368D / K370S asymmetric mutant, which is the N208D / Q295E / N384D / Q418E / N421D pI mutant, on the Fab side. Skeleton 10, based on human IgG2, includes the S364K / E357Q:L368D / K370S asymmetric mutant, which is the N208D / Q295E / N384D / Q418E / N421D pI mutant, on the Fab side, as well as the S267K mutant on both strands. As will be understood by those skilled in the art, and as outlined below, these sequences can be used in any vh and vl pairs outlined herein, with one monomer comprising scFv (optionally comprising a charged scFv linker) and the other monomer comprising a Fab sequence (e.g., vh is bound to the “Fab side heavy chain” and vl is bound to the “steady light chain”). That is, any Fv sequences outlined herein for anti-CTLA-4, anti-PD-1, anti-LAG-3, anti-TIM-3, anti-TIGIT, and anti-BTLA can be incorporated into these Figure 37 skeletons in any combination, whether as scFv (likewise having an optional charged scFv linker) or Fab. The steady light chains shown in Figure 37A can be used in all of the constructs in the figure, but kappa steady light chains can also be substituted.It should be noted that these bottle-pull skeletons are useful in the central-scFv type of Figure 1F when an additional second Fab (vh-CH1 and vl constant light) having the same antigen binding as the first Fab is added to the N-terminus of the scFv on the “bottle-pull side”. Sequences that are 90, 95, 98, and 99% identical (as defined herein) to the listed sequences and / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 further amino acid substitutions (as will be understood by those skilled in the art, compared to parent human IgG1 (or IgG2 or IgG4 depending on the skeleton, compared to the “parent” in the figure which already contains several amino acid modifications) are contained within each of these skeletons. That is, the listed skeletons may contain additional amino acid modifications (generally amino acid substitutions) in addition to the asymmetric, pI, and cleavage variants contained within the skeletons of this figure. [Figure 38A-D]The sequence of the mAb-scFv skeleton used in the present invention, to which the Fv sequence of the present invention is added, is shown. mAb-scFv skeleton 1 is based on human IgG1 (356E / 358M allotype) and includes the N208D / Q295E / N384D / Q418E / N421D pI mutant on the Fab side and the S364K / E357Q:L368D / K370S asymmetric mutant of the E233P / L234V / L235A / G236del / S267K cleavage mutant on both strands. Skeleton 2 is based on human IgG1 (356D / 358L allotype) and includes the N208D / Q295E / N384D / Q418E / N421D pI mutant on the Fab side and the S364K / E357Q:L368D / K370S asymmetric mutant of the E233P / L234V / L235A / G236del / S267K cleavage mutant on both strands. Skeleton 3, based on human IgG1 (356E / 358M allotype), includes the N208D / Q295E / N384D / Q418E / N421D pI variant on the Fab side and the S364K / E357Q:L368D / K370S asymmetric variant of the E233P / L234V / L235A / G236del / S267K cleavage variant on both strands, as well as the N297A variant on both strands. Skeleton 4 is identical to 3 except that the mutation is N297S. Alternative forms of mAb-scFv skeletons 3 and 4 can exclude the E233P / L234V / L235A / G236del / S267K cleavage variant on both strands. Skeleton 5 is based on human IgG4 and includes the N208D / Q295E / N384D / Q418E / N421D pI variant on the Fab side and the S364K / E357Q:L368D / K370S asymmetric variant of the E233P / L234V / L235A / G236del / S267K cleavage variant on both strands, as well as the S228P (EU numbered, which is S241P in Kabat) variant on both strands (cleaving the Fab arm exchange as known in the art). Skeleton 6 is based on human IgG2 and includes the S364K / E357Q:L368D / K370S asymmetric variant of the N208D / Q295E / N384D / Q418E / N421D pI variant on the Fab side.The skeleton 7, based on human IgG2, includes the S364K / E357Q:L368D / K370S asymmetric variant of the N208D / Q295E / N384D / Q418E / N421D pI variant on the Fab side, as well as the S267K variant on both strands. As will be understood by those skilled in the art, and as outlined below, these sequences can be used with any vh and vl pair outlined herein, with one monomer containing both Fab and scFv (optionally including a charged scFv linker) and the other monomer containing the Fab sequence (e.g., vh is bound to the “Fab-side heavy chain” and vl is bound to the “steady light chain”). That is, any Fv sequences outlined herein for anti-CTLA-4, anti-PD-1, anti-LAG-3, anti-TIM-3, anti-TIGIT, and anti-BTLA can be incorporated into the Figure 38 skeleton in any combination, whether as scFv (and optionally as a charged scFv linker) or Fab. The monomer 1 side is the Fab-scFv pI-negative side and includes the heterodimerized mutant L368D / K370S, the isodistributed pI mutant N208D / Q295E / N384D / Q418E / N421D, and the cleavage mutant E233P / L234V / L235A / G236del / S267K (all relative to IgG1). The monomer 2 side is the scFv pI-positive side and includes the heterodimerized mutant 364K / E357Q. However, other asymmetric mutant pairs, particularly [[S364K / E357Q:L368D / K370S], [L368D / K370S:S364K], [L368E / K370S:S364K], [T411T / E360E / Q362E:D401K], [L368D / K370S:S364K / E357L], [K370S:S364K / E357Q], [T366S / L368A / Y407V:T366W], and [T366S / L368A / Y407V / Y394C:T366W / S354C], may be substituted. The constant light chains shown in Figure 38A can be used in all of the constructs shown in the figure, but the kappa constant light chain may also be substituted.It should be noted that these mAb-scFv skeletons are useful for both the mAb-Fv type in Figure 1H (one monomer containing vl at the C-terminus and the other containing vh at the C-terminus) and the scFv-mAb type in Figure 1E (the scFv domain is added to the C-terminus of one of the monomers). Sequences that are 90, 95, 98, and 99% identical (as defined herein) to the listed sequences and / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 further amino acid substitutions (as will be understood by those skilled in the art, parent human IgG1 (or IgG2 or IgG4 depending on the skeleton, compared to the “parent” in the figure which already contains several amino acid modifications) are contained within each of these skeletons. That is, the listed skeletons may contain additional amino acid modifications (generally amino acid substitutions) in addition to the asymmetric, pI, and cleavage variants contained within the skeletons in this figure. [Figure 39A-B]A matrix of possible combinations of the bispecific checkpoint antibodies of the present invention is shown. In Figure 39A, the combinations are not type-constrained, and any type from Figure 1 can be used. "A" in the box means that the CDR from the first ABD (listed on the X-axis) can be combined with the CDR from the second ABD (listed on the Y-axis). "B" in the box means that the vh and vl chains from the first ABD can be combined with the vh and vl chains from the second ABD. "C" in the box means that the CDR from the first ABD can be combined with the vh and vl chains from the second ABD. "D" in the box means that the vh and vl chains from the first ABD can be combined with the CDR from the second ABD. The "E" in the box means that the PD-1 ABD is selected from the group 1G6_H1.279_L1.194, 1G6_H1.280_L1.224, 1G6_L1.194_H1.279, 1G6_L1.210_H1.288, and 2E9_H1L1.The "F" in the box indicates that CTLA-4 ABD is [CTLA-4]_H0.25_L0, [CTLA-4]_H0.26_L0;[CTLA-4]_H0.27_L0, [CTLA-4]_H0.29_L0, [CTLA-4]_H0.38_L0, [CTLA-4]_H0.39_L0, 0[CTLA-4]_H0.40_L0, [CTLA-4]_H0.70_L0, [CTLA-4]_H0_L0.22, [CTLA-4]_H2_L0, [CTLA-4]_H3.21_L0.124, [CTLA-4]_ H3.21_L0.129, [CTLA-4]_H3.21_L0.132, [CTLA-4]_H3.23_L0.124, [CTLA-4]_H3.23_L0.129, [CTLA-4]_H3.23_L0.132, [CTLA- 4]_H3.25_L0.124, [CTLA-4]_H3.25_L0.129, [CTLA-4]_H3.25_L0.132, [CTLA-4]_H3.4_L0.118, [CTLA-4]_H3.4_L0.119, [CTLA- 4]_H3.4_L0.12, [CTLA-4]_H3.4_L0.121, [CTLA-4]_H3.4_L0.122, [CTLA-4]_H3.4_L0.123, [CTLA-4]_H3.4_L0.124, [CTLA-4]_ H3.4_L0.125, [CTLA-4]_H3.4_L0.126, [CTLA-4]_H3.4_L0.127, [CTLA-4]_H3.4_L0.128, [CTLA-4]_H3.4_L0.129, [CTLA-4]_H3 This means that the selection is made from the group consisting of .4_L0.130, [CTLA-4]_H3.4_L0.131, [CTLA-4]_H3.4_L0.132, [CTLA-4]_H3.5_L2.1, [CTLA-4]_H3.5_L2.2, [CTLA-4]_H3.5_L2.3, [CTLA-4]_H3_L0, [CTLA-4]_H3_L0.22, [CTLA-4]_H3_L0.44, [CTLA-4]_H3_L0.67, and [CTLA-4]_H3_L0.74.The "G" in the box means that the TIM-3 ABD is selected from the group 1D10_H0L0, 1D12_H0L0, 3H3_H1_L2.1, 6C8_H0L0, 6D9_H0_1D12_L0;7A9_H0L0, 7B11_H0L0, 7B11var_H0L0, and 7C2_H0L0. The "H" in the box means that LAG-3 ABD has identifiers 2A11_H0L0, 2A11_H1.125_L2.113, 2A11_H1.144_L2.142, 2A11_H1_L2.122, 2A11_H1_L2.123, 2A11_H1_L2.124, 2 A11_H1_L2.25, 2A11_H1_L2.47, 2A11_H1_L2.50, 2A11_H1_L2.91, 2A11_H1_L2.93, 2A11_H1_L2.97, 2A11_H1L1, 2A11_H1L2 This means that the selection is made from the group consisting of 2A11_H2L2, 2A11_H3L1, 2A11_H3L2, 2A11_H4L1, 2A1_H4L2, 7G8_H0L0, 7G8_H1L1, 7G8_H3.18_L1.11, 7G8_H3.23_L1.11, 7G8_H3.28_L1, 7G8_H3.28_L1.11, 7G8_H3.28_L1.13, 7G8_H3.30_L1.34, 7G8_H3.30_L1.34, and 7G8_H3L1. The "I" in the box and the "J" in the box mean that the BTLA ABD is selected from 9C6_H0L0, 9C6_H1.1_L1, and 9C6_H1.11_L1. Figure 39B is identical to Figure 39A except that Figure 39B is specific to the bottle opener type. In B, when the first ABD binds to PD-1, the first ABD is an scFv monomer, while the other ABDs (CTLA-4, LAG-3, TIGIT, TIM-3, and BTLA) are Fab monomers. In B, when the first ABD binds to CTLA-4, it is an scFv monomer (except when combined with PD-1 and on the Fab side), while the other ABDs (CTLA-4, LAG-3, TIGIT, TIM-3, and BTLA) are Fab monomers. [Figure 40]A matrix of possible bottle opener-type combinations is shown. "Q" in the box means that the first ABD domain (likewise listed on the X axis) is scFv and the second ABD (again listed on the Y axis) is Fab-side. "R" in the box means that the first ABD is Fab-side and the second ABD is scFv. "S" in the box means that the first ABD is anti-PD-1 and scFv-side. "T" in the box means that the first ABD is anti-CTLA-4 and scFv-side. "U" in the box means that the first ABD is anti-TIM-3 and scFv-side. "V" in the box means that the first ABD is anti-LAG-3 and scFv-side. "W" in the box means that the first ABD is anti-TIGIT and scFv-side. The "X" in the box indicates that the first ABD is anti-BTLA and on the scFv side. In addition, each combination outlined in Figure 39 can use the CDR, scFv, and vh and vl combinations of Figure 38. Furthermore, a particular embodiment of the bottle opener skeleton in Figure 39 is the array in Figure 36. [Figure 41A-B] A schematic diagram illustrating the benefits that bispecific checkpoint antibodies can offer in combination therapy using two different antibodies or drugs is shown. [Figure 42] A similar schematic diagram is shown, illustrating that because tumor TILs co-express multiple checkpoints, divalent binding increases activity, enhancing antitumor activity and avoiding peripheral toxicity. [Figure 43] This document demonstrates that the bispecific checkpoint antibody of the present invention (e.g., anti-LAG-3x anti-CTLA-4) can be combined with other monospecific checkpoint antibodies (e.g., nivolumab, pembrolizumab). [Figure 44] This study demonstrates that PD-1 and CTLA-4 are co-expressed in various tumor types, including bladder cancer, breast cancer, colon cancer, prostate cancer, lung cancer, melanoma, and ovarian cancer. [Figure 45A-C]B) Comparison of IL-2 enhancement by anti-PD-1 bivalent and anti-CTLA-4x anti-PD-1, and C) and unilateral anti-PD-1 + unilateral anti-CTLA-4 and anti-CTLA-4x anti-PD-1 in SEB-stimulated PBMC assays, as well as C) control experiments without SEB stimulation. [Figure 46A-B] This demonstrates the blockade of PD-L1 and PD-L2 ligands of PD-1 by exemplary anti-CTLA-4x anti-PD-1 bispecific compounds compared to unilateral anti-PD-1 and unilateral anti-CTLA-4 antibodies. [Figure 47] This shows exemplary T cell binding in a SEB-stimulated PBMC assay using an anti-CTLA-4x anti-PD-1 bispecific antibody. [Figure 48] This study demonstrates that the anti-CTLA-4x anti-PD-1 bispecific compound enhances engraftment (measured by human CD45 counting) in human PBMC-transplanted NSG mice. The enhancement is greater than that observed with nivolumab (XENP16432) alone. [Figure 49] This study demonstrates that an anti-BTLAx anti-PD-1 bispecific candidate binds more strongly to T cells compared to a "one-arm" control in a SEB-stimulated PBMC assay. [Figure 50A-B] This study demonstrates that an anti-BTLAx anti-PD-1 chimeric bispecific agent promotes IL-2 secretion from SEB-stimulated PBMCs. PBMCs were stimulated with 10 ng / mL of SEB for 3 days with the indicated test substance. Cell supernatants were collected and assayed with MSD for the indicated analytes. A: 20 μg / mL test substance; B: 5 μg / mL test substance. [Figure 51A-B] This study demonstrates that an anti-BTLAx anti-PD-1 chimeric bispecific agent promotes IFNγ secretion from SEB-stimulated PBMCs. PBMCs were stimulated with 10 ng / mL of SEB for 3 days with the indicated test substance. Cell supernatants were collected and assayed with MSD for the indicated analytes. A: 20 μg / mL test substance; B: 5 μg / mL test substance. [Figure 52A-B]This study demonstrates that anti-BTLAx anti-PD-1 bispecific antibodies (chimeric and humanized / optimized anti-BTLA Fab arms) promote IL-2 secretion and IFN-γ from SEB-stimulated PBMCs. Both panels were used to stimulate PBMCs with 10 ng / mL SEB for 3 days, along with the indicated 20 μg / mL test material. Cell supernatants were collected after 72 hours and assayed for the indicated analytes. [Figure 53A-F] This study demonstrates the enhancement of CD45 cell counting and IFNγ secretion over time (days 10, 14, and 22) by an exemplary anti-BTLAx anti-PD-1 bispecific antibody in a GVHD study. [Figure 54] Technical data for a 9C6 anti-BTLA antigen-binding domain is shown. This includes the XENP code for the bivalent embodiment, the designation of the vh and vl manipulated domains, and the KD binding constant to human BTLA as measured by Octet. [Figure 55A-E] Technical data for a 2A11 anti-LAG-3 antigen-binding domain is shown. This includes the XENP code of the Fab embodiment, the designation of the vh and vl manipulated domains, the KD binding constant to human LAG-3 as measured by Octet, and the Tm of the Fab. [Figure 56A-K] Technical data for a 7G8 anti-LAG-3 antigen-binding domain is shown. This includes the XENP code of the Fab embodiment, the designation of the vh and vl manipulated domains, the KD binding constant to human LAG-3 as measured by Octet, and the Tm of the Fab. [Figure 57A-B] The Kd value is measured by Octet and is based on either the optimized 2A11 or 7G8 anti-LAG-3 Fab arm, representing an anti-LAG-3X anti-CTLA-4 bispecific heterodimer plug-out type. [Figure 58]This study demonstrates that the anti-LAG-3(7G8)×anti-CTLA-4 and anti-LAG-3(2A11)×anti-CTLA-4 bispecific compounds bind more strongly than a single-arm anti-LAG-3 control. PBMCs were stimulated with 100 ng / mL SEB for 3 days. The cells were then treated with the indicated test material for 30 minutes at 4°C and washed twice. The cells were then treated with anti-CD3-FITC and anti-human Fc-APC antibodies. The cells were then washed twice and analyzed by flow cytometry. [Figure 59A-B] This study demonstrates that the 7G8-based anti-LAG-3x anti-CTLA-4 bispecificity exhibits more selective function in PBMCs than the 2A11-based anti-LAG-3x anti-CTLA-4 bispecificity, as indicated by enhancement in IL-2 and IFNγ release. PBMCs were stimulated with 500 ng / mL SEB for 2 days. The cells were then washed twice with culture medium and stimulated with 500 ng / mL SEB in combination with the indicated amounts of the test substance. Cells were assayed for the indicated analytes (either IL-2 or IFN-γ) 24 hours after treatment. Each dot represents a unique donor tested with a technical singlet assay. [Figure 60A-B] This study shows mixed lymphocyte reactions (MLR) with anti-LAG-3X anti-CTLA-4 bispecific antibodies. Forty unique MLR reactions were performed in the presence of the indicated test material at 20 ug / mL. The cell supernatant was then assayed by MSD for A: IL-2 and B: IFNγ 6 days after treatment. [Figure 61A-B] This study demonstrates enhancement of IL-2 and IFNγ release by additional anti-LAG-3X anti-CTLA-4 candidates in the SEB assay. PBMCs were stimulated with 500 ng / mL SEB for 2 days. The cells were then washed twice with culture medium and stimulated with 500 ng / mL SEB in combination with the indicated amounts of test material. Cells were assayed for the indicated analytes (either IL-2 or IFN-γ) 24 hours after treatment. Each point represents a unique donor tested with a technical singlet assay. [Figure 62A-B] The Kd value of the anti-LAG-3X anti-PD-1 bispecific heterodimer plug-out type is measured by Octet, based on either the optimized 2A11 or 7G8 anti-LAG-3 Fab arm. [Figure 63A-B] This demonstrates the ability of humanized / optimized 7G8 and 2A11 anti-LAG-3 clones to block LAG-3 binding to endogenously MHC-II expressing Daudi cells. [Figure 64A-B] This study demonstrates the anti-LAG-3x anti-PD-1 candidate function of SEB-stimulated T cells. PBMCs were stimulated with 500 ng / mL SEB for 2 days. The cells were then washed twice with culture medium and stimulated with 500 ng / mL SEB in combination with the indicated amounts of the test substance. Cells were assayed for the indicated analytes 24 hours after treatment. Each dot represents a unique donor tested with a technical singlet assay. [Figure 65] This graph shows that tumor-infiltrating lymphocytes (TILs) co-express multiple checkpoint receptors in various tumors. Specifically, the graph shows that various tumors co-express PD-1 and CTLA-4, PD-1 and BTLA, PD-1 and LAG-3, and LAG-3 and CTLA-4. The results shown are based on data generated by the TCGA Research network: http: / / cancergenome.nih.gov / . [Figure 66] This specification demonstrates that the bispecific antibodies provided here selectively target bispecific checkpoint-positive T cells. The bispecific PD-1xLAG-3 antibody is used to show PD-1 and LAG-3 receptor occupancy in staphylococcal enterotoxin B (SEB) stimulated CD3+ T cells compared to a negative control. [Figure 67A-F]This graph shows that the constituent antibody domains of the target antibodies provided herein are capable of blocking checkpoint receptor / ligand interactions. In particular, a bispecific antibody containing the 1G6 anti-PD-1 scFv arm is capable of blocking PD-1 / PD-L1 and PD-1 / PD-L2 interactions, a 7G8 anti-LAG-3 single arm is capable of blocking LAG-3 / MHC II interactions, a bispecific antibody containing an exemplary anti-PD-1 Fab arm is capable of blocking CTLA-4 / CD80 and CTLA-4 / CD86 interactions, and a bispecific antibody containing the 9C6 anti-BTLA Fab arm is capable of blocking BTLA / HVEM interactions. [Figure 68] We compare the enhancement of IL-2 release by exemplary anti-CTLA-4x anti-PD-1 bispecific antibodies and nivolumab. [Figure 69] We compare the enhancement of IL-2 release by an exemplary anti-LAG-3x anti-CTLA-4 bispecific antibody, the same bispecific antibody in combination with nivolumab, and nivolumab alone. [Figure 70] We compare the enhancement of IL-2 release by exemplary anti-LAG-3x anti-PD-1 bispecific antibodies and nivolumab. [Figure 71] We compare the enhancement of IL-2 release by exemplary anti-BTLAx anti-PD-1 bispecific antibodies and nivolumab. [Figure 72] We compare the enhancement of GVHD (indicated by CD45 cell counting) by exemplary anti-PD-1x anti-CTLA-4 bispecific antibodies, nivolumab alone, and nivolumab in combination with ipilimumab. [Figure 73] We compare the enhancement of GVHD (indicated by CD45 cell counting) by exemplary anti-BTLAx anti-PD-1 bispecific antibodies and nivolumab. [Figure 74] We compare the enhancement of GVHD (indicated by CD45 cell counting) by exemplary anti-LAG-3x anti-CTLA-4 bispecific antibodies, the same bispecific antibodies combined with nivolumab, and nivolumab alone. [Figure 75]We compare the enhancement of GVHD (as indicated by CD45 cell counting) by exemplary anti-LAG-3x anti-PD-1 bispecific antibodies and nivolumab. [Figure 76A-B] Two studies demonstrate that anti-CTLA-4x anti-PD-1 bispecific antibodies may enhance in vivo T cell-mediated antitumor efficacy. KG1a-luc cancer cells were transplanted into mice. After 21 days, huPMCs were transplanted into the same mice, and antibody treatment (anti-CTLA-4x anti-PD-1 bispecific antibody, anti-PD-1 bivalent antibody, or anti-PD-1 bivalent antibody + anti-CTLA-4 bivalent antibody) was administered weekly. IVIS cancer cell imaging was performed on the mice, and tumor size was assessed based on changes in tumor bundles. [Modes for carrying out the invention]

[0009] A. Incorporating data 1. Figures and Legend All figures and accompanying legends in U.S. Patent Applications No. 62,350,145, No. 62 / 353,511, and No. 62 / 420,500 are incorporated herein by reference in their entirety, expressly and independently, with respect to the amino acid sequences shown herein.

[0010] 2. Array Regarding the attached sequence listing, please note the following: Suitable anti-PD-1 sequences for use as ABD include sequence numbers 6209-11464 (Fv sequences of PD-1sc, although the Fv sequences within them may be in Fab form), sequence numbers 11465-17134 (Fab sequences of PD-1, although the Fv sequences within them may be in scFv form), sequence numbers 33003-33072 (additional Fab sequences of PD-1, although the Fv sequences within them may be in scFv form), sequence numbers 33073-35394 (additional scFv sequences of PD-1, although the Fv sequences within them may be in Fab form), and sequence numbers 36127-36146 (bivalent constructs of PD-1, which may be in either scFv or Fab form). Suitable anti-CTLA-4 sequences for use as ABD include sequence numbers 21-2918 (which are CTLA-4 scFv sequences, but whose Fv sequences may be in Fab format), sequence numbers 2919-6208 (which are CTLA-4 Fab sequences, but whose Fv sequences may be in scFv format), sequence numbers 36739-36818 (which are additional CTLA-4 Fab sequences, but whose Fv sequences may be in scFv format), and sequence numbers 35395-35416 (CTLA-4 uniarm constructs, which may be in either Fab or scFv format).Suitable anti-LAG-3 sequences for use as ABD include SEQ ID NOs: 17135-20764 (LAG-3 Fabs, but the Fv sequences within them may be in scFv format), 36819-36962 (additional LAG-3 Fabs, but the Fv sequences within them may be in scFv format), 35417-35606 (additional LAG-3 Fabs, but the Fv sequences within them may be in scFv format), 25194-32793 (additional LAG-3 Fabs, but the Fv sequences within them may be in scFv format), and 32794-33002 (one-arm LAG-3 constructs, which may be in either Fab or scFv format). Suitable anti-TIM-3 sequences for use as ABD include sequence numbers 20765-20884 (TIM-3 Fabs, but the Fv sequences within them may be in scFv format), sequence numbers 37587-37698 (additional TIM-3 Fabs, the Fv sequences may be in scFv format), and sequence numbers 36347-36706 (bivalent TIM-3 constructs, these may be in either Fab or scFv format). Suitable anti-BTLA sequences for use as ABD include sequence numbers 20885-21503 (BTLA Fabs, but the Fv sequences within them may be in scFv format), and sequence numbers 36707-36738 (additional BTLA Fabs, but the Fv sequences within them may be in scFv format). Suitable anti-TIGIT sequences for use as ABD include sequence numbers 21504-21523 (which are TIGIT Fabs, but whose Fv sequences may be in scFv format) and sequence numbers 37435-37586 (which are additional TIGIT Fabs, but whose Fv sequences may be in scFv format).

[0011] The bispecific antibodies of the present invention include constructs of LAG3 and CTLA4, numbered SEQ ID NOs. 35607-35866 and 21524-22620. The constructs of PD-1 and CTLA4 include SEQ ID NOs. 36167-36346 and 23316-2373. This includes items listed as 5. Constructs of PD-1 and TIM3 include items listed as sequence numbers 25174-25193. Constructs of PD-1 and LAG3 include items listed as sequence numbers 35867-36126 and 23736-25133. Constructs of PD-1 and TIGIT include items listed as sequence numbers 25134-25173. Constructs of PD-1 and BTLA include items listed as sequence numbers 22724-23315 and 36147-36166. Constructs of CTLA4 and BTLA include items listed as sequence numbers 22624-22723. Finally, the names XENP23552, XENP22841, XENP22842, XENP22843, XENP22844, XENP22845, XENP22846, XENP22847, XENP22848, XENP22849, XENP22850, XENP22851, XENP22852, XENP22858, XENP22854, and XENP22855 should all have included the inadvertently omitted "M428L / N434S" designation in their titles.

[0012] B. Overview Therapeutic antibodies targeting immune checkpoint inhibitors such as PD-1 hold great promise in the limited context of cancer treatment. Cancer can be perceived as a condition where the patient is unable to recognize and eliminate cancerous cells. In many cases, these transformed (e.g., cancerous) cells weaken immune surveillance. To prevent uncontrolled T-cell activity, there are innate regulatory mechanisms in the body that limit T-cell activation, which can be exploited by cancerous cells to evade or suppress the immune response. The goal of immunotherapy is to restore the ability of immune effector cells, particularly T cells, to recognize and eliminate cancer. The field of immuno-oncology, sometimes referred to as "immunotherapy," is rapidly developing with the recent approval of several T-cell checkpoint inhibitor antibodies such as Yervoy, Keytruda, and Opdivo. These antibodies are generally referred to as "checkpoint inhibitors" because they block negative regulators of T-cell immunity. It is generally understood that a variety of immunomodulatory signals, both co-stimulatory and co-inhibitory, can be used to modulate an optimal antigen-specific immune response.

[0013] Generally, these monoclonal antibodies bind to checkpoint inhibitor proteins such as CTLA-4 and PD-1, thereby preventing or suppressing the activation of cytotoxic T cells (CTLs) under normal circumstances. For example, an increase in the T cell response to tumors can be achieved by inhibiting checkpoint proteins using antibodies that bind to these proteins. That is, these cancer checkpoint proteins suppress the immune response, and when these proteins are blocked, for example, by using antibodies against them, the immune system is activated, leading to the treatment of conditions such as cancer and infectious diseases.

[0014] However, as mentioned above, studies have shown that TILs generally express multiple checkpoint receptors, which may suggest that blocking a single checkpoint is insufficient to promote a complete T cell response. Furthermore, TILs expressing multiple checkpoints are likely to be the most tumor-responsive, suggesting that therapies blocking two or more checkpoint antigens may be highly effective.

[0015] Therefore, the present invention provides a bispecific checkpoint antibody that binds to cells expressing two antigens, and a method for treating diseases such as cancer and infectious diseases, or other pathological conditions that are treated by enhancing immune activity, by activating T cells and / or NK cells.

[0016] Therefore, the present invention aims to solve toxicity problems and the costs of administering multiple antibodies by providing, in some cases, bispecific antibodies that bind to two different checkpoint inhibitor molecules on a single cell, thereby advantageously requiring the administration of only one therapeutic agent.

[0017] Bispecific antibodies, capable of simultaneously binding to two different targets, offer the potential to improve TIL-versus-peripheral T cell target selectivity while simultaneously reducing therapeutic costs. The bivalent interaction of an antibody with two targets on the cell surface should, in some cases, result in a higher binding affinity compared to a monovalent interaction with a single target at a time. For this reason, conventional bivalent antibodies tend to have high binding affinity to their targets on the cell surface. Bispecific antibodies, by leveraging the higher binding affinity obtained through simultaneous binding to both targets, may create greater selectivity for cells expressing two different targets simultaneously.

[0018] Accordingly, the present invention relates to novel constructs for providing heterodimer antibodies that enable binding to two or more checkpoint antigens or ligands, for example, to enable bispecific binding. For example, an anti-PD1xanti-CTLA4 (PD1×CTLA4) bispecific antibody is expected to be more selective of PD1+CTLA4+ double-positive TILs versus single-positive PD1-only or CTLA4-only T cells. Thus, selective blockade of double-positive TILs versus single-positive T cells is expected to improve the therapeutic index of combined checkpoint blockade. This is similarly true for other possible combinations, as outlined herein. Accordingly, the preferred bispecific antibodies of the present invention bind to PD-1 and CTLA-4, PD-1 and TIM-3, PD-1 and LAG-3, PD-1 and TIGIT, PD-1 and BTLA, CTLA-4 and TIM-3, CTLA-4 and LAG-3, CTLA-4 and TIGIT, CTLA-4 and BTLA, TIM-3 and LAG-3, TIM-3 and TIGIT, TIM-3 and BTLA, LAG-3 and TIGIT, LAG-3 and BTLA, and TIGIT and BTLA. It should be noted that, generally, these bispecific antibodies are named "anti-PD-1X anti-CTLA-4" for each pair, or generally simplified or for ease of use (and therefore interchangeable), "PD-1XCTLA-4," etc.

[0019] The heterodimeric bispecific checkpoint antibodies of the present invention are useful for treating various types of cancer. As those skilled in the art will understand, in contrast to conventional monoclonal antibodies that bind to tumor antigens, or a newer class of bispecific antibodies that bind to, for example, CD3 and tumor antigens (such as those described in USSN 15 / 141,350), checkpoint antibodies are used to enhance the immune response, but are generally not tumor-specific in their action. That is, the bispecific checkpoint antibodies of the present invention inhibit the suppression of the immune system, which generally leads to T cell activation, which leads to a greater immune response against cancer cells, and therefore to treatment. Thus, such antibodies can be expected to find utility for the treatment of a wide variety of tumor types. For example, the FDA recently approved Keytruda®, an anti-PD-1 monospecific antibody based on genetic characteristics rather than tumor type.

[0020] As discussed below, there are various methods for measuring T cell activation. The functional effects of bispecific checkpoint antibodies against NK and T cells can be evaluated in vitro (and, in some cases, in vivo, as fully described below) by measuring changes in the following parameters: proliferation, cytokine release, and cell surface markers. For NK cells, increased cell proliferation, cytotoxicity (the ability to kill target cells, measured by increased CD107a, granzyme, and perforin expression, or by directly measuring target cell killing), cytokine production (e.g., IFN-γ and TNF), and cell surface receptor expression (e.g., CD25) are used for immunomodulation, e.g., cancer cells. These serve as indicators of enhanced cell killing. For T cells, increased proliferation, increased expression of activation cell surface markers (e.g., CD25, CD69, CD137, and PD1), cytotoxicity (ability to kill target cells), and cytokine production (e.g., IL-2, IL-4, IL-6, IFN-γ, TNF-α, IL-10, IL-17A) are indicators of immunomodulation, such as enhanced killing of cancer cells. Therefore, the evaluation of treatment can be performed using assays that assess one or more of the following: (i) increased immune response, (ii) increased activation of αβ and / or γδ T cells, (iii) increased cytotoxic T cell activity, (iv) increased NK and / or NKT cell activity, (v) reduced suppression of αβ and / or γδ T cells, (vi) increased secretion of pro-inflammatory cytokines, (vii) increased secretion of IL-2, (viii) increased interferon-γ production, (ix) increased Th1 response, (x) decreased Th2 response, (xi) decreased number and / or activity of regulatory T cells and at least one other cell type, (xii) increased tumor immune infiltrates.

[0021] Accordingly, in some embodiments, the present invention provides the use of bispecific checkpoint antibodies to perform one or more of the following in subjects requiring: (a) upregulation of pro-inflammatory cytokines, (b) increased T cell proliferation, expansion, or tumor infiltration, (c) increased T cell production of interferon-γ, TNF-α, and other cytokines, (d) increased IL-2 secretion, (e) stimulation of antibody response, (f) inhibition of cancer cell growth, (g) promotion of antigen-specific T cell immunity, (h) promotion of CD4+ and / or CD8+ T cell activation, (i) reduction of T cell suppression, (j) promotion of NK cell activation, (k) promotion of apoptosis or lysis of cancer cells, and / or (l) cytotoxic or cell proliferation inhibitory effects on cancer cells.

[0022] Accordingly, the present invention provides bispecific heterodimer checkpoint antibodies. Heterodimer antibody constructs are based on the self-assembly properties of two "monomers" that organize into, for example, a "dimer" in the two Fc domains of the antibody's heavy chain. Heterodimer antibodies are produced by altering the amino acid sequence of each monomer, which will be discussed in full below. Accordingly, the present invention generally aims to create heterodimer antibodies that can co-capture checkpoint antigens in several ways, by relying on amino acid variants in the constant region that differ in each chain, thereby promoting heterodimer formation and / or facilitating heterodimer purification with respect to homodimers.

[0023] Therefore, the present invention provides a bispecific checkpoint antibody. A continuing problem in antibody technology is the desire for "bispecific" antibodies that bind simultaneously to two (or more) different antigens, generally allowing different antigens to be brought into proximity, resulting in novel functionalities and novel therapies. Generally, these antibodies are made by incorporating the genes for each heavy chain and light chain into a host cell (generally, in the present invention, the genes for the two heavy chain monomers and light chains outlined herein). This generally results in the formation of a desired heterodimer (AB), as well as two homodimers (AA and BB). However, the main obstacle in bispecific antibody formation is the difficulty in purifying heterodimeric antibodies to remove homodimeric antibodies and / or biasing them towards heterodimer formation to outweigh homodimer formation.

[0024] To solve this problem, several mechanisms exist that can be used to generate the heterodimers of the present invention. In addition, as will be understood by those skilled in the art, these mechanisms can be combined to ensure high heterodimerization. Thus, amino acid variants that lead to the production of heterodimeric antibodies are called "heterodimerizing variants." As will be discussed below, heterodimerizing variants include stereovariates (e.g., the "knob-and-hole" or "asymmetric" variants described below, and the "charge pair" variants described below), as well as "pI variants" that allow for the removal of heterodimers and the purification of homodimers. It may include.

[0025] One mechanism, sometimes referred to in the art as “knob-and-hole” (“KIH”) or herein as “asymmetric” variant, may be optionally used, which refers to amino acid manipulations that create steric and / or electrostatic effects that favor heterodimerization and discourage homodimerization, as described in Ridgway et al., Protein Engineering 9(7):617(1996), Atwell et al., J.Mol.Biol.1997 270:26, U.S. Patent No. 8,216,805, and US2012 / 0149876 (all of which are incorporated herein by reference in their entirety). In the figure, several “monomer A-monomer B” pairs containing “knob-and-hole” amino acid substitutions are identified. In addition, as described in Merchant et al., Nature Biotech. 16:677 (1998), these "knob-and-hole" mutations, in combination with disulfide bonds, can make the formation asymmetric to heterodimerization. Those used in the present invention are, in particular, T366S / L368A / Y407V paired with T366W, and T366S / L368A / Y407V / Y349C paired with this mutant with a cross-linking disulfide, T366W / S354C, in combination with other heterodimerizing mutants, including the pI mutant outlined below.

[0026] An additional mechanism useful in the generation of heterodimer antibodies is sometimes called “electrostatic manipulation” or “charge pairing,” as described in Gunasekaran et al., J. Biol. Chem. 285(25):19637(2010) (the whole is incorporated herein by reference). This is sometimes referred to herein as “charge pairing.” In this embodiment, electrostatic manipulation is used to make the formation asymmetric towards heterodimerization. Those skilled in the art will understand that these may also have an effect on pI, and thus may also have an effect on purification, and therefore may, in some cases, be considered pI variants. However, since these are generated to promote heterodimerization and are not used as a means of purification, they are classified as “stereovariants.” These include, but are not limited to, D221E / P228E / L368E paired with D221R / P228R / K409R (for example, these are "the set of monomers that correspond") and C220E / P228E / 368E paired with C220R / E224R / P228R / K409R, the others shown in the figure.

[0027] In some embodiments of the present invention, pI variants are used to alter one or both of the pIs of the monomer, thereby enabling isoelectric focusing of AA dimer proteins, AB dimer proteins, and BB dimer proteins.

[0028] In the present invention, several fundamental mechanisms exist that can facilitate the purification of heterodimeric proteins. Specifically, they rely on the use of pI variants, resulting in monomers each possessing a different pI, thereby enabling isoelectric focus purification of AA, AB, and BB dimer proteins. Alternatively, some scaffold types, such as the "triple F" type, also allow for size-based separation. As will be further outlined below, "asymmetry" is also possible, where heterodimerization outweighs homodimerization. Therefore, combinations of stereodimerizing variants and pI variants or charge-pair variants are particularly useful in the present invention. In addition, as will be further outlined below, scaffolds utilizing scFv(s), such as the triple F type, may include charged scFv linkers (either positive or negative) to provide further pI increases for purification purposes. As those skilled in the art will understand, some triple-F types have only a charged scFv linker and are useful without additional pI adjustment; however, the present invention also provides the use of asymmetric mutants having a charged scFv linker (and combinations of Fc, FcRn, and KO discussed herein). .

[0029] In the present invention, which utilizes pI as a separation mechanism to enable the purification of heterodimeric proteins, amino acid variants may be introduced into one or both of the monomeric polypeptides. That is, the pI of one of the monomers (hereinafter simply referred to as “monomer A”) can be manipulated to remove monomer B, or both monomers A and B can be charged to increase the pI of monomer A and decrease the pI of monomer B. As will be outlined more fully below, changes in the pI of either or both monomers can be made by removing or adding a charged residue (e.g., a neutral amino acid is replaced by a positively or negatively charged amino acid residue, e.g., glycine to glutamic acid), changing a charged residue from positive or negative to the opposite charge (e.g., aspartic acid to lysine), or changing a charged residue to a neutral residue (e.g., lysine to serine, thereby eliminating the charge). Several of these variants are shown in the figure. In addition, suitable pI variants for use in the creation of heterodimer antibodies as described herein are isotypes, for example, those obtained by transferring pI variants from different IgG isotypes to alter pI without introducing significant immunogenicity; see Figure 29 of U.S. Publication No. 2014 / 0288275 (the entire Figure is incorporated herein by reference).

[0030] Therefore, in this embodiment of the present invention, a sufficient change in pI is provided in at least one of the monomers, and as a result, the heterodimer can be separated from the homodimer. As will be understood by those skilled in the art, and as will be discussed further below, this is the “wild-type” heavy chain constant region and its pI increasing or decreasing (wt A-+B or wt This can be done by using a mutant region that has been manipulated to be either A--B, or by increasing one region and decreasing the other (A+-B- or A-B+).

[0031] Therefore, generally, components of some embodiments of the present invention are amino acid variants in the constant region of an antibody, which are directed to alter the isoelectric point (pI) of both, or if not both, at least one, monomer of a dimeric protein in order to form a “pI dimer” (when the protein is an antibody, these are called “pI antibodies”) by incorporating an amino acid substitution (“pI variant” or “pI substitution”) into one or both of the monomers. As shown herein, the separation of heterodimers from two homodimers can be achieved if the pI of the two monomers differs by as little as 0.1 pH units, and differences of 0.2, 0.3, 0.4, and 0.5 or greater are all useful in the present invention.

[0032] As those skilled in the art will understand, the number of pI variants present in each or both monomers to obtain good separation depends in part on the starting pIs of the scFv and Fab of interest. That is, the Fv sequences of the two target antigens are calculated and used to determine which monomer to manipulate or in which "direction" (e.g., more positive or more negative). As is known in the art, different Fvs have different starting pIs that are utilized in this invention. Generally, as outlined herein, the manipulation of the pI results in a total pI difference of at least about 0.1 log in each monomer, and as outlined herein, such total pI difference is preferably 0.2 to 0.5.

[0033] Furthermore, as will be understood by those skilled in the art and outlined herein, in some cases (depending on the type), heterodimers can be separated from homodimers based on size (e.g., molecular weight). For example, as shown in some embodiments of Figure 1, different sizes can be separated depending on the type. This yields homodimers and heterodimers of the antibody (for example, for the bottle opener, one homodimer is of the "double scFv" type, one homodimer is the standard antibody, and the heterodimer has one Fab and one scFv).

[0034] In addition, as shown in Figure 1, it recognizes that some antigens can be bound in a bivalent state (e.g., two antigen-binding sites for a single antigen). As can be understood, any combination of Fab and scFv can be used to achieve the desired results and combinations.

[0035] When pI variants are used to achieve heterodimerization beyond homodimerization by utilizing the constant region(s) of the heavy chain(s), a more modular approach is provided for the design and purification of multispecific proteins, including antibodies. Thus, in some embodiments, the heterodimerization variants (including asymmetric heterodimerization variants and purified heterodimerization variants) are not included in the variable region, and as a result, each antibody must be manipulated individually. In addition, in some embodiments, the potential immunogenicity arising from pI variants is significantly reduced by introducing pI variants derived from different IgG isotypes, thereby altering pI without introducing significant immunogenicity. Therefore, an additional problem to be solved is the elucidation of low pI constant domains with a high human sequence content, such as minimizing or avoiding non-human residues at certain specific locations.

[0036] The additional benefits that can arise from this pI manipulation include an extension of serum half-life and increased FcRn binding. That is, as described in USSN13 / 194,904 (which is incorporated in its entirety by reference), lowering the pI of the antibody constant domain (including those found in the antibody and Fc fusion) can lead to longer retention in serum in vivo. These pI variants, which increase serum half-life, also facilitate pI changes for purification.

[0037] Furthermore, it should be noted that the pI variants of heterodimerized mutants offer additional advantages to the analytical and quality control processes of bispecific antibodies, as they exhibit significant abilities to exclude, minimize, or distinguish homodimers in the presence of homodimers. Similarly, the ability to reliably test the reproducibility of heterodimer protein production is important.

[0038] As will be understood by those skilled in the art and will be fully discussed below, the heterodimer fusion proteins of the present invention can generally take on a wide variety of configurations, as shown in Figure 1. Some figures show a "single-ended" configuration in which one specificity is located on one "arm" of the molecule and a different specificity is located on the other "arm". Other figures show a "dual-ended" configuration in which at least one specificity is located on the "upper" part of the molecule and one or more different specificities are located on the "lower" part of the molecule. Accordingly, the present invention relates to novel immunoglobulin compositions that co-capture a first antigen and a second antigen. The first and second antigens of the present invention are referred herein to as antigen-1 and antigen-2 (or "checkpoint-1" and "checkpoint-2"), respectively.

[0039] One heterodimer scaffold particularly useful in the present invention is the “triple F” or “bottle opener” scaffold type shown in Figure 1A. In this embodiment, one heavy chain of the antibody comprises a single Fv (defined hereafter as “scFv”), and the other heavy chain is a “regular” FAb type, comprising a variable heavy chain and a light chain. This structure is sometimes referred to herein as the “triple F” type (scFv-FAb-Fc) or the “bottle opener” type because it roughly resembles a bottle opener visually (see Figure 1A). The two chains form a heterodimer antibody. In the constant regions that promote this (e.g., the Fc domain and / or hinge region), amino acid variants are used to bring them together, which are described more fully below.

[0040] This "triple F" type has several distinctly different advantages. As is known in the art, antibody analogs that rely on two scFv constructs often have stability and aggregation problems, but in this invention, these problems can be mitigated by adding "typical" heavy and light chain pairs. In addition, in contrast to types that rely on two heavy chains and two light chains, there is no problem of the heavy and light chains being incorrectly paired (for example, heavy 1 being paired with light 2, etc.).

[0041] Furthermore, as outlined herein, additional amino acid variants can be introduced into the bispecific antibodies of the present invention to add additional functionality. For example, amino acid changes within the Fc region can be added (to either one monomer or both monomers) to facilitate the increase of ADCC or CDC (e.g., altering binding to the Fcγ receptor), and to increase binding to FcRn and / or increase the serum half-life of the resulting molecule. As further described herein and understood by those skilled in the art, all variants outlined herein can be optionally and independently combined with other variants.

[0042] Similarly, another category of functional variants is "Fcγ cleavage variants" or "Fc knockout (FcKO or KO) variants." In these embodiments, for several therapeutic applications, it is desirable to reduce or eliminate the normal binding of the Fc domain to one or more or all of the Fcγ receptors (e.g., FcγR1, FcγRIIa, FcγRIIb, FcγRIIIa, etc.) to avoid additional mechanisms of action. That is, for example, it is generally desirable to cleave FcγRIIIa binding to eliminate or significantly reduce ADCC activity. Preferred cleavage variants are shown in Figure 5.

[0043] C. Nomenclature The bispecific antibodies of the present invention are listed in several different types. Each polypeptide is given a unique "XENP" number, although, as understood in the art, longer sequences may have shorter numbers. For example, the heavy chain of the bottle-opener-type scFv monomer of a given sequence has a first XENP number, while the scFv domain has a different XENP number. Some molecules have three polypeptides, and therefore the XENP numbers are used as names along with the components. Thus, the bottle-opener-type molecule XENP20717 contains three sequences, or equivalents, commonly referred to as "XENP20717 HC-Fab," "XENP20717HC-scFv," and "XENP20717LC," which will be readily identifiable by sequence alignment to those skilled in the art. These XENP numbers and identifiers are found in sequence listings and are used in the figures. In addition, a single molecule containing three components may result in multiple sequence identifiers. For example, the Fab monomer list has three CDRs: a full-length sequence, a variable-weight sequence, and a variable-weight sequence; the light chain has three CDRs: a full-length sequence, a variable-light sequence, and a variable-light sequence; and the scFv-Fc domain has a full-length sequence, an scFv sequence, a variable-light sequence, three light CDRs, an scFv linker, a variable-weight sequence, and three heavy CDRs. Note that all molecules herein having an scFv domain use a single charged scFv linker (+H), although others may be used. In addition, the nomenclature for specific variable domains uses the format of the kind "Hx.xx_Ly.yy", where the number is a unique identifier for a particular variable chain sequence. Thus, the Fab-side variable domain of XENP22841 is "7G8_H3.30_L1.34", which indicates that the variable-weight domain H3.30 is combined with the light domain L1.34. When these sequences are used as scFvs, the designation "7G8_H3.30_L1.34" means that the variable weight domain H3.30 is combined with the light domain L1.34, and N This indicates that the C-terminal vh-linker-vl orientation is present. This molecule, which has the same sequence of heavy and light variable domains but in reverse order, would be named "7G8_L1.34_H3.30". Similarly, different constructs can "combine" the heavy and light chains, as is evident from the sequence listing and figures.

[0044] D. Definition To help you fully understand this application, some definitions are provided below. Such definitions are intended to encompass grammatically equivalents.

[0045] In this specification, “cleavage” means reduction or removal of activity. Therefore, for example, “cleavage of the FcγR bond” means that the Fc region amino acid variant has less than 50% of the starting bond compared to the Fc region without the specific variant, preferably with a loss of 70-80-90-95-98% of activity, and generally the activity is below a detectable binding level in Biacore, SPR, or BLI assays. Those specifically used for FcγR bond cleavage are shown in Figure 5, and these are generally added to both monomers.

[0046] As used herein, "ADCC" or "antibody-dependent cytotoxicity" refers to a cell-mediated response in which nonspecific cytotoxic cells expressing FcγR recognize an antibody bound to a target cell, subsequently causing lysis of the target cell. ADCC is correlated with binding to FcγRIIIa, and increased binding to FcγRIIIa leads to increased ADCC activity.

[0047] As used herein, “ADCP” or “antibody-dependent cell phagocytosis” means a cell-mediated response in which nonspecific phagocytic cells expressing FcγR recognize antibodies bound to target cells, and subsequently cause phagocytosis of the target cells.

[0048] In this specification, “antigen-binding domain” or “ABD” means a set of six complementarity-determining regions (CDRs) that, when present as part of a polypeptide sequence, specifically bind to a target antigen as discussed herein. Thus, the “checkpoint antigen-binding domain” binds to the target checkpoint antigen outlined herein. As is known in the art, these CDRs generally exist as a first set of variable heavy CDRs (vhCDR or VHCDR) and a second set of variable light CDRs (vlCDR or VLCDR), each containing three CDRs: vhCDR1, vhCDR2, and vhCDR3 of the heavy chain, and vlCDR1, vlCDR2, and vlCDR3 of the light chain. The CDRs are present in the variable heavy domain and the variable light domain, respectively, and together form the Fv region. (See Table 1 and the relevant considerations above regarding the CDR numbering scheme). Thus, in some cases, the six CDRs of the antigen-binding domain are contributed by the variable heavy domain and the variable light domain. In the "Fab" type, the set of six CDRs is contributed by two distinct polypeptide sequences: a variable heavy domain (vh or VH; containing vhCDR1, vhCDR2, and vhCDR3) and a variable light domain (vl or VL; containing vlCDR1, vlCDR2, and vlCDR3), where the C-terminus of the vh domain binds to the N-terminus of the CH1 domain of the heavy chain, and the C-terminus of the vl domain binds to the N-terminus of the constant light domain (thus forming a light chain). In the scFv type, the vh and vl domains are covalently linked to a single polypeptide sequence by the use of a linker ("scFv linker"), as outlined herein, which may be either vh-linker-vl or vl-linker-vh (starting from the N-terminus), with the former being generally preferred (each side may contain an optional domain linker, depending on the type used (e.g., from Figure 1). Generally, the C-terminus of the scFv domain is linked to the N-terminus of the hinge of the second monomer.

[0049] In this specification, “modification” means the substitution, insertion, and / or deletion of amino acids in a polypeptide sequence, or a change to a portion chemically linked to a protein. For example, a modification may be a change in carbohydrates or a PEG structure linked to a protein. In this specification, “amino acid modification” means the substitution, insertion, and / or deletion of amino acids in a polypeptide sequence. For clarification, unless otherwise stated, an amino acid modification is always a change to an amino acid encoded by DNA, for example, 20 amino acids that have codons in DNA and RNA.

[0050] In this specification, “amino acid substitution” or “substitution” means replacing one amino acid with another at a specific position in the parent polypeptide sequence. In particular, in some embodiments, the substitution is for an amino acid that does not spontaneously occur at a particular position, and does not spontaneously occur either within a living organism or in any organism. For example, substitution E272Y refers to a mutant polypeptide, in this case an Fc mutant in which glutamic acid is replaced with tyrosine at position 272. For clarification, a protein that is manipulated to alter the nucleic acid coding sequence but not alter the starting amino acid (e.g., replacing CGG (which codes for arginine) with CGA (which still codes for arginine) to increase the expression level in a host organism) is not an “amino acid substitution.” That is, if a protein has the same amino acid at the specific position where it started, without creating a new gene that codes for the same protein, it is not an amino acid substitution.

[0051] As used herein, “amino acid insertion” or “insertion” means the addition of an amino acid sequence at a specific position in the parent polypeptide sequence. For example, -233E or 233E specifies the insertion of glutamic acid after position 233 and before position 234. Furthermore, -233ADE or A233ADE specifies the insertion of AlaAspGlu after position 233 and before position 234.

[0052] As used herein, “amino acid deletion” or “deletion” means the removal of an amino acid sequence at a specific position in the parent polypeptide sequence. For example, E233- or E233#, E233() or E233del specify a deletion of glutamic acid at position 233. In addition, EDA233- or EDA233# specifies a deletion of the sequence GluAspAla starting at position 233.

[0053] As used herein, “mutant protein” or “protein variant” or “variant” means a protein that differs from that of the parent protein for reasons of at least one amino acid modification. A protein variant has at least one but not many amino acid modifications compared to the parent protein, and the mutant protein does not align with the parent protein using alignment programs such as those described below. Generally, mutant proteins (such as mutant Fc domains outlined herein) are generally at least 75, 80, 85, 90, 91, 92, 93, 94, 95, 96, 97, 98, or 99% identical to the parent protein using alignment programs such as BLAST described below.

[0054] As described below, in some embodiments, the parent polypeptide, for example, the Fc parent polypeptide, is a human wild-type sequence such as a heavy constant domain or Fc region derived from IgG1, IgG2, IgG3, or IgG4. However, a human sequence having a variant can also function as the “parent polypeptide,” for example, the IgG1 / 2 hybrid described in U.S. Publication No. 2006 / 0134105. The protein variant sequences herein preferably have at least about 80% identity with the parent protein sequence, most preferably at least about 90% identity, and more preferably at least 95-98-99% identity. As used herein, “antibody variant” or “mutant antibody” means an antibody that differs from the parent antibody due to at least one amino acid modification; as used herein, “IgG variant” or “mutant IgG” means an antibody that differs from the parent IgG (and similarly, often from a human IgG sequence) due to at least one amino acid modification; as used herein, “immunoglobulin variant” or “mutant immunoglobulin” means an immunoglobulin sequence that differs from the parent immunoglobulin sequence due to at least one amino acid modification. As used herein, “Fc variant” or “mutant Fc” means a protein that contains an amino acid modification in its Fc domain compared to the Fc domain of human IgG1, IgG2, or IgG4.

[0055] The Fc variants of the present invention are defined according to the amino acid modifications that constitute them. For example, N434S or 434S is an Fc variant having a serine substitution at position 434 relative to the parent Fc polypeptide, and the numbering follows the EU index. Similarly, M428L / N434S defines an Fc variant having substitutions M428L and N434S relative to the parent Fc polypeptide. The identity of the WT amino acids does not need to be specified; in that case, the variant described above is called 428L / 434S. Note that the rules by which substitutions are provided are arbitrary. That is, for example, N434S / M428L is the same Fc variant as M428L / N434S. In all positions discussed in the present invention and related to antibodies, unless otherwise stated, the numbering of amino acid positions follows the EU index. The EU index, or the EU index as in Kabat or the EU numbering scheme, refers to the numbering of EU antibodies. Kabat et al. collected numerous primary sequences of variable regions in the heavy and light chains. Based on the degree of sequence conservation, they classified the individual primary sequences into CDRs and frameworks and compiled a list (see SEQUENCES OF IMMUNOLOGICAL INTEREST, 5th edition, NIH publication, No. 91-3242, E. Kabat et al., the entire list is incorporated by reference). See also Edelman et al., 1969, Proc Natl Acad Sci USA 63:78-85, the entire list is incorporated by reference. Modifications may be additions, deletions, or substitutions.

[0056] In this specification, “protein” means at least two covalently bonded amino acids and includes proteins, polypeptides, oligopeptides, and peptides. In addition, polypeptides used to construct antibodies of the present invention may include synthetic derivatization, glycosylation, PEGylation, circular permutation, cyclization, linking to other molecules, fusion to proteins or protein domains, and addition of peptide tags or peptide labels to one or more side chains or terminals.

[0057] As used herein, “residue” refers to a position in a protein and is associated with amino acid identity. For example, asparagine 297 (also known as Asn297 or N297) is the residue at position 297 in the human antibody IgG1.

[0058] As used herein, “Fab” or “Fab region” generally means a polypeptide comprising VH, CH1, VL, and CL immunoglobulin domains on two different polypeptide chains (e.g., VH-CH1 on one chain and VL-CL on the other). Fab may refer to this region alone or, in the context of the bispecific antibodies of the present invention, to this region. In the context of Fab, Fab includes the Fv region in addition to the CH1 and CL domains.

[0059] As used herein, “Fv,” “Fv fragment,” or “Fv region” means a polypeptide containing the VL and VH domains of ABD. The Fv region is a polypeptide containing two distinct regions, Fab (as described above, generally including the constant region outlined above). ) may be formalized as both and scFv, where the vl and vh domains are combined (generally in the linker discussed herein) to form scFv.

[0060] In this specification, “single-strand Fv” or “scFv” means a variable heavy domain covalently bonded to a variable light domain using the scFv linkers generally discussed herein to form an scFv or scFv domain. The .scFv domain may be oriented either N-to-C-terminus (vh-linker-vl or vl-linker-vh). In the sequences shown in the sequence listings and figures, the order of the vh and vl domains is indicated by their names; for example, H.X_L.Y means that the N-to-C-terminus is vh-linker-vl, and L.Y_H.X means vl-linker-vh.

[0061] As used herein, “IgG subclass modification” or “isotype modification” means an amino acid modification that converts one amino acid in one IgG isotype to a corresponding amino acid in a different, aligned IgG isotype. For example, since IgG1 contains tyrosine at EU position 296 and IgG2 contains phenylalanine, the F296Y substitution in IgG2 is considered an IgG subclass modification.

[0062] As used herein, “non-spontaneous modification” means an amino acid modification that is not isotype. For example, since none of human IgGs contain serine at position 434, the substitution 434S in IgG1, IgG2, IgG3, or IgG4 (or their hybrids) is considered a non-spontaneous modification.

[0063] As used herein, “amino acid” and “amino acid identity” mean one of the 20 naturally occurring amino acids encoded by DNA and RNA.

[0064] As used herein, “effector function” means a biochemical event resulting from the interaction between an antibody Fc region and an Fc receptor or Fc ligand. Effector functions include, but are not limited to, ADCC, ADCP, and CDC.

[0065] As used herein, “IgG Fc ligand” means any biologically derived molecule, preferably a polypeptide, that binds to the Fc region of an IgG antibody to form an Fc / Fc ligand complex. Fc ligands include, but are not limited to, FcγRI, FcγRII, FcγRIII, FcRn, C1q, C3, mannan-binding lectins, mannose receptors, staphylococcal protein A, streptococcal protein G, and viral FcγR. Fc ligands also include Fc receptor homologs (FcRH) (Davis et al., 2002, Immunological Reviews 190:123-136, incorporated in whole by reference), which are a family of Fc receptors homologous to FcγR. Fc ligands may also include undiscovered molecules that bind to Fc. Specific IgG Fc ligands are FcRn and Fc gamma receptors. As used herein, "Fc ligand" means any biologically derived molecule, preferably a polypeptide, that binds to the Fc region of an antibody to form an Fc / Fc ligand complex.

[0066] As used herein, “Fc gamma receptor,” “FcγR,” or “Fc gamma R” means any member of the family of proteins that bind to the Fc region of an IgG antibody and are encoded by the FcγR gene. In humans, this family includes, but is not limited to, FcγRI(CD64) including isoforms FcγRIa, FcγRIb, and FcγRIc, isoform FcγRIIa (including allotypes H131 and R131), isoform FcγRIIb (including FcγRIIb-1 and FcγRIIb-2), and isoforms FcγRI(CD64), isoform FcγRIIa (including allotypes H131 and R131), isoform FcγRIIb (including FcγRIIb-1 and FcγRIIb-2), and isoforms FcγRI(CD64). This includes FcγRII(CD32) including FcγRIIc, and FcγRIII(CD16) including isoforms FcγRIIIa (including allotypes V158 and F158) and FcγRIIIb (including allotypes FcγRIIb-NA1 and FcγRIIb-NA2) (the entirety of Jefferis et al., 2002, Immunol Lett 82:57-65 is incorporated by reference), as well as any undiscovered human FcγR or FcγR isoform or FcγR allotype. FcγR may be of any biological origin, but is not limited to humans, mice, rats, rabbits, and monkeys. Mouse FcγR includes, but is not limited to, FcγRI(CD64), FcγRII(CD32), FcγRIII(CD16), and FcγRIII-2(CD16-2), as well as any undiscovered mouse FcγR or FcγR isoform or FcγR allotype.

[0067] As used herein, “FcRn” or “neonatal Fc receptor” means a protein that binds to the Fc region of an IgG antibody, and is at least partially encoded by the FcRn gene. FcRn may be of any biological origin, but is not limited to humans, mice, rats, rabbits, and monkeys. As is known in the art, functional FcRn proteins consist of two polypeptides, often referred to as the heavy chain and the light chain. The light chain is beta-2-microglobulin, and the heavy chain is encoded by the FcRn gene. Unless otherwise stated, FcRn or FcRn protein refers to the complex of the FcRn heavy chain with beta-2-microglobulin. Various FcRn variants are used to increase binding to the FcRn receptor and, in some cases, to increase the serum half-life. “FcRn variants” are those that increase binding to the FcRn receptor, and preferred FcRn variants are listed below.

[0068] As used herein, “parent polypeptide” means a starting polypeptide that is later modified to produce a mutant. The parent polypeptide may be a naturally occurring polypeptide, or a mutant or engineered form of a naturally occurring polypeptide. Accordingly, as used herein, “parent immunoglobulin” means an unmodified immunoglobulin polypeptide that is modified to produce a mutant, and as used herein, “parent antibody” means an unmodified antibody that is modified to produce a mutant antibody. It should be noted that “parent antibody” includes known commercially available recombinant antibodies, as outlined below. In this context, “parent Fc domain” is relative to the listed mutants; therefore, “mutant human IgG1 Fc domain” is compared to the parent Fc domain of human IgG1, and “mutant human IgG4 Fc domain” is compared to the parent Fc domain of human IgG4, etc.

[0069] As used herein, “Fc,” “Fc region,” or “Fc domain” means a polypeptide comprising the CH2-CH3 domain of an IgG molecule, and possibly the hinge. In the EU numbering of human IgG1, the CH2-CH3 domain comprises amino acids 231-447, and the hinge comprises amino acids 216-230. Therefore, the definition of “Fc domain” includes both amino acids 231-447 (CH2-CH3) or 216-447 (hinge-CH2-CH3), or fragments thereof. In this context, an “Fc fragment” may contain fewer amino acids from either or both of the N and C-terminuses, but still retains the ability to form a dimer having another Fc domain or Fc fragment when detected using standard methods generally based on size (e.g., non-denaturing chromatography, size exclusion chromatography, etc.). The human IgG Fc domain, as used in this invention, is of particular use and may be an Fc domain derived from human IgG1, IgG2, or IgG4.

[0070] The "mutant Fc domain" contains amino acid modifications compared to the parent Fc domain. Therefore, the "mutant human IgG1 Fc domain" contains amino acid modifications (generally amino acid substitutions, but in the case of cleavage mutants, amino acid deletions) compared to the human IgG1 Fc domain. Generally, the mutant Fc domain is the same as the corresponding parent human IgG The mutant Fc domain has at least approximately 80, 85, 90, 95, 97, 98, or 99 percent identity with respect to the Fc domain (using the identity algorithms discussed below, and in one embodiment, using default parameters, utilizing the BLAST algorithm known in the art). Alternatively, the mutant Fc domain may have 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 amino acid modifications compared to the parent Fc domain. In addition, as discussed herein, the mutant Fc domain retains the ability to form dimers with other Fc domains, as measured using known techniques described herein, such as non-denaturing gel electrophoresis.

[0071] In this specification, “heavy chain constant region” means the CH1-hinge-CH2-CH3 portion of an antibody (or fragment thereof) excluding the variable heavy domain, which in the EU numbering of human IgG1 corresponds to amino acids 118-447. In this specification, “heavy chain constant region fragment” means a heavy chain constant region that contains fewer amino acids from either or both of the N and C-terminuses but still retains the ability to form a dimer with another heavy chain constant region.

[0072] As used herein, “position” means the location in a protein sequence. Positions may be sequentially numbered or follow an established format, for example, the EU index for antibody numbering.

[0073] As used herein, “target antigen” means a molecule that is specifically bound by the antigen-binding domain, which includes the variable region of a given antibody. In this example, as will be discussed below, the target antigen is a checkpoint inhibitor protein.

[0074] In this specification, “strandedness” in the monomer composition of the heterodimer antibody of the present invention means that, similar to two strands of “fitting” DNA, the heterodimerizing variants are incorporated into each monomer in such a way that they retain the ability to “fit” and form a heterodimer. For example, if some pI variants are manipulated into monomer A (e.g., to increase pI), then stereovariates that are “charge pairs” can also be used without interfering with the pI variants. For example, by adding a charge variant that increases pI to the same “strand” or the same “monomer,” both functionalities are preserved. Similarly, for “asymmetric” variants occurring in pairs of sets, as fully outlined below, those skilled in the art will consider pI when deciding which strands or monomers to incorporate into a pair of sets so that pI separation is also maximized using the asymmetric pI.

[0075] As used herein, “target cell” means a cell that expresses a target antigen.

[0076] In this specification, in the context of producing bispecific antibodies according to the present invention, "host cell" means a cell that contains foreign nucleic acids encoding components of the bispecific antibody and that can express the bispecific antibody under suitable conditions. Suitable host cells are discussed below.

[0077] As used herein, “variable region” or “variable domain” means substantially formed by one of the Vκ, Vλ, and / or VH genes that make up the loci of kappa immunoglobulin, lambda immunoglobulin, and heavy chain immunoglobulin, respectively. This refers to a region of immunoglobulin containing one or more Ig domains encoded by and containing a CDR that confers antigen specificity. Thus, a "variable heavy domain" is paired with a "variable light domain" to form an antigen-binding domain ("ABD"). In addition, each variable domain contains three hypervariable regions ("complementarity-determining regions," "CDRs") (vhCDR1, vhCDR2, and vhCDR3 for variable heavy domains, and vlCDR1, vlCDR2, and vlCDR3 for variable light domains) and four framework (FR) regions, arranged from the amino terminus to the carboxyl terminus in the following order: FR1-CDR1-FR2-CDR2-FR3-CDR3-FR4.

[0078] In this specification, "wild-type" or "WT" means the amino acid or nucleotide sequence found in nature, including allelic variation. WT proteins have an amino acid or nucleotide sequence that has not been intentionally modified.

[0079] The present invention provides several antibody domains having sequence identity with respect to human antibody domains. Sequence identity between two similar sequences (e.g., antibody variable domains) is described in Smith, TF & Waterman, MS (1981) “Comparison of Biosequences, Adv.Appl.Math.2:482 [Local homology algorithm], Needleman, SB & Wunsch, CD. (1970) "A General Method Applicable To The Search For Similarities In The Amino Acid Sequence Of Two Proteins," J.Mol.Biol.48:443 [Homologous alignment algorithm], Pearson, WR & Lipman, DJ (1988) "Improved Tools For Biological Sequence Comparison," Proc.Natl.Acad.Sci.(USA)85:2444 [Similarity search method], or Altschul, SF et al, (1990) "Basic Local Alignment Search Sequence identity can be measured using algorithms such as the "BLAST" algorithm (see https: / / blast.ncbi.nlm.nih.gov / Blast.cgi), described in Tool, J.Mol.Biol.215:403-10. When using any of the aforementioned algorithms, default parameters (window length, gap penalty, etc.) are used. In one embodiment, sequence identity is performed using the BLAST algorithm with default parameters.

[0080] The antibodies of the present invention are generally isolated or recombinant. When “isolated” is used in the description of the various polypeptides disclosed herein, it means a polypeptide that has been identified, isolated, and / or recovered from the cells or cell cultures in which it is expressed. Typically, isolated polypeptides will be prepared by at least one purification step. “Isolated antibody” refers to an antibody that substantially does not contain other antibodies with different antigen specificities. “Recombinant” means that the antibody is produced in exogenous host cells using recombinant nucleic acid techniques, and they can also be isolated.

[0081] "Specific binding", or "specifically bind to" a particular antigen or epitope, or "specific for" a particular antigen or epitope, means binding that is measurably different from non-specific interactions. Specific binding can generally be measured, for example, by determining the binding of a molecule compared to the binding of a control molecule, which is a molecule of a similar structure that does not have binding activity. For example, specific binding can be determined by competition with a control molecule similar to the target.

[0082] Specific binding to a particular antigen or epitope can be demonstrated, for example, when an antibody directed against the antigen or epitope has a KD of at least about 10 -4 M, at least about 10 -5 M, at least about 10 -6 M, at least about 10 -7 M, at least about 10 -8 M, at least about 10 -9 M, or at least about 10 -10 M, at least about 10 -11 M, at least about 10 -12 M or greater, where KD refers to the dissociation rate of a particular antibody-antigen interaction. Typically, an antibody that specifically binds to an antigen has a KD for the antigen or epitope that is 20-fold, 50-fold, 100-fold, 500-fold, 1000-fold, 5,000-fold, 10,000-fold, or more than these multiples of the KD of the control molecule.

[0083] Furthermore, specific binding to a particular antigen or epitope can be demonstrated, for example, by the antibody having a KA or Ka toward that antigen or epitope at a multiple of at least 20-, 50-, 100-, 500-, 1000-, 5000-, 10000-, or greater than these multiples compared to a control, where KA or Ka refers to the association rate of a particular antibody-antigen interaction. Binding affinity is generally measured using Biacore, SPR, or BLI assays.

[0084] E. Antibodies The present invention relates to the production of bispecific checkpoint antibodies that bind to two different checkpoint antigens, as discussed herein. The term “antibody” is used in general terms, as discussed below. Antibodies useful in the present invention exhibit several types as described herein and include traditional antibodies, as well as derivatives, fragments, and mimetic forms of antibodies described herein and shown in the figures.

[0085] Traditional antibody structural units typically consist of tetramers. Each tetramer typically comprises two identical pairs of polypeptide chains, each pair having one "light" chain (typically having a molecular weight of about 25 kDa) and one "heavy" chain (typically having a molecular weight of about 50–70 kDa). Human light chains are classified as kappa light chains and lambda light chains. This invention generally deals with bispecific antibodies based on the IgG class, which has several subclasses, but is not limited to IgG1, IgG2, IgG3, and IgG4. Generally, IgG1, IgG2, and IgG4 are used more frequently than IgG3. It should be noted that IgG1 has different allotypes with polymorphisms in 356 (D or E) and 358 (L or M). The sequences shown herein use the 356E / 358M allotype, but other allotypes are also included herein. In other words, any sequence containing an IgG1 Fc domain as included herein may have 356D / 358L replacing the 356E / 358M allotype.

[0086] In addition, many of the antibodies herein have at least one cysteine ​​that is replaced by serine at position 220, which is generally on the “scFv monomer” side with respect to most of the sequences shown herein, but may be on the “Fab monomer” side or both to reduce disulfide formation. One or both of these replaced cysteines (C220S) are specifically included in the sequences herein.

[0087] Therefore, as used herein, “isotype” means any subclass of immunoglobulin defined by the chemical and antigenic properties of its constant region. It should be understood that therapeutic antibodies may also include hybrids of isotypes and / or subclasses. For example, as shown in U.S. Publication No. 2009 / 0163699 incorporated by reference, the present invention relates to the use of human IgG1 / G2 hybrids.

[0088] The hypervariable region consists of approximately amino acid residues 24-34 (LCDR1; "L" indicates the light chain), 50-56 (LCDR2), and 89-97 (LCDR3) in the light chain variable region, and the heavy chain In the variable region, approximately 31-35B (HCDR1; "H" indicates the heavy chain), approximately 50-65 (HCDR2), and approximately 95-102 (HCDR3) amino acid residues; Kabat et al., SEQUENCES OF PROTEINS OF IMMUNOLOGICAL INTEREST, 5th Ed. Public Health Service, National Institutes of Health, Bethesda, Md. (1991), and / or those residues that form the hypervariable loop (e.g., residues 26-32 (LCDR1), 50-52 (LCDR2), and 91-96 (LCDR3) in the light chain variable region, and 26-32 (HCDR1), 53-55 (HCDR2), and 96-101 (HCDR3) in the heavy chain variable region; Chothia and It is common to include amino acid residues derived from Lesk (1987) J.Mol.Biol.196:901-917. Specific CDRs of the present invention are described below.

[0089] As those skilled in the art will understand, the precise numbering and arrangement of CDRs may differ between different numbering systems. However, it should be understood that the disclosure of variable weight and / or variable light sequences includes the disclosure of the relevant (unique) CDRs. Thus, the disclosure of each variable weight region is a disclosure of vhCDRs (e.g., vhCDR1, vhCDR2, and vhCDR3), and the disclosure of each variable light region is a disclosure of vlCDRs (e.g., vlCDR1, vlCDR2, and vlCDR3). A useful comparison of CDR numbering is given below; see Lafranc et al., Dev. Comp. Immunol. 27(l):55-77 (2003). Table 1

[0090] [Table 1] [Table 2]

[0091] Throughout this specification, the Kabat numbering system is generally used when referring to residues in the variable domain (approximately residues 1-107 in the light chain variable region and residues 1-113 in the heavy chain variable region), and the EU numbering system is generally used for the Fc region (e.g., Kabat et al., above (1991)).

[0092] Another type of heavy chain Ig domain is the hinge region. Hereinafter, “hinge,” “hinge region,” “antibody hinge region,” or “hinge domain” refers to a mobile polypeptide containing amino acids between the first and second constant domains of an antibody. Structurally, the IgG CH1 domain terminates at EU position 215, and the IgG CH2 domain begins at residue EU position 231. Therefore, for IgG, the antibody hinge is defined herein as containing positions 216 (E216 in IgG1) to 230 (p230 in IgG1), with numbering following the EU index as found in Kabat. In some cases, a “hinge fragment” is used, which contains fewer amino acids at either the N-terminus or C-terminus, or both, of the hinge domain. As shown herein, pI variants can also be constructed in the hinge region.

[0093] A light chain generally consists of two domains: a variable light domain (including the light chain CDR) and a variable heavy domain. It includes a constant light chain region (often called CL or Cκ) and a Fv region (which together with the nucleotides).

[0094] Another region of interest for additional substitutions, outlined below, is the Fc region.

[0095] The present invention provides a number of different sets of CDRs. In this case, a “complete set of CDRs” includes three variable light CDRs and three variable heavy CDRs, e.g., vlCDR1, vlCDR2, vlCDR3, vhCDR1, vhCDR2, and vhCDR3. These may each be part of a larger variable light domain or variable heavy domain. In addition, as will be outlined more fully herein, the variable heavy domains and variable light domains may be on separate polypeptide chains when heavy and light chains are used (e.g., when Fab is used), or on a single polypeptide chain in the case of scFv sequences.

[0096] CDRs contribute to the formation of antigen binding, or more specifically, the formation of an antibody epitope binding site. An "epitope," also known as a paratope, refers to a determinant that interacts with a specific antigen-binding site in the variable region of an antibody molecule. Epitopes are classifications of molecules, such as amino acids or sugar side chains, that typically possess specific structural and charge properties. A single antigen may have two or more epitopes.

[0097] An epitope may include amino acid residues directly involved in binding (also known as the immunodominant components of the epitope), as well as other amino acid residues not directly involved in binding, such as amino acid residues that are effectively blocked by specific antigen-binding peptides. In other words, amino acid residues fall within the footprint of the specific antigen-binding peptide.

[0098] Epitopes can be either steric or linear. Steric epitopes are produced from different segments of a linear polypeptide chain by spatially juxtaposed amino acids. Linear epitopes are produced by adjacent amino acid residues in a polypeptide chain. Steric and non-steric epitopes may be distinguished in that binding to the former but not to the latter disappears in the presence of a denaturing solvent.

[0099] An epitope typically contains at least three amino acids in its unique spatial structure, and more commonly, at least five or eight to ten amino acids. Antibodies that recognize the same epitope can be validated in a simple immunoassay demonstrating the ability of one antibody to block the binding of another antibody to a target antigen, e.g., "binning." As outlined below, the present invention includes not only the antigen-binding domains and antibodies listed herein, but also those that compete for binding with the epitopes bound by the listed antigen-binding domains.

[0100] Accordingly, the present invention provides different antibody domains. As described herein and known in the art, the heterodimeric antibodies of the present invention contain different domains in the chain and light chain, and these domains may also overlap. These domains include, but are not limited to, Fc domains, CH1 domains, CH2 domains, CH3 domains, hinge domains, heavy constant domains (CH1-hinge-Fc domain or CH1-hinge-CH2-CH3), variable heavy domains, variable light domains, light constant domains, FAb domains, and scFv domains.

[0101] Therefore, the “Fc domain” includes the -CH2-CH3 domain and optionally a hinge domain (-H-CH2-CH3). In embodiments of this specification, when scFv is bonded to an Fc domain, it is bonded to all or part of the hinge of the Fc domain. The C-terminus of the scFv construct is bound to the sequence EPKS, which is the hinge initiation, for example. The heavy chain comprises a variable heavy domain and a constant domain, and a CH1-optional hinge-Fc domain containing CH2-CH3. The light chain comprises a variable light chain and a light constant domain. The scFv comprises the variable heavy chain, the scFv linker, and the variable light domain. In most constructs and sequences outlined herein, the C-terminus of the variable heavy chain is bound to the N-terminus of the scFv linker, and its C-terminus is bound to the N-terminus of the variable light chain (N-vh-linker-vh-C), although this can be interchanged (N-vl-linker-vh-C).

[0102] Some embodiments of the present invention include at least one scFv domain containing a variable heavy domain and a variable light domain, which do not occur spontaneously but are generally linked together by an scFv linker. As outlined herein, scFv domains are generally oriented from N to C-terminus, such as vh-scFv linker-vl, but this can be reversed with respect to either scFv domain (or one constructed using vh and vl sequences from Fab), and the linker of any choice at one or both ends depends on the type (see generally Figure 1).

[0103] As shown herein, there are several suitable linkers that can be used (for use as either domain linkers or scFv linkers), which can be used to covalently bond listed domains, including traditional peptide bonds produced by recombinant technology. In some embodiments, the linker peptide may mainly consist of the following amino acid residues: Gly, Ser, Ala, or Thr. The linker peptide should have a length appropriate for linking two molecules so that they take on the correct conformational structure relative to each other, and as a result they retain the desired activity. In one embodiment, the linker is about 1 to 50 amino acids long, preferably about 1 to 30 amino acids long. In one embodiment, a linker with a length of 1 to 20 amino acids may be used with about 5 to about 10 amino acids, which are useful in some embodiments. Useful linkers include glycine-serine polymers (e.g., (GS)n, (GSGGS)n (SEQ ID NO: 37756), (GGGGS)n (SEQ ID NO: 37757), and (GGGS)n (SEQ ID NO: 37758), where n is an integer at least 1 (and generally 3-4)), glycine-alanine polymers, alanine-serine polymers, and other mobile linkers. Alternatively, various non-proteinaceous polymers, including but not limited to polyethylene glycol (PEG), polypropylene glycol, polyoxyalkylene, or copolymers of polyethylene glycol and polypropylene glycol, may be useful as linkers.

[0104] Other linker sequences may contain any sequence of some length from the CL / CH1 domain, but not all residues of the CL / CH1 domain, for example, the first 5 to 12 amino acid residues of the CL / CH1 domain. Linkers may be derived from immunoglobulin light chains, e.g., Cκ or Cλ. Linkers may be derived from immunoglobulin heavy chains of some isotype, e.g., Cγ1, Cγ2, Cγ3, Cγ4, Cα1, Cα2, Cδ, Cε, and Cμ. Linker sequences may be derived from other proteins, such as Ig-like proteins (e.g., TCR, FcR, KIR), hinge region sequences, and other native sequences from other proteins.

[0105] In some embodiments, the linker is a “domain linker” used to link any two domains outlined together herein. For example, in Figure 1F, there may be a domain linker that links the C-terminus of the CH1 domain of Fab to the N-terminus of scFv, and another optional domain linker that links the C-terminus of scFv to the CH2 domain (however, in many embodiments, a hinge is used as this domain linker). Any preferred linker may be used, but many embodiments utilize a glycine-serine polymer as the domain linker, which includes, for example, (GS)n, (GSGGS)n (SEQ ID NO: 37756), (GGGGS)n (SEQ ID NO: 37757), and (GGGS)n (SEQ ID NO: 37758) (where n is an integer at least 1 (and generally 3-4-5)), as well as any peptide sequence that enables recombination linking of two domains, having sufficient length and flexibility to allow each domain to retain its biological function. In some cases, and when considering "chain-like" properties as outlined below, charged domain linkers may be used, as in some embodiments of scFv linkers.

[0106] In some embodiments, the linker is an scFv linker used to covalently bond the vh and vl domains, as discussed herein. In many cases, the scFv linker is a charged scFv linker, and some of them are

[0107] This is shown in Figure 7. Accordingly, the present invention further provides a charged scFv linker to facilitate separation at pI between the first monomer and the second monomer. That is, by incorporating a charged scFv linker, whether positive or negative (or both in the case of a scaffold using scFv on different monomers), this allows the monomer containing the charged linker to change the pI without further modifying the Fc domain. These charged linkers can be substituted for any scFv containing a standard linker. Furthermore, as will be understood by those skilled in the art, the charged scFv linker is used on the correct “chain” or monomer according to the desired change at pI. For example, as discussed herein, to produce a triple F-type heterodimer antibody, the initial pI of the Fv region for each of the desired antigen-binding domains is calculated, one is selected to produce an scFv, and by pI, either a positive or negative linker is selected.

[0108] The charged domain linker can also be used to increase the pI separation of monomers according to the present invention, and therefore,

[0109] The components shown in Figure 7 may be used in any embodiment of this specification in which a linker is utilized.

[0110] Specifically, the type shown in Figure 1 is an antibody commonly called a "heterodimal antibody," meaning that the protein has at least two associated Fc sequences that self-assemble into a heterodimer Fc domain, and at least two Fv regions (whether as Fab or scFv).

[0111] Chimeric antibodies and humanized antibodies In certain embodiments, the antibody of the present invention comprises a heavy chain variable region derived from a specific germline heavy chain immunoglobulin gene and / or a light chain variable region derived from a specific germline light chain immunoglobulin gene. For example, such an antibody comprises or may comprise a human antibody that is a "product of" a specific germline sequence or that contains a heavy chain variable region or light chain variable region "derived from" a specific germline sequence. A human antibody that is a "product of" a human germline immunoglobulin sequence or that "derived from" a human germline immunoglobulin sequence may be identified in this way by comparing the amino acid sequence of the human antibody with the amino acid sequence of a human germline immunoglobulin and selecting the human germline immunoglobulin sequence whose sequence is closest to (i.e., the sequence with the greatest identity %) of the human antibody (using the method outlined herein). A human antibody that is a "product of" a specific human germline immunoglobulin sequence or that "derived from" a specific human germline immunoglobulin sequence may contain amino acid differences compared to the germline sequence, for example, by the intentional introduction of spontaneously occurring somatic mutations or site-directed mutations. Humanized antibodies typically have an amino acid sequence that is at least 90% identical to the amino acid sequence encoded by the human germline immunoglobulin gene, and contain amino acid residues that identify the antibody as derived from a human sequence when compared to the germline immunoglobulin amino acid sequence of another species (e.g., mouse germline sequence). In certain cases, the humanized antibody may be at least 95, 96, 97, 98, or 99% identical to the amino acid sequence encoded by the germline immunoglobulin gene, or even at least 96%, 97%, 98, or 99% identical. Typically, a humanized antibody derived from a particular human germline sequence shows no more than 10-20 amino acid differences from the amino acid sequence encoded by the human germline immunoglobulin gene (the number of variants before the introduction of any asymmetric, pI, and cleavage variants as defined herein, i.e., before the introduction of the variants of the present invention, is generally small). In certain cases, the humanized antibody does not exhibit five or more amino acid differences from the amino acid sequence encoded by the germline immunoglobulin gene, or even four, three, two, or one or more amino acids (similarly, the number of variants before the introduction of any asymmetric, pI, and cleavage variants herein, i.e., before the introduction of the variants of the present invention, is generally small).

[0112] In one embodiment, the parent antibody is affinity-mature, as is known in the art. Structure-based methods may be used for humanization and affinity maturation, as described, for example, in USSN 11 / 004, 590. Selection-based methods may be used for humanization and / or affinity maturation of the antibody variable region, but are not limited to, Wu et al., 1999, J.Mol.Biol.294:151-162, Baca et al., 1997, J.Biol.Chem.272(16):10678-10684, Rosok et al., 1996, J.Biol.Chem.271(37):22611-22618, Rader et al., 1998, Proc.Natl.Acad.Sci.USA 95:8910-8915, Krauss et al. This includes methods described in al., 2003, Protein Engineering 16(10):753-759, all of which are incorporated in their entirety by reference. Other humanization methods may involve only partial transplantation of the CDR and are not limited to those described in USSN09 / 810, 510, Tan et al., 2002, J.Immunol.169:1119-1125, and De Pascalis et al., 2002, J.Immunol.169:3076-3084, all of which are incorporated in their entirety by reference.

[0113] IV. Heterodimal antibodies Therefore, in some embodiments, the present invention provides heterodimer checkpoint antibodies that rely on the use of two different heavy chain mutant Fc sequences that self-assemble to form heterodimer Fc domains and heterodimer antibodies.

[0114] Accordingly, the present invention relates to novel constructs for providing heterodimer antibodies that enable binding to two or more checkpoint antigens or ligands, for example, to enable bispecific binding. Heterodimer antibody constructs are based on the self-assembly properties of two "monomers" that organize into, for example, two Fc domains of the antibody's heavy chain, "dimer". Heterodimer antibodies are produced by altering the amino acid sequence of each monomer, which will be discussed in full below. Accordingly, the present invention generally relates to the creation of heterodimer checkpoint antibodies that can co-capture antigens in several ways, by relying on amino acid variants in the constant region that differ in each chain, thereby promoting heterodimerization and / or facilitating telodimer purification with respect to homodimers.

[0115] Therefore, the present invention provides a bispecific antibody. The ongoing problem in antibody technology is to bind to two different antigens simultaneously, and generally, to bring different antigens into close proximity in this way. There is a strong demand for "bispecific" antibodies that enable new functionalities and therapeutics. Generally, these antibodies are produced by incorporating the heavy and light chain genes, respectively, into the host cell. This generally results in the formation of the desired heterodimer (AB), as well as two homodimers (AA and BB (without the issue of light chain heterodimers)). However, the main obstacle in bispecific antibody formation is the difficulty in purifying heterodimeric antibodies to remove homodimeric antibodies and / or biasing heterodimer formation to outweigh homodimer formation.

[0116] Several mechanisms exist that can be used to generate the heterodimers of the present invention. In addition, as will be understood by those skilled in the art, these mechanisms can be combined to ensure high heterodimerization. Thus, amino acid variants that lead to the production of heterodimers are called “heterodimerizing variants.” As will be discussed below, heterodimerizing variants may include stereovariates (e.g., the “knob-and-hole” or “asymmetric” variants described below, and the “charge pair” variants described below), as well as “pI variants” that allow for the removal of heterodimers and the purification of homodimers. As a result, the whole is incorporated by reference, and useful mechanisms of heterodimerization include the "knob and hole" ("KIH"), sometimes referred to herein as the "asymmetric" variant (see discussion in WO2014 / 145806), the "electrostatic maneuvering" or "charge pair" described in WO2014 / 145806, the pI variant described in WO2014 / 145806, and additional general Fc variants outlined in WO2014 / 145806 and below.

[0117] The present invention involves several fundamental mechanisms that facilitate the purification of heterodimer antibodies. Specifically, it relies on the use of pI variants, resulting in monomers each possessing a different pI, thereby enabling isoelectric focus purification of AA, AB, and BB dimer proteins. Alternatively, some scaffold types, such as the "triple F" type, also allow for size-based separation. As will be further outlined below, "asymmetric" configurations are also possible, where heterodimerization outweighs homodimerization. Therefore, combinations of stereodimerizing variants and pI variants or charge-pair variants are particularly useful in the present invention.

[0118] Generally, embodiments particularly useful in the present invention depend on a set of mutants including asymmetric mutants, which, together with pI mutants that increase the pI difference between two monomers, promote heterodimerization in preference to homodimerization, thereby facilitating the removal of homodimers and the purification of heterodimers.

[0119] In addition, as will be fully outlined below, depending on the type of heterodimeric antibody, the pI variant may be contained either within the constant and / or Fc domain of the monomer, or a charged linker, which is either a domain linker or an scFv linker, may be used. That is, scaffolds utilizing scFv(s), such as the triple-F type, may include a charged scFv linker(s) (either positive or negative) to provide a further increase in pI for purification purposes. As will be understood by those skilled in the art, some triple-F types have only a charged scFv linker and are useful without additional pI adjustment, but the present invention also provides pI variants and / or charged domain linkers that are present in one or both of the monomers. In addition, additional amino acids manipulated for alternative functionality may also give pI changes such as Fc, FcRn, and KO variants.

[0120] In this invention, which utilizes pI as a separation mechanism to enable the purification of heterodimeric proteins, the amino acid variant is introduced into one or both of the monomeric polypeptides. It is also possible that one pI of the monomers (hereinafter simply referred to as “monomer A”) may be manipulated to remove monomer B, or both changes in monomers A and B may be charged to increase the pI of monomer A and decrease the pI of monomer B. As will be discussed, changes in the pI of either or both monomers can be made by removing or adding a charged residue (e.g., a neutral amino acid is replaced by a positively or negatively charged amino acid residue, e.g., glycine to glutamic acid), changing a charged residue from positive or negative to the opposite charge (e.g., aspartic acid to lysine), or changing a charged residue to a neutral residue (e.g., lysine to serine, which eliminates the charge). Several of these variants are shown in the figure.

[0121] Accordingly, this embodiment of the present invention provides the creation of a sufficient change in pI in at least one of the monomers, as a result the heterodimer can be separated from the homodimer. As will be understood by those skilled in the art, and as will be discussed further below, this can be done by using a “wild-type” heavy chain constant region and a mutant region manipulated to have either an increased or decreased pI (wt A-+B or wt A--B), or by increasing one region and decreasing the other (A+-B- or A-B+).

[0122] Therefore, generally, components of some embodiments of the present invention are amino acid variants in the constant region of an antibody, which are directed to alter the isoelectric point (pI) of both, or if not both, at least one, monomer of a dimeric protein in order to form a “pI antibody” by incorporating an amino acid substitution (“pI variant” or “pI substitution”) into one or both of the monomers. As shown herein, the separation of heterodimers from two homodimers can be achieved if the pI of the two monomers differs by as little as 0.1 pH units, and differences of 0.2, 0.3, 0.4, and 0.5 or greater are all useful in the present invention.

[0123] As will be understood by those skilled in the art, the number of pI variants present in each or both monomers to obtain good separation depends in part on the starting pIs of the constituent elements, e.g., triple F type, and the starting pIs of the target scFv and Fab. That is, the Fv sequences of the two target antigens are calculated and a decision is made from there to determine which monomer to manipulate or in which "direction" (e.g., more positive or more negative). As is known in the art, different Fvs have different starting pIs that are utilized in the present invention. Generally, as outlined herein, the manipulation of the pI results in a total pI difference of at least about 0.1 log in each monomer, and as outlined herein, such total pI difference is preferably 0.2 to 0.5.

[0124] Furthermore, as will be understood by those skilled in the art and outlined herein, in some embodiments, heterodimers can be separated from homodimers based on size. Some types, for example, as shown in Figure 1, allow for the separation of heterodimers and homodimers based on size.

[0125] A. Heterodimerized mutants The present invention provides heterodimer proteins, which include various types of heterodimer antibodies that utilize heterodimer variants to remove homodimers and enable heterodimer formation and / or purification.

[0126] There are pairs of several suitable sets of heterodimerizing asymmetric mutants. These mutants occur in pairs of sets; that is, one set of the pair is incorporated into the first monomer. Rarely, the other set of a pair is incorporated into the second monomer. Note that these sets do not necessarily function as “knob-in-hole” mutants with a one-to-one correspondence between residues on one monomer and those on the other; rather, these pairs form an interface between the two monomers that promotes heterodimerization and prevents homodimerization, allowing the percentage of heterodimers spontaneously formed under biological conditions to exceed 90% rather than the expected 50% (25% homodimer A / A: 50% heterodimer A / B: 25% homodimer B / B).

[0127] B. Stereomorphs In some embodiments, heterodimer formation may be facilitated by the addition of stereomutants. That is, by altering the amino acids in each heavy chain, different heavy chains are more likely to associate to form a heterodimer structure than to form homodimers with the same Fc amino acid sequence. Suitable stereomutants are shown in the figure.

[0128] A mechanism commonly referred to in the art as "knob and hole" may also be optionally used, which involves amino acid manipulation to create steric effects that promote heterodimer formation and prevent homodimer formation. (USSN61 / 596,846, Ridgway et al.) As described in al., Protein Engineering 9(7):617(1996); Atwell et al., J.Mol.Biol.1997 270:26, and U.S. Patent No. 8,216,805 (all of which are incorporated herein by reference in their entirety), these mutations are sometimes referred to as "knob-and-hole." Several "monomer A-monomer B" pairs that depend on "knob-and-hole" mutations are identified in the figure. In addition, as described in Merchant et al., Nature Biotech.16:677(1998), these "knob-and-hole" mutations, in combination with disulfide bonds, can make the formation of heterodimers asymmetric.

[0129] An additional mechanism useful for heterodimer formation is sometimes called “electrostatic manipulation,” as described in Gunasekaran et al., J. Biol. Chem. 285(25):19637(2010) (the whole of which is incorporated herein by reference). This is sometimes referred to herein as “charge pairing.” In this embodiment, electrostatics are used to make the formation asymmetric towards heterodimerization. Those skilled in the art will understand that these may also have an effect on pI, and thus may also have an effect on purification, and therefore, in some cases, may be considered to be pI mutants. However, since these are generated to promote heterodimerization and are not used as a means of purification, they are classified as “stereomutants.” These include, but are not limited to, D221E / P228E / L368E paired with D221R / P228R / K409R (for example, these are "the set of monomers that correspond") and C220E / P228E / 368E paired with C220R / E224R / P228R / K409R.

[0130] Additional monomer A and monomer B variants, the pI variants outlined herein, or other figures, descriptions, and sequence numbers of this patent document shown in Figure 37 of US2012 / 0149876 are expressly incorporated herein by reference.

[0131] In some embodiments, the stereomutae outlined herein can be optionally and independently incorporated into one or both monomers together with any pI mutant (or other mutants such as Fc mutants, FcRn mutants, etc.), thereby independently and optionally including or excluding the protein of the present invention.

[0132] A list of preferred asymmetric variants is shown in Figure 3, and Figure 8 shows specific variants in many embodiments. Several pairs of usefulness are shown. Those particularly used in many embodiments are, but are not limited to, pairs of sets including S364K / E357Q:L368D / K370S, L368D / K370S:S364K, L368E / K370S:S364K, T411T / E360E / Q362E:D401K, L368D / K370S:S364K / E357L, K370S:S364K / E357Q, and T366S / L368A / Y407V:T366W (optionally cross-linked disulfide, T366S / L368A / Y407V / Y349C:T366W / S354C). From a nomenclature standpoint, as mentioned above, the pair "S364K / E357Q:L368D / K370S" means that one monomer has the double mutant set S364K / E357Q and the other has the double mutant set L368D / K370S, and the "chain-likeness" of these pairs depends on the starting pI.

[0133] C. Heterodimal pI (isoelectric point) mutant Generally, as those skilled in the art will understand, pI variants fall into two common categories: those that increase the pI of a protein (changes in basicity) and those that decrease the pI of a protein (changes in acidity). As described herein, all combinations of these variants are possible; that is, one monomer may be wild-type or a variant that does not exhibit a significantly different pI from the wild-type, while the other may be more basic or more acidic. Alternatively, each monomer may be altered, one becoming more basic and the other more acidic.

[0134] Preferred combinations of pI variants are shown in Figure 4. These modifications are shown for IgG1, as outlined herein and shown in the figure, but all isotypes and isotype hybrids can be modified in this way. R133E and R133Q can also be used if the heavy chain constant domain is derived from IgG2-4.

[0135] In one embodiment, for example, in Figures 1A, E, F, G, H, and Type I, a preferred combination of pI mutants comprises one monomer containing the 208D / 295E / 384D / 418E / 421D mutant (N208D / Q295E / N384D / Q418E / N421D for human IgG1) and a second monomer (positive scFv side) containing a positively charged scFv linker (containing (GKPGS)4) (SEQ ID NO: 37755). However, as will be understood by those skilled in the art, the first monomer contains a CH1 domain including position 208. Therefore, in constructs that do not contain the CH1 domain (for example, in dual scFv or "one-arm" types, such as those shown in Figures 1B, C, or D, for antibodies that do not utilize the CH1 domain in one of their domains), the preferred negative pI variant Fc set includes the 295E / 384D / 418E / 421D variant (Q295E / N384D / Q418E / N421D for human IgG1).

[0136] Therefore, in some embodiments, one monomer has a set of substitutions from Figure 4, and the other monomer has a charged linker (either in the form of a charged scFv linker, as the monomer includes an scFv or a charged domain linker, as the type indicates, which can be selected from those shown in Figure 7).

[0137] 1. Isotype variants In addition, many embodiments of the present invention rely on the "incorporation" of pI amino acids at specific sites from one IgG isotype to another, thus reducing or eliminating the potential for unwanted immunogenicity to be introduced into the variant. Some of these are shown in Figure 21 of U.S. Publication No. 2014 / 0370013, which is incorporated herein by reference. That is, IgG1 is a common isotype for therapeutic antibodies for a variety of reasons, including high effector function. However, the polyconstant region of IgG1 has a higher pI compared to that of IgG2 (8.10 vs. 7.31). At specific sites, IgG2 residues Introducing a specific amino acid into the IgG1 backbone reduces (or increases) the pI of the resulting monomer, and additionally prolongs the serum half-life. For example, IgG1 has glycine at position 137 (pI 5.97), and IgG2 has glutamic acid (pI 3.22). In other words, the incorporation of glutamic acid affects the pI of the resulting protein. As described below, several amino acid substitutions are generally required to have a significant effect on the pI of mutant antibodies. However, it should be noted that even changes in the IgG2 molecule can lead to an increase in serum half-life, as discussed below.

[0138] As further described below, in other embodiments, non-isotype amino acid changes are made either to reduce the overall charge state of the resulting protein (for example, by changing higher pI amino acids to lower pI amino acids) or to enable structural adaptation toward stability.

[0139] In addition, significant changes can be observed in each monomer of the heterodimer by manipulating the pI of both the heavy and light steady domains. As discussed herein, by differentiating the pI of the two monomers by at least 0.5, separation by ion-exchange chromatography, isoelectric focusing, or other isoelectric-sensitive methods becomes possible.

[0140] D.pI calculation The pI of each monomer may depend on the pI of the variable heavy chain constant region, including the mutant heavy chain constant domain and its fusion partner, and the total monomer. Therefore, in some embodiments, the change in pI is calculated based on the mutant heavy chain constant domain using the chart in Figure 19 of U.S. Publication No. 2014 / 0370013. As discussed herein, the choice of monomer to work with is generally determined by the intrinsic pI of the Fv and scaffold region. Alternatively, the pI of each monomer can be compared.

[0141] E. A pI variant that also confers better FcRn in vivo binding than the pI variant. If a pI variant reduces monomeric pI, an additional benefit may be improved retention in serum in vivo.

[0142] Although still under investigation, the Fc region is thought to have a longer half-life in vivo, because binding to FcRn at pH 6 in endosomes sequesters Fc (Ghetie and Ward, 1997 Immunol Today. 18(12):592-598 is incorporated in its entirety by reference). Subsequently, the endosomal compartment recirculates Fc to the cell surface. Once the compartment opens to the extracellular space at a higher pH of approximately 7.4, the release of Fc is induced and it returns to the bloodstream. In mice, Dall'Acqua et al. showed that Fc mutants with increased FcRn binding at pH 6 and pH 7.4 actually decreased serum concentrations and had the same half-life as wild-type Fc (Dall'Acqua et al. 2002, J.Immunol. 169:5171-5180 is incorporated in its entirety by reference). The increased affinity of Fc towards FcRn at pH 7.4 is thought to prevent Fc from being released and returning to the bloodstream. Therefore, Fc mutations that would increase the half-life of Fc in vivo would ideally increase FcRn binding at lower pH levels while still allowing Fc release at higher pH levels. The amino acid histidine changes its charge state in the pH range of 6.0–7.4. Therefore, the finding that His residues occupy an important position in the Fc / FcRn complex is not surprising.

[0143] Recently, antibodies with variable regions that have lower isoelectric points have also been shown to have longer serum half-lives. This has been suggested (Igawa et al., 2010 PEDS. 23(5):385-392, incorporated in its entirety by reference). However, the understanding of this mechanism remains insufficient. Moreover, the variable region differs from antibody to antibody. As described herein, constant region variants with reduced pI and extended half-life offer a more modular approach to improving the pharmacokinetic properties of antibodies.

[0144] F. Additional Fc variants for additional functionality In addition to pI amino acid variants, there are several useful Fc amino acid modifications that can be implemented for various reasons, including, but are not limited to, alterations to binding to one or more FcγR receptors, alterations to binding to FcRn receptors, etc.

[0145] Accordingly, the proteins of the present invention may include amino acid modifications, which include heterodimerized mutants outlined herein, including pI mutants and stereomutae. Each set of mutants may independently and selectively include or exclude any particular heterodimerized protein.

[0146] G.FcγR mutant Therefore, there are several useful Fc substitutions that can be made to alter binding to one or more of the FcγR receptors. Substitutions that result in increased binding and decreased binding may be useful. For example, increased binding to FcγRIIIa is known to result in increased ADCC (antibody-dependent cell-mediated cytotoxicity, a cell-mediated reaction in which nonspecific cytotoxic cells expressing FcγR recognize antibodies bound to target cells, subsequently causing lysis of the target cells). Similarly, decreased binding to FcγRIIb (an inhibitory receptor) may also be beneficial in some circumstances. Useful amino acid substitutions in the present invention include those listed in USSN11 / 124,620 (particularly Figure 41), USSN11 / 174,287, USSN11 / 396,495, and USSN11 / 538,406, all of which are explicitly incorporated herein in whole by reference, and in particular incorporated toward the variants disclosed therein. Useful specific variants include, but are not limited to, 236A, 239D, 239E, 332E, 332D, 239D / 332E, 267D, 267E, 328F, 267E / 328F, 236A / 332E, 239D / 332E / 330Y, 239D, 332E / 330L, 243A, 243L, 264A, 264V, and 299T.

[0147] In addition, there are additional Fc substitutions that are useful in increasing binding to the FcRn receptor and increasing serum half-life, as specifically disclosed in USSN12 / 341,769, which is incorporated in whole by reference, and include, but are not limited to, 434S, 434A, 428L, 308F, 259I, 428L / 434S, 259I / 308F, 436I / 428L, 436I, or V / 434S, 436V / 428L, and 259I / 308F / 428L.

[0148] H. Cutting Mutant Similarly, another category of functional variants is “Fcγ cleavage variants” or “Fc knockout (FcKO or KO)” variants. In these embodiments, for several therapeutic applications, it is desirable to reduce or eliminate the normal binding of the Fc domain to one or more or all of the Fcγ receptors (e.g., FcγR1, FcγRIIa, FcγRIIb, FcγRIIIa, etc.) to avoid additional mechanisms of action. That is, for example, in many embodiments, particularly in the use of bispecific checkpoint antibodies, it is desirable to cleave the FcγRIIIa binding so that one of the Fc domains contains one or more Fcγ receptor cleavage variants to eliminate or significantly reduce ADCC activity. These cleavage variants are shown in Figure 5, and each is G236R / L328R These may be independently and optionally included or excluded, along with preferred embodiments that utilize cleavage mutants selected from the group consisting of E233P / L234V / L235A / G236del / S239K, E233P / L234V / L235A / G236del / S267K, E233P / L234V / L235A / G236del / S239K / A327G, and E233P / L234V / L235A / G236del. Note that the cleavage mutants referenced herein generally cleave FcγR bonds rather than FcRn bonds.

[0149] As is known in the art, the Fc domain of human IgG1 has the best binding to the Fcγ receptor, and therefore, cleavage variants can be used when the constant domain (or Fc domain) in the heterodimeric antibody backbone is IgG1. Alternatively, or in addition to cleavage variants in the IgG1 background, mutations at glycosylation site 297 (generally to A or S) can significantly cleave binding to FcγRIIIa, for example. Human IgG2 and IgG4 have naturally reduced binding to the Fcγ receptor, and therefore, these backbones can be used with or without cleavage variants.

[0150] I. Combinations of heterodimers and Fc mutants As those skilled in the art will understand, all of the listed heterodimerized mutants (including asymmetric and / or pI mutants) can be combined in any way, arbitrarily and independently, as long as the "chain-like" or "monomer partition" is preserved. In addition, all of these mutants can be incorporated into any of the heterodimerized forms.

[0151] In the case of pI mutants, particularly useful embodiments are shown in the figure, but other combinations can be generated by following basic rules that alter the pI difference between the two monomers for ease of purification.

[0152] In addition, any of the heterodimerized, asymmetric, and pI variants can also be independently and arbitrarily combined with Fc-cleaved variants, Fc variants, and FcRn variants, as generally outlined herein.

[0153] V. Useful types of the present invention As will be understood by those skilled in the art and will be fully discussed below, the bispecific heterodimer antibodies of the present invention can generally take on a wide variety of configurations, as shown in Figure 1. Some figures show a "single-ended" configuration in which one specificity is located on one "arm" of the molecule and a different specificity is located on the other "arm". Other figures show a "dual-ended" configuration in which at least one specificity is located on the "upper" part of the molecule and one or more different specificities are located on the "lower" part of the molecule. Accordingly, the present invention relates to novel immunoglobulin compositions that co-capture different first and second antigens.

[0154] As will be understood by those skilled in the art, the heterodimer forms of the present invention may have different binding valencies and be bispecific. That is, the heterodimer antibody of the present invention may be bivalent and bispecific, in which one checkpoint target is bound by one ABD and the other checkpoint target is bound by a second ABD. The heterodimer antibody may also be trivalent and bispecific, in which the first antigen is bound by two ABDs and the second antigen is bound by a second ABD.

[0155] A. Bottle opener type One heterodimer scaffold particularly useful in the present invention is the “triple F” or “bottle opener” scaffold type shown in Figure 1A. In this embodiment, one heavy chain of the antibody contains a single Fv (defined hereafter as “scFv”), and the other heavy chain is a “typical” Fab type, comprising a variable heavy chain and a light chain. This structure is sometimes referred to herein as the “triple F” type (scFv-Fab-Fc), or the “bottle opener” type because it roughly resembles a bottle opener visually (see Figure 1A). The two chains are brought together by the use of amino acid variants in the constant region (e.g., the Fc domain, CH1 domain, and / or hinge region) that promotes the formation of the heterodimer antibody, which is described more fully below.

[0156] This "triple F" type has several distinctly different advantages. As is known in the art, antibody analogs that rely on two scFv constructs often have stability and aggregation problems, but in this invention, these problems can be mitigated by adding "typical" heavy and light chain pairs. In addition, in contrast to types that rely on two heavy chains and two light chains, there is no problem of the heavy and light chains being incorrectly paired (for example, heavy 1 being paired with light 2, etc.).

[0157] Many of the embodiments outlined herein generally depend on a bottle opener type comprising a first monomer containing an scFv, which comprises a variable heavy domain and a variable light domain covalently bonded using an scFv linker (which is often, but not always, charged), and the scFv is typically covalently bonded to the N-terminus of the first Fc domain via a domain linker (which may be either uncharged or charged, as outlined herein) and may be exogenous or endogenous (e.g., all or part of the native hinge domain). The second monomer of the bottle opener type is a heavy chain, and the composition further comprises a light chain.

[0158] In addition, the bottle-opener type Fc domain is generally asymmetric mutant (for example, the amino acid substitution sets shown in Figures 3 and 8, particularly useful asymmetric mutants include S364K / E357Q:L368D / K370S, L368D / K370S:S364K, L368E / K370S:S364K, T411T / E360E / Q362E:D401K, L368D / K370S:S364K / E357L, K The group consists of 370S:S364K / E357Q, T366S / L368A / Y407V:T366W, and T366S / L368A / Y407V / Y349C:T366W / S354C, optionally includes cleavage mutants (including those shown in Figure 5), optionally includes charged scFv linkers (including those shown in Figure 7), and the heavy chain includes pI mutants (including those shown in Figure 4).

[0159] In some embodiments, the bottle opener type includes asymmetric mutants, pI mutants, and cleavage mutants. Thus, some embodiments provide a bottle opener type which includes a) a first monomer ("scFv monomer") comprising a charged scFv linker (in some embodiments, the +H sequence in Figure 7 is preferred), an asymmetric mutant S364K / E357Q, a cleavage mutant E233P / L234V / L235A / G236del / S267K, and Fv that binds to the checkpoint receptor outlined herein, and b) c) light chains, and a second monomer ("Fab monomer") comprising a variable heavy domain that, together with a variable light domain, forms an Fv that binds to a second checkpoint receptor outlined herein, as an asymmetric mutant L368D / K370S, a pI mutant N208D / Q295E / N384D / Q418E / N421D, a cleavage mutant E233P / L234V / L235A / G236del / S267K, and a light chain. In this particular embodiment, preferred monomer Fv pairs include (Fab first, then scFv) PD-1 and CTLA-4, CTLA-4 and PD-1, PD-1 and TIM-3, TIM-3 and PD-1, PD-1 Examples include LAG-3, LAG-3XPD1, PD-1 and TIGIT, TIGIT and PD-1, PD-1 and BTLA, BTLA and PD-1, CTLA-4 and TIM-3, TIM-3 and CTLA-4, CTLA-4 and LAG-3, LAG-3 and CTLA-4, CTLA-4 and TIGIT, TIGIT and CTLA-4, CTLA-4 and BTLA, BTLA and CTLA-4, TIM-3 and LAG-3, LAG-3 and TIM-3, TIM-3 and TIGIT, TIGIT and TIM-3, TIM-3 and BTLA, BTLA and TIM-3, LAG-3 and TIGIT, TIGIT and LAG-3, LAG-3 and BTLA, BTLA and LAG-3, BTLA and TIGIT, and TIGIT and BTLA. In this particular embodiment, a bottle opener having these variants has an scFv side containing ABD 1G6_L1.194_H1.279 that binds to PD-1, and is particularly useful. In this particular embodiment, a bottle opener having these variants has an scFv side containing [CTLA-4]_H3.23_L0.129 ABD that binds to CTLA-4, and is particularly useful.

[0160] Some embodiments, particularly those used in bottle opener types, include CTLA-4XPD-1, LAG-3XPD-1, BTLAXPD-1, TIM-3XPD-1, and LAG-3XCTLA-4.

[0161] The ABD sequences of these combinations are disclosed in the sequence list or shown in Figures 9-13, and may be any combination shown in Figures 39 and 40.

[0162] In some embodiments, the bottle opener type includes asymmetric mutants, pI mutants, cleavage mutants, and FcRn mutants. Thus, some embodiments provide a bottle opener type which includes a) a first monomer ("scFv monomer") comprising a charged scFv linker (in some embodiments, the +H sequence in Figure 7 is preferred), an asymmetric mutant S364K / E357Q, a cleavage mutant E233P / L234V / L235A / G236del / S267K, an FcRn mutant M428L / N434S, and Fv that binds to checkpoint inhibitors outlined herein, and b) The second monomer ("Fab monomer") includes, together with the asymmetric mutant L368D / K370S, the pI mutant N208D / Q295E / N384D / Q418E / N421D, the cleavage mutant E233P / L234V / L235A / G236del / S267K, the FcRn mutant M428L / N434S, and a variable heavy domain that, together with the variable light domain, forms an Fv that binds to the second checkpoint inhibitor outlined herein, and c) a light chain. In this particular embodiment, preferred pairs of Fv include (Fab first, then scFv) PD-1 and CTLA-4, CTLA-4 and PD-1, PD-1 and TIM-3, TIM-3 and PD-1, PD-1 and LAG-3, LAG-3XPD1, PD-1 and TIGIT, TIGIT and PD-1, PD-1 and BTLA, BTLA and PD-1, CTLA-4 and TIM-3, TIM-3 and CTLA-4, CTLA-4 and LAG-3, LAG-3 and CTLA-4, CTLA-4 and TIGIT, TIGIT and CTLA-4, CTLA-4 and BTLA, BTLA and CTLA-4, TIM-3 and LAG-3, LAG-3 and TIM-3, TIM-3 and TIGIT, TIGIT and TIM-3, TIM-3 and BTLA, BTLA and TIM-3. LAG-3 and TIGIT, TIGIT and LAG-3, LAG-3 and BTLA, BTLA and LAG-3, BTLA and TIGIT, and TIGIT and BTLA. In this particular embodiment, bottle openers having these variants have an scFv side containing ABD 1G6_L1.194_H1.279 that binds to PD-1, which is particularly useful.In this particular embodiment, bottle openers having these variants have an scFv side containing [CTLA-4]_H3.23 L0.129 ABD that binds to CTLA-4, and are particularly useful.

[0163] Some embodiments, particularly those used in bottle opener types, include CTLA-4XPD-1, LAG-3XPD-1, BTLAXPD-1, TIM-3XPD-1, and LAG-3XCTLA-4.

[0164] Specifically, Figure 37 shows several bottle opener "skeleton" sequences lacking an Fv sequence that may be used in the present invention. That is, the Fv sequences of the scFv and Fab portions may be any combination of PD-1 and CTLA-4, PD-1 and TIM-3, PD-1 and LAG-3, PD-1 and TIGIT, PD-1 and BTLA, CTLA-4 and TIM-3, CTLA-4 and LAG-3, CTLA-4 and TIGIT, CTLA-4 and BTLA, TIM-3 and LAG-3, TIM-3 and TIGIT, TIM-3 and BTLA, LAG-3 and TIGIT, LAG-3 and BTLA, and TIGIT and BTLA. The sequences may be any of those disclosed herein in the sequence listing and / or in Figures 9-13.

[0165] Regarding the bottle opener skeleton 1 in Figure 37, specific combinations of Fv use in the present invention include PD-1 and CTLA-4, PD-1 and TIM-3, PD-1 and LAG-3, PD-1 and TIGIT, PD-1 and BTLA, CTLA-4 and TIM-3, CTLA-4 and LAG-3, CTLA-4 and TIGIT, CTLA-4 and BTLA, TIM-3 and LAG-3, TIM-3 and TIGIT, TIM-3 and BTLA, LAG-3 and TIGIT, LAG-3 and BTLA, and TIGIT and BTLA. The sequences may be any of those disclosed in the sequence listing and / or in Figures 9-13 herein.

[0166] Regarding the bottle opener skeleton 1 in Figure 37, specific combinations of Fv use in the present invention include CTLA-4(Fab)XPD-1(scFv), PD-1(Fab)XCTLA-4(scFv), LAG-3(Fab)XPD-1(scFv), BTLA(Fab)XPD-1(scFv), and LAG-3(Fab)XCTLA-4(scFv).

[0167] For bottle opener skeleton 1 in Figure 37 (including the 428L / 434S variant at will), specific ABDs that bind to human PD-1 include, but are not limited to, 1G6_H1.279_L1.194, 1G6_H1.280_L1.224, 1G6_L1.194_H1.279, 1G6_L1.210_H1.288, and 2E9_H1L1, as well as those listed in sequence numbers 6209-11464, 11465-17134, 33003-33072, 33073-35394, and 36127-36146.

[0168] Regarding the bottle opener skeleton 1 in Figure 37 (including the 428L / 434S mutant at will), the specific ABDs that bind to human CTLA-4 are not limited to [CTLA-4]_H0.25_L0, [CTLA-4]_H0.26_L0, [CTLA-4]_H0.27_L0, [CTLA-4]_H0.29_L0, [CTLA-4]_H0.38_L0, [CTLA-4]_H0.39_L0, [CTLA-4]_H0.40_L0, [CTLA-4]_H0.70_L0, [CTLA-4]_H0_L0.22, [CTLA-4]_H2_L0, and [CTLA-4]_H3.2 1_L0.124, [CTLA-4]_H3.21_L0.129, [CTLA-4]_H3.21_L0.132, [CTLA-4]_H3. 23_L0.124, [CTLA-4]_H3.23_L0.129, [CTLA-4]_H3.23_L0.132, [CTLA-4]_H3 .25_L0.124, [CTLA-4]_H3.25_L0.129, [CTLA-4]_H3.25_L0.132, [CTLA-4]_H 3.4_L0.118, [CTLA-4]_H3.4_L0.119, [CTLA-4]_H3.4_L0.12, [CTLA-4]_H3.4 _L0.121, [CTLA-4]_H3.4_L0.122, [CTLA-4]_H3.4_L0.123, [CTLA-4]_H3.4_L0.124, [CTLA-4]_H3.4_L0.125, [CTLA-4]_H3.4_L0.126, [CTL A-4]_H3.4_L0.127, [CTLA-4]_H3.4_L0.128, [CTLA-4]_H3.4_L0.129, [CTLA-4]_H3.4_L0.130, [CTLA-4]_H3.4_L0.131, [CTLA-4]_H3.4_L0 Examples include .132, [CTLA-4]_H3.5_L2.1, [CTLA-4]_H3.5_L2.2, [CTLA-4]_H3.5_L2.3, [CTLA-4]_H3_L0, [CTLA-4]_H3_L0.22, [CTLA-4]_H3_L0.44, [CTLA-4]_H3_L0.67, and [CTLA-4]_H3_L0.74, as well as those listed in sequence numbers 21-2918, 2919-6208, 36739-36818, and 35395-35416.

[0169] Regarding the bottle opener skeleton 1 in Figure 37 (including the 428L / 434S mutant at arbitrary selection), the specific ABDs that bind to human LAG-3 are not limited to, but include 2A11_H0L0, 2A11_H1.125_L2.113, 2A11_H1.144_L2.142, 2A11_H1_L2.122, and 2A11_H1_ L2.123, 2A11_H1_L2.124, 2A11_H1_L2.25, 2A11_H1_L2.47, 2A11_H1_L2.50, 2A11_H 1_L2.91, 2A11_H1_L2.93, 2A11_H1_L2.97, 2A11_H1L1, 2A11_H1L2, 2A11_H2L2, 2A11 _H3L1, 2A11_H3L2, 2A11_H4L1, 2A11_H4L2, 7G8_H0L0, 7G8_H1L1, 7G8_H3.18_L1.11, 7G8_H3.23_L1.11, 7G8_H3.28_L1, 7G8_H3.28_L1.11, 7G8_H3.28_L1.13, 7G8_H3.30 Examples include those listed as _L1.34, 7G8_H3.30_L1.34, and 7G8_H3L1, as well as those listed as sequence numbers 17135-20764, 36819-36962, 35417-35606, 25194-32793, and 32794-33002.

[0170] For the bottle opener skeleton 1 in Figure 37 (including the 428L / 434S variant, if selected), specific ABDs that bind to human BTLA include, but are not limited to, 9C6_H0L0, 9C6_H1.1_L1, and 9C6_H1.11_L1, as well as those listed in sequence numbers 20885-21503 and 36707-36738.

[0171] For bottle opener skeleton 1 in Figure 37 (including the 428L / 434S variant at will), specific ABDs that bind to human TIM-3 include, but are not limited to, 1D10_H0L0, 1D12_H0L0, 3H3_H1_L2.1, 6C8_H0L0, 6D9_H0_1D12_L0, 7A9_H0L0, 7B11_H0L0, 7B11var_H0L0, and 7C2_H0L0, as well as those listed in sequence numbers 20765-20884, 37587-37698, and 36347-36706.

[0172] A specific embodiment of the bottle opener is outlined below.

[0173] B.mAb-Fv type One heterodimer scaffold particularly useful in the present invention is the mAb-Fv type shown in Figure 1H. In this embodiment, the type relies on the use of C-terminal binding of an "extra" variable heavy domain to one monomer and C-terminal binding of an "extra" variable light domain to the other monomer, thus forming a third antigen-binding domain, where the Fab portions of the two monomers bind to one checkpoint target and the "extra" scFv domain binds to a different checkpoint target.

[0174] In this embodiment, the first monomer comprises a first heavy chain including a first variable heavy domain and a first steady heavy domain, the first steady heavy domain comprising a first Fc domain having a first variable light domain covalently bonded to the C-terminus of the first Fc domain using a domain linker (vh1-CH1-hinge-CH2-CH3-[optional linker]-vl2). The second monomer comprises a second variable weight domain of a second stationary weight domain containing a second Fc domain, and a third variable weight domain covalently bonded to the C-terminus of the second Fc domain using a domain linker (vh1-CH1-hinge-CH2-CH3-[optional linker]-vh2). The variable domains bonded to the two C-terminuses make up the scFv. This embodiment further utilizes a common light chain containing a variable light domain and a stationary light domain, which is related to the heavy chain to form two identical Fabs. For many of the embodiments herein, these constructs include asymmetric variants, pI variants, cleavage variants, additional Fc variants, etc., as desired and described herein. In this embodiment, preferred pairs of Fvs are PD-1 and CTLA-4, CTLA-4 and PD-1, PD-1 (Fabs are described first, then scFvs). Examples include TIM-3, TIM-3 and PD-1, PD-1 and LAG-3, LAG-3XPD1, PD-1 and TIGIT, TIGIT and PD-1, PD-1 and BTLA, BTLA and PD-1, CTLA-4 and TIM-3, TIM-3 and CTLA-4, CTLA-4 and LAG-3, LAG-3 and CTLA-4, CTLA-4 and TIGIT, TIGIT and CTLA-4, CTLA-4 and BTLA, BTLA and CTLA-4, TIM-3 and LAG-3, LAG-3 and TIM-3, TIM-3 and TIGIT, TIGIT and TIM-3, TIM-3 and BTLA, BTLA and TIM-3. LAG-3 and TIGIT, TIGIT and LAG-3, LAG-3 and BTLA, BTLA and LAG-3, BTLA and TIGIT, and TIGIT and BTLA.

[0175] The ABD sequences of these combinations are disclosed in the sequence listing or are shown in Figures 9-13, and may be any combination shown in Figures 39 and 40.

[0176] In addition, the mAb-Fv type Fc domain has asymmetric mutants (for example, the amino acid substitution sets shown in Figures 3 and 8, particularly useful asymmetric mutants include S364K / E357Q:L368D / K370S, L368D / K370S:S364K, L368E / K370S:S364K, T411T / E360E / Q362E:D401K, L368D / K370S:S364K / E357L, K The group consists of 370S:S364K / E357Q, T366S / L368A / Y407V:T366W, and T366S / L368A / Y407V / Y349C:T366W / S354C, optionally includes cleavage mutants (including those shown in Figure 5), optionally includes charged scFv linkers (including those shown in Figure 7), and the heavy chain includes pI mutants (including those shown in Figure 4).

[0177] In some embodiments, the mAb-Fv type includes asymmetric mutants, pI mutants, and cleavage mutants. Thus, some embodiments include a) a first monomer comprising the asymmetric mutant S364K / E357Q, the cleavage mutant E233P / L234V / L235A / G236del / S267K, and a first variable weight domain and a second variable weight domain that, together with a first variable light domain of the light chain, form an Fv that binds to a first checkpoint inhibitor, and b) the asymmetric mutant L368D / K370S, the pI mutant N208D / Q295 A second monomer comprising E / N384D / Q418E / N421D, the cleavage mutant E233P / L234V / L235A / G236del / S267K, a first variable heavy domain which, together with the first variable light domain, creates an Fv that binds to the first checkpoint inhibitor outlined herein, and a second variable light chain which, together with the second variable heavy chain, forms an Fv(ABD) that binds to the second checkpoint inhibitor, and c) the first This includes a bottle opener type, which comprises a light chain containing variable light domains and constant light domains. Specific uses in some embodiments of this type are (Fab-scFv order) CTLA-4XPD-1, LAG-3XPD-1, BTLAXPD-1, and LAG-3XCTLA-4.

[0178] In some embodiments, the mAb-Fv type includes asymmetric mutants, pI mutants, cleavage mutants, and FcRn mutants. Thus, some embodiments include a) a first monomer comprising the asymmetric mutant S364K / E357Q, the cleavage mutant E233P / L234V / L235A / G236del / S267K, the FcRn mutant M428L / N434S, and a first variable weight domain and a second variable weight domain that, together with a first variable light domain of the light chain, form an Fv that binds to a first checkpoint inhibitor, and b) the asymmetric mutant L368D / K370S, the pI mutant N208D / Q295E / N384D / Q418E / N421 D) A bottle opener type comprising a second monomer including a first variable heavy domain that, together with a first variable light domain, makes up an Fv that binds to a first checkpoint inhibitor outlined herein, and a second variable light chain that, together with a second variable heavy chain, forms an Fv(ABD) that binds to a second checkpoint inhibitor; and c) a light chain including a first variable light domain and a constant light domain. Specific uses in some embodiments of this type include (Fab-scFv order) CTLA-4XPD-1, LAG-3XPD-1, BTLAXPD-1, and LAG-3XCTLA-4.

[0179] Regarding mAb-Fv sequences similar to mAb-scFv skeleton 1 (including M428L / N434S, optionally selected) in Figure 38, specific ABDs that bind to human PD-1 include, but are not limited to, 1G6_H1.279_L1.194, 1G6_H1.280_L1.224, 1G6_L1.194_H1.279, 1G6_L1.210_H1.288, and 2E9_H1L1, as well as those listed in sequence numbers 6209-11464, 11465-17134, 33003-33072, 33073-35394, and 36127-36146.

[0180] Regarding mAb-Fv sequences similar to mAb-scFv skeleton 1 (including M428L / N434S, optionally selected) in Figure 38, specific ABDs that bind to human CTLA-4 include, but are not limited to, [CTLA-4]_H0.25_L0, [CTLA-4]_H0.26_L0, [CTLA-4]_H0.27_L0, [CTLA-4]_H0.29_L0, [CTLA-4]_H0.38_L0, [CTLA-4]_H0.39_L0, [CTLA -4]_H0.40_L0, [CTLA-4]_H0.70_L0, [CTLA-4]_H0_L0.22, [CTLA-4]_H2_L0, [CTLA-4]_H3.21_L0.124, [CTLA-4]_H3.21 _L0.129, [CTLA-4]_H3.21_L0.132, [CTLA-4]_H3.23_L0.124, [CTLA-4]_H3.23_L0.129, [CTLA-4]_H3.23_L0.132, [CTL A-4]_H3.25_L0.124, [CTLA-4]_H3.25_L0.129, [CTLA-4]_H3.25_L0.132, [CTLA-4]_H3.4_L0.118, [CTLA-4]_H3.4_L0. 119, [CTLA-4]_H3.4_L0.12, [CTLA-4]_H3.4_L0.121, [CTLA-4]_H3.4_L0.122, [CTLA-4]_H3.4_L0.123, [CTLA-4]_H3.4 _L0.124, [CTLA-4]_H3.4_L0.125, [CTLA-4]_H3.4_L0.126, [CTLA-4]_H3.4_L0.127, [CTLA-4]_H3.4_L0.128, [CTLA-4] _H3.4_L0.129, [CTLA-4]_H3.4_L0.130, [CTLA-4]_H3.4_L0.131, [CTLA-4]_H3.4_L0.132, [CTLA-4]_H3.5_L2.1, [CTLA Examples include [CTLA-4]_H3.5_L2.2, [CTLA-4]_H3.5_L2.3, [CTLA-4]_H3_L0, [CTLA-4]_H3_L0.22, [CTLA-4]_H3_L0.44, [CTLA-4]_H3_L0.67, and [CTLA-4]_H3_L0.74, as well as those listed in sequence numbers 21-2918, 2919-6208, 36739-36818, and 35395-35416.

[0181] Regarding mAb-scFv sequences similar to mAb-scFv skeleton 1 in Figure 38 (including M428L / N434S, optional), specific ABDs that bind to human LAG-3 include, but are not limited to, 2A11_H0L0, 2A11_H1.125_L2.113, 2A11_H1.144_L2.142, and 2A11_H1_L 2.122, 2A11_H1_L2.123, 2A11_H1_L2.124, 2A11_H1_L2.25, 2A11_H1_L2.47, 2A11_H1_ L2.50, 2A11_H1_L2.91, 2A11_H1_L2.93, 2A11_H1_L2.97, 2A11_H1L1, 2A11_H1L2, 2A11_ H2L2, 2A11_H3L1, 2A11_H3L2, 2A11_H4L1, 2A11_H4L2, 7G8_H0L0, 7G8_H1L1, 7G8_H3.18 _L1.11, 7G8_H3.23_L1.11, 7G8_H3.28_L1, 7G8_H3.28_L1.11, 7G8_H3.28_L1.13, 7G8_ Examples include those listed as H3.30_L1.34, 7G8_H3.30_L1.34, and 7G8_H3L1, as well as those listed as sequence numbers 17135-20764, 36819-36962, 35417-35606, 25194-32793, and 32794-33002.

[0182] For mAb-Fv sequences similar to mAb-scFv skeleton 1 (optionally including M428L / N434S) in Figure 38, specific ABDs that bind to human BTLA include, but are not limited to, 9C6_H0L0, 9C6_H1.1_L1, and 9C6_H1.11_L1, as well as those listed in sequence numbers 20885-21503 and 36707-36738.

[0183] For mAb-Fv sequences similar to mAb-scFv skeleton 1 (including M428L / N434S, optionally selected) in Figure 38, specific ABDs that bind to human TIM-3 include, but are not limited to, 1D10_H0L0, 1D12_H0L0, 3H3_H1_L2.1, 6C8_H0L0, 6D9_H0_1D12_L0, 7A9_H0L0, 7B11_H0L0, 7B11var_H0L0, and 7C2_H0L0, as well as those listed in sequence numbers 20765-20884, 37587-37698, and 36347-36706.

[0184] C.mAb-scFv One heterodimer scaffold particularly useful in the present invention is the mAb-scFv type shown in Figure 1I. In this embodiment, the type relies on the use of a C-terminal binding of scFv to one of the monomers, thus forming a third antigen-binding domain, where the Fab portions of the two monomers bind to one checkpoint target and the "extra" scFv domain binds to a different checkpoint target.

[0185] In this embodiment, the first monomer comprises a first heavy chain (including a variable heavy domain and a constant domain), and the C-terminally covalently bonded scFv comprises an scFv variable light domain, an scFv linker, and an scFv variable heavy domain in either orientation (vh1-CH1-hinge-CH2-CH3-[optional linker]-vh2-scFv linker-vl2 or vh1-CH1-hinge-CH2-CH3-[optional linker]-vl2-scFv linker-vh2). This embodiment includes a variable light domain and a constant light domain. Further utilizing a common light chain, this light chain is associated with the heavy chain to form two identical Fabs that bind to one of the target antigens. For many of the embodiments herein, these constructs include asymmetric variants, pI variants, cleavage variants, additional Fc variants, etc., as desired and described herein. In this embodiment, preferred pairs of Fv include (Fab first, then scFv) PD-1 and CTLA-4, CTLA-4 and PD-1, PD-1 and TIM-3, TIM-3 and PD-1, PD-1 and LAG-3, LAG-3XPD1, PD-1 and TIGIT, TIGIT and PD-1, PD-1 and BTLA, BTLA and PD-1, CTLA-4 and TIM-3, TIM-3 and CTLA-4, CTLA-4 and LAG-3, LAG-3 and CTLA-4, CTLA-4 and TIGIT, TIGIT and CTLA-4, CTLA-4 and BTLA, BTLA and CTLA-4, TIM-3 and LAG-3, LAG-3 and TIM-3, TIM-3 and TIGIT, TIGIT and TIM-3, TIM-3 and BTLA, BTLA and TIM-3. LAG-3 and TIGIT, TIGIT and LAG-3, LAG-3 and BTLA, BTLA and LAG-3, BTLA and TIGIT, and TIGIT and BTLA.

[0186] The ABD sequences of these combinations are disclosed in the sequence listing or are shown in Figures 9-13, and may be any combination shown in Figures 39 and 40.

[0187] In addition, the mAb-scFv type Fc domain generally exhibits asymmetric mutants (e.g., the amino acid substitution sets shown in Figures 3 and 8, particularly useful asymmetric mutants include S364K / E357Q:L368D / K370S, L368D / K370S:S364K, L368E / K370S:S364K, T411T / E360E / Q362E:D401K, L368D / K370S:S364K / E357 The molecule is selected from the group consisting of L, K370S:S364K / E357Q, T366S / L368A / Y407V:T366W, and T366S / L368A / Y407V / Y349C:T366W / S354C), optionally includes a cleavage mutant (including those shown in Figure 5), optionally includes a charged scFv linker (including those shown in Figure 7), and the heavy chain includes a pI mutant (including those shown in Figure 4).

[0188] In some embodiments, the mAb-scFv type includes asymmetric mutants, pI mutants, and cleavage mutants. Thus, some embodiments include a) a first monomer comprising the asymmetric mutant S364K / E357Q, the cleavage mutant E233P / L234V / L235A / G236del / S267K, and a first variable weight domain and a second variable weight domain that, together with a first variable light domain of the light chain, form an Fv that binds to a first checkpoint inhibitor, and b) the asymmetric mutant L368D / K370S, the pI mutant N208D / Q295E / N384D / Q418E / N421 D) A bottle opener type comprising a second monomer including a cleavage mutant E233P / L234V / L235A / G236del / S267K, a first variable heavy domain which, together with a first variable light domain, makes up an Fv that binds to a first checkpoint inhibitor outlined herein, and a second variable light chain which, together with a second variable heavy chain, forms an Fv(ABD) that binds to a second checkpoint inhibitor, and c) a light chain including a first variable light domain and a constant light domain. Specific uses in some embodiments of this type include (Fab-scFv order) CTLA-4XPD-1, LAG-3XPD-1, BTLAXPD-1, and LAG-3XCTLA-4.

[0189] In some embodiments, the mAb-scFv type includes asymmetric mutants, pI mutants, cleavage mutants, and FcRn mutants. Therefore, in some embodiments, a) the asymmetric mutant S364K / E357Q, the cleavage mutant E233P / L234V / L235A / G236del / S267K, the FcRn mutant M428L / N434S, and the first variable light domain of the light chain, together form an Fv that binds to a first checkpoint inhibitor. The present invention includes a first monomer comprising a first variable weight domain and a second variable weight domain, a) a second monomer comprising an asymmetric mutant L368D / K370S, a pI mutant N208D / Q295E / N384D / Q418E / N421D, a cleavage mutant E233P / L234V / L235A / G236del / S267K, a FcRn mutant M428L / N434S, and a first variable light domain which, together with the first variable light domain, creates an Fv that binds to a first checkpoint inhibitor outlined herein, and a second variable light chain which, together with the second variable heavy chain, forms an Fv(ABD) that binds to a second checkpoint inhibitor, and c) a light chain comprising a first variable light domain and a constant light domain. In the mAb-scFv type, specific combinations of Fv usage in the present invention include CTLA-4(Fab)XPD-1(scFv), PD-1(Fab)XCTLA-4(scFv), LAG-3(Fab)XPD-1(scFv), BTLA(Fab)XPD-1(scFv), and LAG-3(Fab)XCTLA-4(scFv).

[0190] In the mAb-scFv skeleton 1 of Figure 38 (including M428L / N434S at the discretion of choice), specific ABDs that bind to human PD-1 include, but are not limited to, 1G6_H1.279_L1.194, 1G6_H1.280_L1.224, 1G6_L1.194_H1.279, 1G6_L1.210_H1.288, and 2E9_H1L1, as well as those listed in sequence numbers 6209-11464, 11465-17134, 33003-33072, 33073-35394, and 36127-36146.

[0191] In the mAb-scFv skeleton 1 shown in Figure 38 (including M428L / N434S at will), specific ABDs that bind to human CTLA-4 include, but are not limited to, [CTLA-4]_H0.25_L0, [CTLA-4]_H0.26_L0, [CTLA-4]_H0.27_L0, [CTLA-4]_H0.29_L0, [CTLA-4]_H0.38_L0, [CTLA-4]_H0.39_L0, [CTLA-4]_H0.40_L0, [CTLA-4]_H0.70_L0, [CTLA-4]_H0_L0.22, and [CTLA-4]_H2_L 0, [CTLA-4]_H3.21_L0.124, [CTLA-4]_H3.21_L0.129, [CTLA-4]_H3.21_L0.132, [CTLA-4]_H3.23_L0.124, [CTLA-4]_H3.23_L0.129, [CTLA-4]_H3. 23_L0.132, [CTLA-4]_H3.25_L0.124, [CTLA-4]_H3.25_L0.129, [CTLA-4]_H3.25_L0.132, [CTLA-4]_H3.4_L0.118, [CTLA-4]_H3.4_L0.119, [CTLA- 4]_H3.4_L0.12, [CTLA-4]_H3.4_L0.121, [CTLA-4]_H3.4_L0.122, [CTLA-4]_H3.4_L0.123, [CTLA-4]_H3.4_L0.124, [CTLA-4]_H3.4_L0.125, [CTLA -4]_H3.4_L0.126, [CTLA-4]_H3.4_L0.127, [CTLA-4]_H3.4_L0.128, [CTLA-4]_H3.4_L0.129, [CTLA-4]_H3.4_L0.130, [CTLA-4]_H3.4_L0.131, [CT Examples include those listed in the sequence numbers LA-4_H3.4_L0.132, [CTLA-4]_H3.5_L2.1, [CTLA-4]_H3.5_L2.2, [CTLA-4]_H3.5_L2.3, [CTLA-4]_H3_L0, [CTLA-4]_H3_L0.22, [CTLA-4]_H3_L0.44, [CTLA-4]_H3_L0.67, and [CTLA-4]_H3_L0.74, as well as those listed in sequence numbers 21-2918, 2919-6208, 36739-36818, and 35395-35416.

[0192] In the mAb-scFv skeleton 1 shown in Figure 38 (including M428L / N434S at will), the specific ABD that binds to human LAG-3 is not limited to, but includes 2A11_H0. L0, 2A11_H1.125_L2.113, 2A11_H1.144_L2.142, 2A11_H1_L2.122, 2A11 _H1_L2.123, 2A11_H1_L2.124, 2A11_H1_L2.25, 2A11_H1_L2.47, 2A11_H1 _L2.50, 2A11_H1_L2.91, 2A11_H1_L2.93, 2A11_H1_L2.97, 2A11_H1L1, 2 A11_H1L2, 2A11_H2L2, 2A11_H3L1, 2A11_H3L2, 2A11_H4L1, 2A11_H4L2, 7G Examples include those listed as 8_H0L0, 7G8_H1L1, 7G8_H3.18_L1.11, 7G8_H3.23_L1.11, 7G8_H3.28_L1, 7G8_H3.28_L1.11, 7G8_H3.28_L1.13, 7G8_H3.30_L1.34, 7G8_H3.30_L1.34, and 7G8_H3L1, as well as those listed as sequence numbers 17135-20764, 36819-36962, 35417-35606, 25194-32793, and 32794-33002.

[0193] In the mAb-scFv skeleton 1 (optionally including M428L / N434S) of Figure 38, specific ABDs that bind to human BTLA include, but are not limited to, 9C6_H0L0, 9C6_H1.1_L1, and 9C6_H1.11_L1, as well as those listed in sequence numbers 20885-21503 and 36707-36738.

[0194] In the mAb-scFv skeleton 1 of Figure 38 (including M428L / N434S at the discretion of choice), specific ABDs that bind to human TIM-3 include, but are not limited to, 1D10_H0L0, 1D12_H0L0, 3H3_H1_L2.1, 6C8_H0L0, 6D9_H0_1D12_L0, 7A9_H0L0, 7B11_H0L0, 7B11var_H0L0, and 7C2_H0L0, as well as those listed in sequence numbers 20765-20884, 37587-37698, and 36347-36706.

[0195] D. Central scFv One heterodimer scaffold particularly useful in the present invention is the central scFv type shown in Figure 1F. In this embodiment, the type relies on the use of an inserted scFv domain, thus forming a third antigen-binding domain, where the Fab portions of the two monomers bind to one checkpoint target and the "extra" scFv domain binds to the other checkpoint target. The scFv domain is inserted between the Fc domain and the CH1-Fv region of one of the monomers, thus providing a third antigen-binding domain.

[0196] In this embodiment, one monomer comprises a first heavy chain containing a first variable heavy domain, a CH1 domain (and an optional hinge), and an Fc domain, while scFv comprises an scFv variable light domain, an scFv linker, and an scFv variable heavy domain. scFv is covalently bonded between the C-terminus of the CH1 domain of the heavy steady domain and the N-terminus of the first Fc domain using an optional domain linker (vh1-CH1-[optional linker]-vh2-scFv linker-vl2-[optional linker with hinge]-CH2-CH3, or the opposite orientation for scFv, vh1-CH1-[optional linker]-vl2-scFv linker-vh2-[optional linker with hinge]-CH2-CH3). The other monomer is the standard Fab side. This embodiment further utilizes a common light chain including a variable light domain and a constant light domain, which is related to a heavy chain to form two identical Fabs that bind to a checkpoint inhibitor. For many of the embodiments herein, these constructs include asymmetric variants, pI variants, cleavage variants, additional Fc variants, etc., as desired and described herein. In this embodiment, a preferred pair of Fvs is PD (where the Fab is described first, followed by the scFv). Examples include -1 and CTLA-4, CTLA-4 and PD-1, PD-1 and TIM-3, TIM-3 and PD-1, PD-1 and LAG-3, LAG-3XPD1, PD-1 and TIGIT, TIGIT and PD-1, PD-1 and BTLA, BTLA and PD-1, CTLA-4 and TIM-3, TIM-3 and CTLA-4, CTLA-4 and LAG-3, LAG-3 and CTLA-4, CTLA-4 and TIGIT, TIGIT and CTLA-4, CTLA-4 and BTLA, BTLA and CTLA-4, TIM-3 and LAG-3, LAG-3 and TIM-3, TIM-3 and TIGIT, TIGIT and TIM-3, TIM-3 and BTLA, and BTLA and TIM-3. LAG-3 and TIGIT, TIGIT and LAG-3, LAG-3 and BTLA, BTLA and LAG-3, BTLA and TIGIT, and TIGIT and BTLA.

[0197] The ABD sequences of these combinations are disclosed in the sequence listing or are shown in Figures 9-13, and may be any combination shown in Figures 39 and 40.

[0198] In addition, the central scFv-type Fc domain is generally asymmetric mutant (for example, the amino acid substitution sets shown in Figures 3 and 8, particularly useful asymmetric mutants include S364K / E357Q:L368D / K370S, L368D / K370S:S364K, L368E / K370S:S364K, T411T / E360E / Q362E:D401K, L368D / K370S:S364K / E357L (Selected from the group consisting of K370S:S364K / E357Q, T366S / L368A / Y407V:T366W, and T366S / L368A / Y407V / Y349C:T366W / S354C), optionally including cleavage mutants (including those shown in Figure 5), optionally including charged scFv linkers (including those shown in Figure 7), and the heavy chain includes pI mutants (including those shown in Figure 4).

[0199] In some embodiments, the central scFv type includes asymmetric mutants, pI mutants, and cleavage mutants. Thus, some embodiments include a) a first monomer comprising the asymmetric mutant S364K / E357Q, the cleavage mutant E233P / L234V / L235A / G236del / S267K, and a first variable weight domain and a second variable weight domain that, together with a first variable light domain of the light chain, form an Fv that binds to a first checkpoint inhibitor, and b) the asymmetric mutant L368D / K370S, the pI mutant N208D / Q295E / N384D / Q418E / N421 D, a second monomer comprising a cleavage mutant E233P / L234V / L235A / G236del / S267K, and a first variable heavy domain that, together with a first variable light domain, makes up an Fv that binds to a first checkpoint inhibitor outlined herein, and a second variable light chain that, together with a second variable heavy chain, forms an Fv(ABD) that binds to a second checkpoint inhibitor; and c) a bottle opener type comprising a light chain comprising a first variable light domain and a constant light domain. In this embodiment, preferred pairs of Fv include CTLA-4XPD-1, PD-1XCTLA-4, LAG-3XPD-1, BTLAXPD-1, and LAG-3XCTLA-4 (Fab is described first, then scFv).

[0200] In some embodiments, the central scFv type includes asymmetric mutants, pI mutants, cleavage mutants, and FcRn mutants. Therefore, some embodiments include a) a first monomer comprising an asymmetric mutant S364K / E357Q, a cleavage mutant E233P / L234V / L235A / G236del / S267K, an FcRn mutant M428L / N434S, and a first variable weight domain and a second variable weight domain that, together with a first variable light domain of the light chain, form an Fv that binds to a first checkpoint inhibitor; and b) an asymmetric mutant L368D / K370S, a pI mutant N208D / Q295E / N384D / Q418E / N421D, a cleavage mutant E233P / L234V / L235A / G236del / S267K, an FcRn mutant M428L / N434S, and a first variable light domain c) a bottle opener type comprising a second monomer including a first variable heavy domain which, together with the first variable heavy chain, makes up an Fv that binds to the first checkpoint inhibitor outlined herein, and a second variable light chain which, together with the second variable heavy chain, forms an Fv(ABD) that binds to the second checkpoint inhibitor, and c) a light chain including the first variable light domain and a constant light domain. In this embodiment, preferred pairs of Fv include CTLA-4XPD-1, PD-1XCTLA-4, LAG-3XPD-1, BTLAXPD-1, and LAG-3XCTLA-4 (Fab is described first, then scFv).

[0201] Regarding central scFv sequences (including M428L / N434S optionally) that are similar to / utilize the bottle opener skeleton 1 in Figure 37, specific combinations of Fv use in the present invention include CTLA-4(Fab)XPD-1(scFv), PD-1(Fab)XCTLA-4(scFv), LAG-3(Fab)XPD-1(scFv), BTLA(Fab)XPD-1(scFv), and LAG-3(Fab)XCTLA-4(scFv).

[0202] Regarding central scFv sequences (including M428L / N434S, optionally selected) that are similar to / utilize the bottle opener skeleton 1 in Figure 37, specific ABDs that bind to human PD-1 include, but are not limited to, 1G6_H1.279_L1.194, 1G6_H1.280_L1.224, 1G6_L1.194_H1.279, 1G6_L1.210_H1.288, and 2E9_H1L1, as well as those listed in sequence numbers 6209-11464, 11465-17134, 33003-33072, 33073-35394, and 36127-36146.

[0203] Regarding central scFv sequences (including M428L / N434S, optionally selected) that are similar to / utilize the bottle opener skeleton 1 in Figure 37, specific ABDs that bind to human CTLA-4 are not limited to [CTLA-4]_H0.25_L0, [CTLA-4]_H0.26_L0, [CTLA-4]_H0.27_L0, [CTLA-4]_H0.29_L0, [CTLA-4]_H0.38_L0, [CTLA-4]_H0.39_L0, [CTLA-4]_H0.40_L0, and [CTLA-4]_H0.70. _L0, [CTLA-4]_H0_L0.22, [CTLA-4]_H2_L0, [CTLA-4]_H3.21_L0.124, [CTLA-4]_H3.21_L0.129, [CTLA-4]_H3.21_L0.132, [CTLA-4]_H3. 23_L0.124, [CTLA-4]_H3.23_L0.129, [CTLA-4]_H3.23_L0.132, [CTLA-4]_H3.25_L0.124, [CTLA-4]_H3.25_L0.129, [CTLA-4]_H3.25_L0. 132, [CTLA-4]_H3.4_L0.118, [CTLA-4]_H3.4_L0.119, [CTLA-4]_H3.4_L0.12, [CTLA-4]_H3.4_L0.121, [CTLA-4]_H3.4_L0.122, [CTLA-4 ]_H3.4_L0.123, [CTLA-4]_H3.4_L0.124, [CTLA-4]_H3.4_L0.125, [CTLA-4]_H3.4_L0.126, [CTLA-4]_H3.4_L0.127, [CTLA-4]_H3.4_L0.1 28, [CTLA-4]_H3.4_L0.129, [CTLA-4]_H3.4_L0.130, [CTLA-4]_H3.4_L0.131, [CTLA-4]_H3.4_L0.132, [CTLA-4]_H3.5_L2.1, [CTLA-4]_ H3.5_L2.2, [CTLA-4]_H3.5_L2.3, [CTLA-4]_H3_L0, [CTLA-4]_H3_L0.22, [CTLA-4]_H3_L0.44, [CTLA-4]_H3_L0.67, and [CTLA-4]_H3_L0.This includes items 74, as well as those listed in sequence numbers 21-2918, 2919-6208, 36739-36818, and 35395-35416.

[0204] Regarding central scFv sequences (including M428L / N434S, optionally selected) that are similar to / utilize the bottle opener skeleton 1 in Figure 37, specific ABDs that bind to human LAG-3 are not limited to, but include 2A11_H0L0, 2A11_H1.125_L2.113, 2A11_H1.144_L2.142, and 2A11_H1 _L2.122, 2A11_H1_L2.123, 2A11_H1_L2.124, 2A11_H1_L2.25, 2A11_H1_L2.47, 2A11_H 1_L2.50, 2A11_H1_L2.91, 2A11_H1_L2.93, 2A11_H1_L2.97, 2A11_H1L1, 2A11_H1L2, 2A1 1_H2L2, 2A11_H3L1, 2A11_H3L2, 2A11_H4L1, 2A11_H4L2, 7G8_H0L0, 7G8_H1L1, 7G8_H3. 18_L1.11, 7G8_H3.23_L1.11, 7G8_H3.28_L1, 7G8_H3.28_L1.11, 7G8_H3.28_L1.13, 7G8 Examples include those listed as _H3.30_L1.34, 7G8_H3.30_L1.34, and 7G8_H3L1, as well as those listed as sequence numbers 17135-20764, 36819-36962, 35417-35606, 25194-32793, and 32794-33002.

[0205] Regarding central scFv sequences (including M428L / N434S, optionally selected) that are similar to / utilize the bottle opener skeleton 1 in Figure 37, specific ABDs that bind to human BTLA include, but are not limited to, 9C6_H0L0, 9C6_H1.1_L1, and 9C6_H1.11_L1, as well as those listed in sequence numbers 20885-21503 and 36707-36738.

[0206] Regarding central scFv sequences (including M428L / N434S, optionally selected) that are similar to / utilize the bottle opener skeleton 1 in Figure 37, specific ABDs that bind to human TIM-3 include, but are not limited to, 1D10_H0L0, 1D12_H0L0, 3H3_H1_L2.1, 6C8_H0L0, 6D9_H0_1D12_L0, 7A9_H0L0, 7B11_H0L0, 7B11var_H0L0, and 7C2_H0L0, as well as those listed in sequence numbers 20765~20884, 37587~37698, and 36347~36706.

[0207] E. Center Fv type One heterodimer scaffold particularly useful in the present invention is the central Fv type shown in Figure 1G. In this embodiment, the type relies on the use of an inserted scFv domain, thus forming a third antigen-binding domain, where the Fab portions of the two monomers bind to one checkpoint target and the "extra" scFv domain binds to the other checkpoint target. The scFv domain is inserted between the Fc domain and the CH1-Fv region of the monomer, thus providing a third antigen-binding domain, and each monomer contains components of the scFv (for example, one monomer contains a variable weight domain and the other contains a variable light domain).

[0208] In this embodiment, one monomer comprises a first heavy chain containing a first variable heavy domain, a CH1 domain, and an Fc domain, as well as an additional variable light domain. The light domain is covalently bonded between the C-terminus of the CH1 domain of the heavy steady domain and the N-terminus of the first Fc domain using a domain linker (vh1-CH1-[optional linker]-vh2-hinge-CH2-CH3). The other monomer comprises a first heavy chain containing a first variable heavy domain, a CH1 domain, and an Fc domain, as well as an additional variable heavy domain (vh1-CH1-[optional linker]-vh2-hinge-CH2-CH3). The light domain is covalently bonded between the C-terminus of the CH1 domain of the heavy steady domain and the N-terminus of the first Fc domain using a domain linker. This embodiment further utilizes a common light chain containing a variable light domain and a steady light domain, which is bonded to two TTAs. The heavy chain is involved in forming the same Fab. For many of the embodiments herein, these constructs include asymmetric mutants, pI mutants, cleavage mutants, additional Fc mutants, etc., as desired and described herein. In this embodiment, preferred pairs of Fv include (Fab first, then scFv) PD-1 and CTLA-4, CTLA-4 and PD-1, PD-1 and TIM-3, TIM-3 and PD-1, PD-1 and LAG-3, LAG-3XPD1, PD-1 and TIGIT, TIGIT and PD-1, PD-1 and BTLA, BTLA and PD-1, CTLA-4 and TIM-3, TIM-3 and CTLA-4, CTLA-4 and LAG-3, LAG-3 and CTLA-4, CTLA-4 and TIGIT, TIGIT and CTLA-4, CTLA-4 and BTLA, BTLA and CTLA-4, TIM-3 and LAG-3, LAG-3 and TIM-3, TIM-3 and TIGIT, TIGIT and TIM-3, TIM-3 and BTLA, BTLA and TIM-3. LAG-3 and TIGIT, TIGIT and LAG-3, LAG-3 and BTLA, BTLA and LAG-3, BTLA and TIGIT, and TIGIT and BTLA.

[0209] The ABD sequences of these combinations are disclosed in the sequence listing or are shown in Figures 9-13, and may be any combination shown in Figures 39 and 40.

[0210] In the central scFv type, specific combinations of Fv usage in the present invention include CTLA-4(Fab)XPD-1(scFv), PD-1(Fab)XCTLA-4(scFv), LAG-3(Fab)XPD-1(scFv), BTLA(Fab)XPD-1(scFv), and LAG-3(Fab)XCTLA-4(scFv).

[0211] In the central scFv type, specific ABDs that bind to human PD-1 include, but are not limited to, 1G6_H1.279_L1.194, 1G6_H1.280_L1.224, 1G6_L1.194_H1.279, 1G6_L1.210_H1.288, and 2E9_H1L1, as well as those listed in sequence numbers 6209-11464, 11465-17134, 33003-33072, 33073-35394, and 36127-36146.

[0212] In the central scFv type, specific ABDs that bind to human CTLA-4 are not limited to [CTLA-4]_H0.25_L0, [CTLA-4]_H0.26_L0, [CTLA-4]_H0.27_L0, [CTLA-4]_H0.29_L0, [CTLA-4]_H0.38_L0, [CTLA-4]_H0.39_L0, [CTLA-4]_H0.40_L0, [CTLA-4]_H0.70_L0, CTLA-4]_H0_L0.22, [CTLA-4]_H2_L0, [CTLA-4]_H3.21_L0.124, [CTLA-4]_H3.21_L0.129, [CTLA-4]_H3.21_L0.1 32, [CTLA-4]_H3.23_L0.124, [CTLA-4]_H3.23_L0.129, [CTLA-4]_H3.23_L0.132, [CTLA-4]_H3.25_L0.124, [CTL A-4]_H3.25_L0.129, [CTLA-4]_H3.25_L0.132, [CTLA-4]_H3.4_L0.118, [CTLA-4]_H3.4_L0.119, [CTLA-4]_H3.4 _L0.12, [CTLA-4]_H3.4_L0.121, [CTLA-4]_H3.4_L0.122, [CTLA-4]_H3.4_L0.123, [CTLA-4]_H3.4_L0.124, [CTL A-4]_H3.4_L0.125, [CTLA-4]_H3.4_L0.126, [CTLA-4]_H3.4_L0.127, [CTLA-4]_H3.4_L0.128, [CTLA-4]_H3.4_L 0.129, [CTLA-4]_H3.4_L0.130, [CTLA-4]_H3.4_L0.131, [CTLA-4]_H3.4_L0.132, [CTLA-4]_H3.5_L2.1, [CTLA-4 Examples include [CTLA-4]_H3.5_L2.2, [CTLA-4]_H3.5_L2.3, [CTLA-4]_H3_L0, [CTLA-4]_H3_L0.22, [CTLA-4]_H3_L0.44, [CTLA-4]_H3_L0.67, and [CTLA-4]_H3_L0.74, as well as those listed in sequence numbers 21-2918, 2919-6208, 36739-36818, and 35395-35416.

[0213] In the central scFv type, specific ABDs that bind to human LAG-3 include, but are not limited to, 2A11_H0L0, 2A11_H1.125_L2.113, 2A11_H1.144_L2.142, 2A11_H1_L2.122, 2A11_H1_L2.123, 2A11_H1_L2.124, 2A11_H1_L2.25, 2A11_H1_L2.47, 2A11_H1_L2.50, 2A11_H1_L2.91, 2A11_H1_L2.93, 2A11_H1_L2.97, 2A11_H1L1, 2A11_H1L2, 2A11_H2L2, 2A11_H3L1, 2A11_H3L2, 2A11_H4L1, 2A11_H4L2, 7G8_H0L0, 7G8_H1L1, 7G8_H3.18_L1.11, 7G8_H3.23_L1.11, 7G8_H3.28_L1, 7G8_H3.28_L1.11, 7G8_H3.28_L1.13, 7G8_H3.30_L1.34, 7G8_H3.30_L1.34, and 7G8_H3L1, as well as those listed in the lists of SEQ ID NO: 17135-20764, SEQ ID NO: 36819-36962, SEQ ID NO: 35417-35606, SEQ ID NO: 25194-32793, and SEQ ID NO: 32794-33002.

[0214] In the central scFv type, specific ABDs that bind to human BTLA include, but are not limited to, 9C6_H0L0; 9C6 H1.1_L1 and 9C6_H1.11_L1, as well as those listed in the lists of SEQ ID NO: 20885-21503 and SEQ ID NO: 36707-36738.

[0215] In the central scFv type, specific ABDs that bind to human TIM-3 include, but are not limited to, 1D10_H0L0, 1D12_H0L0, 3H3_H1_L2.1, 6C8_H0L0, 6D9_H0_1D12_L0, 7A9_H0L0, 7B11_H0L0, 7B11var_H0L0, and 7C2_H0L0, as well as those listed in the lists of SEQ ID NO: 20765-20884, SEQ ID NO: 37587-37698, and SEQ ID NO: 36347-36706.

[0216] F. Monovalent-centered scFv One heterodimeric scaffold particularly useful in the present invention is the monovalent-centered scFv type shown in Figure 1C. In this embodiment, one monomer contains only the Fc domain, and the other monomer uses an inserted scFv domain and thus forms a second antigen-binding domain. In this type, either Fab portion binds to one checkpoint target and the scFv binds to another checkpoint target. The scFv domain is inserted between the CH1-Fv region and one of the Fc domains of the monomer.

[0217] In this embodiment, one monomer includes a first heavy chain having a first variable heavy domain containing an scFv comprising an scFv variable light domain, an scFv linker, and an scFv variable heavy domain, a CH1 domain, and an Fc domain. The scFv is covalently linked between the C-terminus of the CH1 domain of the heavy constant domain and the N-terminus of the first Fc domain using a domain linker. The second monomer contains an Fc domain. This embodiment further utilizes a light chain comprising a variable light domain and a constant light domain, which is associated with the heavy chain to form a Fab. For many of the embodiments herein, these constructs include asymmetric variants, pI variants, cleavage variants, additional Fc variants, etc. as desired and described herein This includes. In this embodiment, preferred pairs of Fv include (Fab first, then scFv) PD-1 and CTLA-4, CTLA-4 and PD-1, PD-1 and TIM-3, TIM-3 and PD-1, PD-1 and LAG-3, LAG-3XPD1, PD-1 and TIGIT, TIGIT and PD-1, PD-1 and BTLA, BTLA and PD-1, CTLA-4 and TIM-3, TIM-3 and CTLA-4, CTLA-4 and LAG-3, LAG-3 and CTLA-4, CTLA-4 and TIGIT, TIGIT and CTLA-4, CTLA-4 and BTLA, BTLA and CTLA-4, TIM-3 and LAG-3, LAG-3 and TIM-3, TIM-3 and TIGIT, TIGIT and TIM-3, TIM-3 and BTLA, BTLA and TIM-3. LAG-3 and TIGIT, TIGIT and LAG-3, LAG-3 and BTLA, BTLA and LAG-3, BTLA and TIGIT, and TIGIT and BTLA.

[0218] The ABD sequences of these combinations are disclosed in the sequence listing or are shown in Figures 9-13, and may be any combination shown in Figures 39 and 40.

[0219] In addition, the single-arm central scFv type Fc domain is generally asymmetric mutant (for example, the amino acid substitution sets shown in Figures 3 and 8, particularly useful asymmetric mutants include S364K / E357Q:L368D / K370S, L368D / K370S:S364K, L368E / K370S:S364K, T411T / E360E / Q362E:D401K, L368D / K370S:S364K / E357L, K370S:S364K / E357Q, T366S / L368A / Y407V:T366W, and T366S / L368A / Y407V The molecule is selected from the group consisting of Y349C:T366W / S354C, optionally includes a cleavage mutant (including those shown in Figure 5), optionally includes a charged scFv linker (including those shown in Figure 7), and the heavy chain includes a pI mutant (including those shown in Figure 4).

[0220] In some embodiments, the one-armed central scFv type includes asymmetric mutants, pI mutants, and cleavage mutants. Thus, some embodiments include a) a first monomer comprising the asymmetric mutant S364K / E357Q, the cleavage mutant E233P / L234V / L235A / G236del / S267K, and a first variable weight domain and a second variable weight domain that, together with a first variable light domain of the light chain, form an Fv that binds to a first checkpoint inhibitor, and b) the asymmetric mutant L368D / K370S, the pI mutant N208D / Q295E / N384D / Q418E / N421 D, a second monomer comprising a cleavage mutant E233P / L234V / L235A / G236del / S267K, and a first variable heavy domain that, together with a first variable light domain, makes up an Fv that binds to a first checkpoint inhibitor outlined herein, and a second variable light chain that, together with a second variable heavy chain, forms an Fv(ABD) that binds to a second checkpoint inhibitor; and c) a bottle opener type comprising a light chain comprising a first variable light domain and a constant light domain. In this embodiment, preferred pairs of Fv include CTLA-4XPD-1, PD-1XCTLA-4, LAG-3XPD-1, BTLAXPD-1, and LAG-3XCTLA-4 (Fab is described first, then scFv).

[0221] In some embodiments, the one-armed central scFv type includes asymmetric mutants, pI mutants, cleavage mutants, and FcRn mutants. Thus, some embodiments include a) a first monomer comprising the asymmetric mutant S364K / E357Q, the cleavage mutant E233P / L234V / L235A / G236del / S267K, the FcRn mutant M428L / N434S, and a first variable weight domain and a second variable weight domain that, together with a first variable light domain of the light chain, form an Fv that binds to a first checkpoint inhibitor, and b) the asymmetric mutant L368D / K370S, the pI mutant N208D / Q295E / N384D / Q418E / N421D, and the cleavage mutant E233P / L234V / L235A / G236d c) a bottle opener type comprising a second monomer comprising a first variable heavy domain which, together with a first variable light domain, forms an Fv that binds to a first checkpoint inhibitor outlined herein, and a second variable light chain which, together with a second variable heavy chain, forms an Fv(ABD) that binds to a second checkpoint inhibitor, and c) a light chain comprising a first variable light domain and a constant light domain. In this embodiment, preferred pairs of Fv include CTLA-4XPD-1, PD-1XCTLA-4, LAG-3XPD-1, BTLAXPD-1, and LAG-3XCTLA-4 (Fab is described first, then scFv).

[0222] In uniarm-centered scFv type, specific ABDs that bind to human PD-1 include, but are not limited to, 1G6_H1.279_L1.194, 1G6_H1.280_L1.224, 1G6_L1.194_H1.279, 1G6_L1.210_H1.288, and 2E9_H1L1, as well as those listed in sequence numbers 6209-11464, 11465-17134, 33003-33072, 33073-35394, and 36127-36146.

[0223] In uniarm-centered scFv type CTLA-4, specific ABDs that bind to human CTLA-4 are not limited to [CTLA-4]_H0.25_L0, [CTLA-4]_H0.26_L0, [CTLA-4]_H0.27_L0, [CTLA-4]_H0.29_L0, [CTLA-4]_H0.38_L0, [CTLA-4]_H0.39_L0, [CTLA-4]_H0.40_L0, [CTLA-4]_H0.70_L0, [CTLA-4]_H0_L0.22, [CTLA-4]_H2_L0, and [CTLA-4]_H3.21_L0. 124, [CTLA-4]_H3.21_L0.129, [CTLA-4]_H3.21_L0.132, [CTLA-4]_H3.23_L0.124, [CTLA-4]_H3.23_L0.129, [CTLA-4]_H3.23_L0.132, [CTLA- 4]_H3.25_L0.124, [CTLA-4]_H3.25_L0.129, [CTLA-4]_H3.25_L0.132, [CTLA-4]_H3.4_L0.118, [CTLA-4]_H3.4_L0.119, [CTLA-4]_H3.4_L0.12 , [CTLA-4]_H3.4_L0.121, [CTLA-4]_H3.4_L0.122, [CTLA-4]_H3.4_L0.123, [CTLA-4]_H3.4_L0.124, [CTLA-4]_H3.4_L0.125, [CTLA-4]_H3.4_ L0.126, [CTLA-4]_H3.4_L0.127, [CTLA-4]_H3.4_L0.128, [CTLA-4]_H3.4_L0.129, [CTLA-4]_H3.4_L0.130, [CTLA-4]_H3.4_L0.131, [CTLA-4]_ Examples include H3.4_L0.132, [CTLA-4]_H3.5_L2.1, [CTLA-4]_H3.5_L2.2, [CTLA-4]_H3.5_L2.3, [CTLA-4]_H3_L0, [CTLA-4]_H3_L0.22, [CTLA-4]_H3_L0.44, [CTLA-4]_H3_L0.67, and [CTLA-4]_H3_L0.74, as well as those listed in sequence numbers 21-2918, 2919-6208, 36739-36818, and 35395-35416.

[0224] In uniarm-centered scFv type, specific ABDs that bind to human LAG-3 are not limited to, but include 2A11_H0L0, 2A11_H1.125_L2.113, 2A11_H1.144_L2.142, 2A11_H1_L2.122, 2A11_H1_L2.123, 2A11_H1_L2.124, 2A11_H1_L2.25, and 2A11_H1 _L2.47, 2A11_H1_L2.50, 2A11_H1_L2.91, 2A11_H1_L2.93, 2A11_H1_L2.97, 2A11_H1L1, 2A11_H1 L2, 2A11_H2L2, 2A11_H3L1, 2A11_H3L2, 2A11_H4L1, 2A11_H4L2, 7G8_H0L0, 7G8_H1L1, 7G8_H3.18 Examples include those listed as _L1.11, 7G8_H3.23_L1.11, 7G8_H3.28_L1, 7G8_H3.28_L1.11, 7G8_H3.28_L1.13, 7G8_H3.30_L1.34, 7G8_H3.30_L1.34, and 7G8_H3L1, as well as those listed as sequence numbers 17135-20764, 36819-36962, 35417-35606, 25194-32793, and 32794-33002.

[0225] In uniarm-centered scFv type, specific ABDs that bind to human BTLA include, but are not limited to, 9C6_H0L0, 9C6_H1.1_L1, and 9C6_H1.11_L1, as well as those listed in sequence numbers 20885-21503 and 36707-36738.

[0226] In the uniarm-centered scFv type, specific ABDs that bind to human TIM-3 include, but are not limited to, 1D10_H0L0, 1D12_H0L0, 3H3_H1_L2.1, 6C8_H0L0, 6D9_H0_1D12_L0, 7A9_H0L0, 7B11_H0L0, 7B11var_H0L0, and 7C2_H0L0, as well as those listed in sequence numbers 20765-20884, 37587-37698, and 36347-36706.

[0227] G. One-arm scFv-mAb One heterodimer scaffold particularly useful in the present invention is the one-armed scFv-mAb type shown in Figure 1D. In this embodiment, one monomer contains only an Fc domain, and the other monomer uses an scFv domain that is bound at the N-terminus of the heavy chain, generally by the use of a linker:vh-scFv linker-vl-[optional domain linker]-CH1-hinge-CH2-CH3 or (in the reverse orientation) vl-scFv linker-vh-[optional domain linker]-CH1-hinge-CH2-CH3. In this type, either Fab portion binds to one checkpoint target, and the scFv binds to another checkpoint target. This embodiment further utilizes a light chain containing a variable light domain and a constant light domain, which is related to the heavy chain to form the Fab. For many of the embodiments herein, these constructs include asymmetric mutants, pI mutants, cleavage mutants, additional Fc mutants, etc., as desired and described herein. In this embodiment, preferred pairs of Fv include (Fab first, then scFv) PD-1 and CTLA-4, CTLA-4 and PD-1, PD-1 and TIM-3, TIM-3 and PD-1, PD-1 and LAG-3, LAG-3XPD1, PD-1 and TIGIT, TIGIT and PD-1, PD-1 and BTLA, BTLA and PD-1, CTLA-4 and TIM-3, TIM-3 and CTLA-4, CTLA-4 and LAG-3, LAG-3 and CTLA-4, CTLA-4 and TIGIT, TIGIT and CTLA-4, CTLA-4 and BTLA, BTLA and CTLA-4, TIM-3 and LAG-3, LAG-3 and TIM-3, TIM-3 and TIGIT, TIGIT and TIM-3, TIM-3 and BTLA, BTLA and TIM-3. LAG-3 and TIGIT, TIGIT and LAG-3, LAG-3 and BTLA, BTLA and LAG-3, BTLA and TIGIT, and TIGIT and BTLA.

[0228] The ABD sequences of these combinations are disclosed in the sequence listing or are shown in Figures 9-13, and may be any combination shown in Figures 39 and 40.

[0229] In addition, the Fc domain has asymmetric mutants (e.g., the amino acid substitution sets shown in Figures 3 and 8, with particularly useful asymmetric mutants being S364K / E357Q:L368D / K370S, L368D / K370S:S364K, L368E / K370S:S364K) (Selected from the group consisting of T411T / E360E / Q362E:D401K, L368D / K370S:S364K / E357L, K370S:S364K / E357Q, T366S / L368A / Y407V:T366W, and T366S / L368A / Y407V / Y349C:T366W / S354C), optionally include cleavage mutants (including those shown in Figure 5), optionally include charged scFv linkers (including those shown in Figure 7), and the heavy chain includes pI mutants (including those shown in Figure 4).

[0230] In some embodiments, the single-arm scFv-mAb type includes asymmetric mutants, pI mutants, and cleavage mutants. Thus, some embodiments include a) a first monomer comprising the asymmetric mutant S364K / E357Q, the cleavage mutant E233P / L234V / L235A / G236del / S267K, and a first variable weight domain and a second variable weight domain that, together with a first variable light domain of the light chain, form an Fv that binds to a first checkpoint inhibitor, and b) the asymmetric mutant L368D / K370S, the pI mutant N208D / Q295E / N384D / Q418E / N421 D, a second monomer comprising a cleavage mutant E233P / L234V / L235A / G236del / S267K, and a first variable heavy domain that, together with a first variable light domain, makes up an Fv that binds to a first checkpoint inhibitor outlined herein, and a second variable light chain that, together with a second variable heavy chain, forms an Fv(ABD) that binds to a second checkpoint inhibitor; and c) a bottle opener type comprising a light chain comprising a first variable light domain and a constant light domain. In this embodiment, preferred pairs of Fv include CTLA-4XPD-1, PD-1XCTLA-4, LAG-3XPD-1, BTLAXPD-1, and LAG-3XCTLA-4 (Fab is described first, then scFv).

[0231] In some embodiments, the single-arm scFv-mAb type includes asymmetric mutants, pI mutants, cleavage mutants, and FcRn mutants. Thus, some embodiments include a) a first monomer comprising the asymmetric mutant S364K / E357Q, the cleavage mutant E233P / L234V / L235A / G236del / S267K, the FcRn mutant M428L / N434S, and a first variable weight domain and a second variable weight domain that, together with the first variable light domain of the light chain, make up an Fv that binds to the first checkpoint inhibitor, and b) the asymmetric mutant L368D / K370S, the pI mutant N208D / Q295E / N384D / Q418E / N421 D, a second monomer comprising a cleavage mutant E233P / L234V / L235A / G236del / S267K, an FcRn mutant M428L / N434S, and a second variable light chain comprising a first variable heavy domain that, together with a first variable light domain, makes up an Fv that binds to a first checkpoint inhibitor outlined herein, and a second variable heavy chain that, together with a second variable heavy chain, forms an Fv(ABD) that binds to a second checkpoint inhibitor; and c) a bottle opener type comprising a light chain comprising a first variable light domain and a constant light domain. In this embodiment, preferred pairs of Fv include CTLA-4XPD-1, PD-1XCTLA-4, LAG-3XPD-1, BTLAXPD-1, and LAG-3XCTLA-4 (Fab is described first, then scFv).

[0232] In uniarm scFv-mAb type, specific ABDs that bind to human PD-1 include, but are not limited to, 1G6_H1.279_L1.194, 1G6_H1.280_L1.224, 1G6_L1.194_H1.279, 1G6_L1.210_H1.288, and 2E9_H1L1, as well as those listed in sequence numbers 6209-11464, 11465-17134, 33003-33072, 33073-35394, and 36127-36146.

[0233] In unilateral scFv-mAb type cytoplasmic receptacles, specific ABDs that bind to human CTLA-4 are not limited to [CTLA-4]_H0.25_L0, [CTLA-4]_H0.26_L0, [CTLA-4]_H0.27_L0, and [CTLA-4]_H0.29_L0. , [CTLA-4]_H0.38_L0, [CTLA-4]_H0.39_L0, [CTLA-4]_H0.40_L0, [CTLA-4]_H0.70_L0, [CTLA-4]_H0_L0.22, [CTLA-4]_H2_L0, [CTLA-4]_H3.21_L0.124, [CTLA-4]_H3.21_L0.129, [CTLA-4]_H3.21_L0.132, [CTLA-4]_H3.23_L0.124, [CTLA-4]_H3.23_L0. 129, [CTLA-4]_H3.23_L0.132, [CTLA-4]_H3.25_L0.124, [CTLA-4]_H3.25_L0.129, [CTLA-4]_H3.25_L0.132, [CTLA-4]_H3.4_ L0.118, [CTLA-4]_H3.4_L0.119, [CTLA-4]_H3.4_L0.12, [CTLA-4]_H3.4_L0.121, [CTLA-4]_H3.4_L0.122, [CTLA-4]_H3.4_L0 .123, [CTLA-4]_H3.4_L0.124, [CTLA-4]_H3.4_L0.125, [CTLA-4]_H3.4_L0.126, [CTLA-4]_H3.4_L0.127, [CTLA-4]_H3.4_L0. 128, [CTLA-4]_H3.4_L0.129, [CTLA-4]_H3.4_L0.130, [CTLA-4]_H3.4_L0.131, [CTLA-4]_H3.4_L0.132, [CTLA-4]_H3.5_L2.1 Examples include [CTLA-4]_H3.5_L2.2, [CTLA-4]_H3.5_L2.3, [CTLA-4]_H3_L0, [CTLA-4]_H3_L0.22, [CTLA-4]_H3_L0.44, [CTLA-4]_H3_L0.67, and [CTLA-4]_H3_L0.74, as well as those listed in sequence numbers 21-2918, 2919-6208, 36739-36818, and 35395-35416.

[0234] In the single-arm scFv-mAb type, specific ABDs that bind to human LAG-3 include, but are not limited to, 2A11_H0L0, 2A11_H1.125_L2.113, 2A11_H1.144_L2.142, 2A11_H1_L2.122, 2A11_H1_L2.123, 2A11_H1_L2.124, 2A11_H1_L2.25, 2A11_H1_L2.47, 2A11_H1_L2.50, 2A11_H1_L2.91, 2A11_H1_L2.93, 2A11_H1_L2.97, 2A11_H1L1, 2A11_H1L2, 2A11_H2L2, 2A11_H3L1, 2A11_H3L2, 2A11_H4L1, 2A11_H4L2, 7G8_H0L0, 7G8_H1L1, 7G8_H3.18_L1.11, 7G8_H3.23_L1.11, 7G8_H3.28_L1, 7G8_H3.28_L1.11, 7G8_H3.28_L1.13, 7G8_H3.30_L1.34, 7G8_H3.30_L1.34, and 7G8_H3L1, as well as those listed in the lists of SEQ ID NO: 17135 to 20764, SEQ ID NO: 36819 to 36962, SEQ ID NO: 35417 to 35606, SEQ ID NO: 25194 to 32793, and SEQ ID NO: 32794 to 33002.

[0235] In the single-arm scFv-mAb type, specific ABDs that bind to human BTLA include, but are not limited to, 9C6_H0L0, 9C6_H1.1_L1, and 9C6_H1.11_L1, as well as those listed in the lists of SEQ ID NO: 20885 to 21503 and SEQ ID NO: 36707 to 36738.

[0236] In the single-arm scFv-mAb type, specific ABDs that bind to human TIM-3 include, but are not limited to, 1D10_H0L0; 1D12_H0L0, 3H3_H1_L$.1, 6C8_H0L0, 6D9_H0_1D12 L0, 7A9_H0L0, 7B11_H0L0, 7B11var_H0L0, and 7C2_H0L0, as well as those listed in the lists of SEQ ID NO: 20765 to 20884, SEQ ID NO: 37587 to 37698, and SEQ ID NO: 36347 to 36706.

[0237] H.scFv-mAb type One heterodimer scaffold particularly useful in the present invention is the mAb-scFv type shown in Figure 1E. In this embodiment, the type relies on the use of an N-terminal binding of scFv to one of the monomers, thus forming a third antigen-binding domain, where the Fab portions of the two monomers bind to one checkpoint target and the "extra" scFv domain binds to a different checkpoint target.

[0238] In this embodiment, the first monomer comprises a first heavy chain (including a variable heavy domain and a constant domain), and the N-terminally covalently bonded scFv is in either orientation ((vh1-scFv linker-vl1-[optional domain linker]-vh2-CH1-hinge-CH2-CH3) or (in the reverse orientation with scFv) ((vl1-scFv linker-vh1-[optional domain linker]-vh2-CH1-hinge-CH2-CH3)), with the scFv variable light domain and scFv The construct includes a linker and a variable heavy domain of the scFv. This embodiment further utilizes a common light chain including a variable light domain and a constant light domain, which is related to the heavy chain to form two identical Fabs that bind to one of the target antigens. For many of the embodiments herein, these constructs include asymmetric variants, pI variants, cleavage variants, additional Fc variants, etc., as desired and described herein. In this embodiment, preferred Fv pairs include (Fab first, then scFv) PD-1 and CTLA-4, CTLA-4 and PD-1, PD-1 and TIM-3, TIM-3 and PD-1, PD-1 and LAG-3, LAG-3XPD1, PD-1 and TIGIT, TIGIT and PD-1, PD-1 and BTLA, BTLA and PD-1, CTLA-4 and TIM-3, TIM-3 and CTLA-4, CTLA-4 and LAG-3, LAG-3 and CTLA-4, CTLA-4 and TIGIT Examples include TIGIT and CTLA-4, CTLA-4 and BTLA, BTLA and CTLA-4, TIM-3 and LAG-3, LAG-3 and TIM-3, TIM-3 and TIGIT, TIGIT and TIM-3, TIM-3 and BTLA, BTLA and TIM-3. LAG-3 and TIGIT, TIGIT and LAG-3, LAG-3 and BTLA, BTLA and LAG-3, BTLA and TIGIT, and TIGIT and BTLA.

[0239] The ABD sequences of these combinations are disclosed in the sequence listing or are shown in Figures 9-13, and may be any combination shown in Figures 39 and 40.

[0240] In addition, the scFv-mAb type Fc domain has asymmetric mutants (for example, the amino acid substitution sets shown in Figures 3 and 8, particularly useful asymmetric mutants include S364K / E357Q:L368D / K370S, L368D / K370S:S364K, L368E / K370S:S364K, T411T / E360E / Q362E:D401K, L368D / K370S:S364K / E357L, The molecule is selected from the group consisting of K370S:S364K / E357Q, T366S / L368A / Y407V:T366W, and T366S / L368A / Y407V / Y349C:T366W / S354C, optionally includes a cleavage mutant (including those shown in Figure 5), optionally includes a charged scFv linker (including those shown in Figure 7), and the heavy chain includes a pI mutant (including those shown in Figure 4).

[0241] In some embodiments, the mAb-scFv type includes asymmetric mutants, pI mutants, and cleavage mutants. Thus, some embodiments include a) a first monomer comprising the asymmetric mutant S364K / E357Q, the cleavage mutant E233P / L234V / L235A / G236del / S267K, and a first variable weight domain and a second variable weight domain that, together with a first variable light domain of the light chain, form an Fv that binds to a first checkpoint inhibitor, and b) the asymmetric mutant L368D / K370S, the pI mutant N208D / Q295E / N384D / Q418E / N421D, and the cleavage mutant E233P / L234 c) a bottle opener type comprising a second monomer including V / L235A / G236del / S267K, a first variable heavy domain which, together with a first variable light domain, makes up an Fv that binds to a first checkpoint inhibitor outlined herein, and a second variable light chain which, together with a second variable heavy chain, forms an Fv(ABD) that binds to a second checkpoint inhibitor; and c) a light chain which includes a first variable light domain and a constant light domain. Specific uses in some embodiments of this type include (Fab-scFv order) CTLA-4XPD-1, LAG-3XPD-1, BTLAXPD-1, and LAG-3XCTLA-4.

[0242] In some embodiments, the mAb-scFv type includes asymmetric mutants, pI mutants, cleavage mutants, and FcRn mutants. Therefore, some embodiments include a) a first monomer comprising the asymmetric mutant S364K / E357Q, the cleavage mutant E233P / L234V / L235A / G236del / S267K, the FcRn mutant M428L / N434S, and a first variable weight domain and a second variable weight domain that, together with a first variable light domain of the light chain, make up an Fv that binds to a first checkpoint inhibitor, and b) the asymmetric mutant L368D / K370S, the pI mutant N208D / Q295E / N384D / Q418E / N421 D) A bottle opener type comprising a second monomer including a first variable heavy domain that, together with a first variable light domain, makes up an Fv that binds to a first checkpoint inhibitor outlined herein, and a second variable light chain that, together with a second variable heavy chain, forms an Fv(ABD) that binds to a second checkpoint inhibitor; and c) a light chain including a first variable light domain and a constant light domain. Specific uses in some embodiments of this type include (Fab-scFv order) CTLA-4XPD-1, LAG-3XPD-1, BTLAXPD-1, and LAG-3XCTLA-4.

[0243] For mAb-scFv type skeleton 1 in Figure 38 (including M428L / N434S, if optional), specific ABDs that bind to human PD-1 include, but are not limited to, 1G6_H1.279_L1.194, 1G6_H1.280_L1.224, 1G6_L1.194_H1.279, 1G6_L1.210_H1.288, and 2E9_H1L1, as well as those listed in sequence numbers 6209-11464, 11465-17134, 33003-33072, 33073-35394, and 36127-36146.

[0244] Regarding mAb-scFv type skeleton 1 in Figure 38 (including M428L / N434S at will), the specific ABDs that bind to human CTLA-4 are not limited to [CTLA-4]_H0.25_L0, [CTLA-4]_H0.26_L0, [CTLA-4]_H0.27_L0, [CTLA-4]_H0.29_L0, and [CTLA-4]_H0.3 8_L0, [CTLA-4]_H0.39_L0, [CTLA-4]_H0.40_L0, [CTLA-4]_H0.70_L0, [CTLA-4]_H0_L0.22, [CT LA-4]_H2_L0, [CTLA-4]_H3.21_L0.124, [CTLA-4]_H3.21_L0.129, [CTLA-4]_H3.21_L0.132, [C TLA-4]_H3.23_L0.124, [CTLA-4]_H3.23_L0.129, [CTLA-4]_H3.23_L0.132, [CTLA-4]_H3.25_L 0.124, [CTLA-4]_H3.25_L0.129, [CTLA-4]_H3.25_L0.132, [CTLA-4]_H3.4_L0.118, [CTLA-4]_ H3.4_L0.119, [CTLA-4]_H3.4_L0.12, [CTLA-4]_H3.4_L0.121, [CTLA-4]_H3.4_L0.122, [CTLA- 4]_H3.4_L0.123, [CTLA-4]_H3.4_L0.124, [CTLA-4]_H3.4_L0.125, [CTLA-4]_H3.4_L0.126, [C TLA-4]_H3.4_L0.127, [CTLA-4]_H3.4_L0.128, [CTLA-4]_H3.4_L0.129, [CTLA-4]_H3.4_L0.130, [CTLA- 4]_H3.4_L0.131, [CTLA-4]_H3.4_L0.132, [CTLA-4]_H3.5_L2.1, [CTLA-4]_H3.5_L2.2, [CTLA-4]_H3.5_L 2.3. Examples include [CTLA-4]_H3_L0, [CTLA-4]_H3_L0.22, [CTLA-4]_H3_L0.44, [CTLA-4]_H3_L0.67, and [CTLA-4]_H3_L0.74, as well as those listed in sequence numbers 21-2918, 2919-6208, 36739-36818, and 35395-35416.

[0245] Regarding the mAb-scFv type skeleton 1 in Figure 38 (including M428L / N434S at will), the specific ABDs that bind to human LAG-3 are not limited to, but include 2A11_H0L0, 2A11_H1.125_L2.113, 2A11_H1.144_L2.142, 2A11_H1_L2.122, and 2A11_ H1_L2.123, 2A11_H1_L2.124, 2A11_H1_L2.25, 2A11_H1_L2.47, 2A11_H1_L2.50, 2A1 1_H1_L2.91, 2A11_H1_L2.93, 2A11_H1_L2.97, 2A11_H1L1, 2A11_H1L2, 2A11_H2L2, 2A 11_H3L1, 2A11_H3L2, 2A11_H4L1, 2A11_H4L2, 7G8_H0L0, 7G8_H1L1, 7G8_H3.18_L1.1 1, 7G8_H3.23_L1.11, 7G8_H3.28_L1, 7G8_H3.28_L1.11, 7G8_H3.28_L1.13, 7G8_H3.3 Examples include 0_L1.34, 7G8_H3.30_L1.34, and 7G8_H3L1, as well as those listed in sequence numbers 17135-20764, 36819-36962, 35417-35606, 25194-32793, and 32794-33002.

[0246] For mAb-scFv type skeleton 1 in Figure 38 (including M428L / N434S, optionally selected), specific ABDs that bind to human BTLA include, but are not limited to, 9C6_H0L0, 9C6_H1.1_L1, and 9C6_H1.11_L1, as well as those listed in sequence numbers 20885-21503 and 36707-36738.

[0247] For mAb-scFv type skeleton 1 in Figure 38 (including M428L / N434S, if optional), specific ABDs that bind to human TIM-3 include, but are not limited to, 1D10_H0L0, 1D12_H0L0, 3H3_H1_L2.1, 6C8_H0L0, 6D9_H0 1D12_L0, 7A9_H0L0, 7B11_H0L0, 7B11var_H0L0, and 7C2_H0L0, as well as those listed in sequence numbers 20765-20884, 37587-37698, and 36347-36706.

[0248] I. Dual scFv type The present invention also provides a bi-scFv type known in the art and shown in Figure 1B. In this embodiment, the heterodimer bispecific antibody consists of two scFv-Fc monomers (both of either the (vh-scFv linker-vl-[optional domain linker]-CH2-CH3) type or the (vl-scFv linker-vh-[optional domain linker]-CH2-CH3) type, or one monomer in one orientation and the other in the other orientation.

[0249] In this case, all ABDs are of type scFv, and include PD-1 and CTLA-4, PD-1 and TIM-3, PD-1 and LAG-3, PD-1 and TIGIT, PD-1 and Any combination of BTLA, CTLA-4 and TIM-3, CTLA-4 and LAG-3, CTLA-4 and TIGIT, CTLA-4 and BTLA, TIM-3 and LAG-3, TIM-3 and TIGIT, TIM-3 and BTLA, LAG-3 and TIGIT, LAG-3 and BTLA, or TIGIT and BTLA is useful. The ABD sequences of these combinations are disclosed in the sequence listing or shown in Figures 9-13, and may be any combination shown in Figures 39 and 40.

[0250] In addition, the double scFv type Fc domain has asymmetric mutants (for example, the amino acid substitution sets shown in Figures 3 and 8, particularly useful asymmetric mutants include S364K / E357Q:L368D / K370S, L368D / K370S:S364K, L368E / K370S:S364K, T411T / E360E / Q362E:D401K, L368D / K370S:S364K / E357L, K The group consists of 370S:S364K / E357Q, T366S / L368A / Y407V:T366W, and T366S / L368A / Y407V / Y349C:T366W / S354C, optionally includes cleavage mutants (including those shown in Figure 5), optionally includes charged scFv linkers (including those shown in Figure 7), and the heavy chain includes pI mutants (including those shown in Figure 4).

[0251] In some embodiments, the double scFv type includes asymmetric mutants, pI mutants, and cleavage mutants. Thus, some embodiments include a) a first monomer comprising the asymmetric mutant S364K / E357Q, the cleavage mutant E233P / L234V / L235A / G236del / S267K, and a first variable weight domain and a second variable weight domain that, together with a first variable light domain of the light chain, form an Fv that binds to a first checkpoint inhibitor, and b) the asymmetric mutant L368D / K370S, the pI mutant N208D / Q295E / N384D / Q418E / N421 D) A bottle opener type comprising a second monomer including a cleavage mutant E233P / L234V / L235A / G236del / S267K, a first variable heavy domain which, together with a first variable light domain, makes up an Fv that binds to a first checkpoint inhibitor outlined herein, and a second variable light chain which, together with a second variable heavy chain, forms an Fv(ABD) that binds to a second checkpoint inhibitor, and c) a light chain including a first variable light domain and a constant light domain. Specific uses in some embodiments of this type include (Fab-scFv order) CTLA-4XPD-1, LAG-3XPD-1, BTLAXPD-1, and LAG-3XCTLA-4.

[0252] In some embodiments, the double scFv type includes asymmetric mutants, pI mutants, cleavage mutants, and FcRn mutants. Thus, some embodiments include a) a first monomer comprising the asymmetric mutant S364K / E357Q, the cleavage mutant E233P / L234V / L235A / G236del / S267K, the FcRn mutant M428L / N434S, and a first variable weight domain and a second variable weight domain that, together with a first variable light domain of the light chain, form an Fv that binds to a first checkpoint inhibitor, and b) the asymmetric mutant L368D / K370S, the pI mutant N208D / Q295E / N384D / Q418E / N421 D) A bottle opener type comprising a second monomer including a first variable heavy domain that, together with a first variable light domain, makes up an Fv that binds to a first checkpoint inhibitor outlined herein, and a second variable light chain that, together with a second variable heavy chain, forms an Fv(ABD) that binds to a second checkpoint inhibitor; and c) a light chain including a first variable light domain and a constant light domain. Specific uses in some embodiments of this type include (Fab-scFv order) CTLA-4XPD-1, LAG-3XPD-1, BTLAXPD-1, and LAG-3XCTLA-4.

[0253] J. Non-heterodimeric bispecific antibody As those skilled in the art will understand, the Fv sequences outlined herein can be used in both monospecific antibodies (e.g., "conventional monoclonal antibodies") and non-heterodimerial bispecific antibodies.

[0254] Preferred non-heterodimerary bispecific types are known in the art and generally include several different types shown in Spiess et al., Molecular Immunology (67):95-106 (2015) and Kontermann, mAbs 4:2,182-197 (2012) (both of which are clearly incorporated by reference, particularly with respect to figures, legends, and citations for the types therein).

[0255] K monospecific monoclonal antibody As those skilled in the art will understand, the novel Fv sequences outlined herein can be used in both monospecific antibodies (e.g., "conventional monoclonal antibodies") and non-heterodimal bispecific antibodies. Accordingly, the present invention generally provides monoclonal (monospecific) antibodies comprising six CDRs, and / or the vh and vl sequences shown in the figure, having IgG1, IgG2, IgG3, or IgG4 constant regions, with IgG1, IgG2, and IgG4 (including the IgG4 constant region with an S228P amino acid substitution) being particularly useful in some embodiments. That is, any sequence herein having the "H_L" designation can be ligated to the constant region of a human IgG1 antibody.

[0256] VI. Antigen-targeting antigen-binding domains The bispecific antibodies of the present invention are generally bivalent or trivalent, as shown in Figure 1, and have two distinct antigen-binding domains (ABDs) that bind to two different target checkpoint antigens ("target pairs"). Suitable target checkpoint antigens include human (and sometimes cyno) PD-1, CTLA-4, TIM-3, LAG-3, TIGIT, and BTLA, whose sequences are shown in Figure 2. Accordingly, suitable bispecific antibodies bind to PD-1 and CTLA-4, PD-1 and TIM-3, PD-1 and LAG-3, PD-1 and TIGIT, PD-1 and BTLA, CTLA-4 and TIM-3, CTLA-4 and LAG-3, CTLA-4 and TIGIT, CTLA-4 and BTLA, TIM-3 and LAG-3, TIM-3 and TIGIT, TIM-3 and BTLA, LAG-3 and TIGIT, LAG-3 and BTLA, and TIGIT and BTLA. It should be noted that, generally, these bispecific antibodies are named "anti-PD-1X anti-CTLA-4" for each pair, or, more commonly, simply, for ease of use (and therefore interchangeability), "PD-1XCTLA-4," etc. Unless otherwise specified herein, the order of the antigen list in the name does not confer structure. That is, a PD-1XCTLA-4 capping antibody may have scFvs that bind to PD-1 or CTLA-4, but in some cases, the order specifies the structure as shown.

[0257] As will be outlined more fully in this specification, these combinations of ABDs may be of various types, as outlined below, and are generally combinations in which one ABD is of the Fab type and the other is of the scFv type. Some types discussed in this specification and shown in Figure 1 use a single Fab and a single scFv (Figures 1A, C, and D), and some types use two Fabs and a single scFv (Figures 1E, F, G, H, and I).

[0258] A. Antigen-binding domain As discussed herein, the bispecificity checkpoint heterodimer antibody of the present invention The antibody comprises two antigen-binding domains (ABDs), each of which binds to a different checkpoint protein. As outlined herein, these heterodimer antibodies may be bispecific and bivalent (each antigen is bound by a single ABD, e.g., in the form shown in Figure 1A) or bispecific and trivalent (one antigen is bound by a single ABD and the other by two ABDs, e.g., as shown in Figure 1F).

[0259] In addition, generally, one of the ABDs contains scFv, as outlined herein, with an N-to-C-terminal orientation of vh-scFv linker-vl or vl-scFv linker-vh. Depending on the type, one or both of the other ABDs are generally Fab, containing a vh domain on one protein chain (generally as a component of the heavy chain) and a vl domain on the other protein chain (generally as a component of the light chain).

[0260] The present invention provides several ABDs that bind to several different checkpoint proteins, as outlined below. As will be understood by those skilled in the art, any set of six CDRs or vh and vl domains can be of the scFv or Fab type, which are then added to a heavy constant domain and a light constant domain, where the heavy constant domain includes variants (contained within the CH1 domain and the Fc domain). The scFv sequences included in the sequence listing utilize a specific charged linker, but as outlined herein, uncharged or other charged linkers can be used, including those shown in Figure 7.

[0261] In addition, as mentioned above, the numbering used in the identification sequence listing of CDRs is Kabat, but different numbering can be used, and the amino acid sequence of the CDR changes, as shown in Table 1.

[0262] Further variants can be created for all of the variable weight and variable light domains listed herein. As outlined herein, in some embodiments, a set of six CDRs may have 0, 1, 2, 3, 4, or 5 amino acid modifications (amino acid substitutions are particularly useful), as well as changes in the framework regions of the variable weight and variable light domains, as long as the framework (excluding the CDRs) retains at least 80, 85, or 90% identity with a human germline sequence selected from those listed in Figure 1 of U.S. Patent No. 7,657,380 (the figure and legend are incorporated herein by reference in their entirety). Thus, for example, identical CDRs described herein can be combined with different framework sequences from human germline sequences, as long as the framework regions retain at least 80, 85, or 90% identity with a human germline sequence selected from those listed in Figure 1 of U.S. Patent No. 7,657,380. Alternatively, a CDR may have amino acid modifications (for example, one, two, three, four, or five amino acid modifications in a set of CDRs (i.e., any combination of CDRs may be altered as long as the total number of changes in a set of six CDRs is less than six), and the framework region may be altered, as long as the framework region maintains at least 80, 85, or 90% identity with respect to a human germline sequence selected from those listed in Figure 1 of U.S. Patent No. 7,657,380).

[0263] B.PD-1 antigen-binding domain In some embodiments, one of the ABDs binds to PD-1. A preferred set of six CDRs and / or vh and vl domains, as well as scFv sequences, are found in SEQ ID NOs. 6209-11464, 11465-17134, 33003-33072, 33073-35394, and 36127-36146. As shown, ABD sequences for specific purposes in some embodiments are shown in Figure 9 and include those sequences in the sequence listing, having identifiers 1G6_H1.279_L1.194, 1G6_H1.280_L1.224, 1G6_L1.194_H1.279, 1G6_L1.210_H1.288, and 2E9_H1L1.

[0264] As will be understood by those skilled in the art, preferred anti-PD-1 ABDs may include any of the following CDRs identified using other alignments within the vh and vl sequences of SEQ ID NOs. 6209-11464, SEQ ID NOs. 11465-17134, SEQ ID NOs. 33003-33072, SEQ ID NOs. 33073-35394, and SEQ ID NOs. 36127-36146, as underlined, or when a different numbering scheme is used as described herein and shown in Table 1, including these sequences and the set of six CDRs shown in the figure. Preferred ABDs may also include these sequences and the entire vh and vl sequences shown in the figure, used as scFv or Fab. In many of the embodiments herein containing Fv vs. PD-1, it is the scFv monomer that binds to PD-1. As discussed herein, when PD-1 is one of the antigens, the other of the target pair is CTLA-4 (preferred sequences are SEQ ID NOs. 21-2918, 2919-6208, 36739-36818, and 35395-35416 (which may be scFv sequences, CDR sequence sets, or vh and vl sequences)), TIM-3 (preferred sequences are SEQ ID NOs. 20765-20884, 37587-37698, and 36347-36706 (which may be scFv sequences, CDR sequence sets, or vh and vl sequences)), LAG-3 (preferred sequences are SEQ ID NOs. 17135-2076 4. Select from sequence numbers 36819-36962, 35417-35606, 25194-32793, and 32794-33002 (which may be scFv sequences, CDR sequence sets, or vh and vl sequences), BTLA (preferred sequences are sequence numbers 20885-21503 and 36707-36738 (which may be scFv sequences, CDR sequence sets, or vh and vl sequences)), and TIGIT (preferred sequences are sequence numbers 21504-21523 and 37435-37586 (which may be scFv sequences, CDR sequence sets, or vh and vl sequences)).

[0265] Particularly useful ABDs that bind to human PD-1 include, but are not limited to, 1G6_H1.279_L1.194, 1G6_H1.280_L1.224, 1G6_L1.194_H1.279, 1G6_L1.210_H1.288, and 2E9_H1L1.

[0266] In addition to the parental CDR sets disclosed in the sequence listings that form ABD vs. PD-1, the present invention provides mutant CDR sets. In one embodiment, a set of six CDRs may have one, two, three, four, or five amino acid changes from the parental CDRs, as long as ABD can still bind to the target antigen, when measured by at least one of the following assays: Biacore, surface plasmon resonance (SPR), and / or BLI (biolayer interferometry, e.g., Octet assay) (the latter being particularly useful in many embodiments).

[0267] In addition to the parental variable heavy and variable light domains disclosed herein that form ABD vs. PD-1, the present invention provides mutant vh and vl domains. In one embodiment, the mutant vh and vl domains may have 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid changes from the parental vh and vl domains, respectively, as long as ABD can still bind to the target antigen when measured by at least one of Biacore, surface plasmon resonance (SPR), and / or BLI (biolayer interference, e.g., Octet assay) assays (the latter being particularly useful in many embodiments). In another embodiment, the mutant vh and vl may have 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid changes from the parental vh and vl domains, as long as ABD can still bind to the target antigen when measured by at least one of Biacore, surface plasmon resonance (SPR), and / or BLI (biolayer interference, e.g., Octet assay). When measured by at least one of the assays (the latter being particularly useful in many embodiments), ABD is at least 90, 95, 97, 98, or 99% identical to the respective parent vh or vl, insofar as it can still bind to the target antigen.

[0268] A specific preferred embodiment is the scFv type 1G6_L1.194_H1.279 anti-PD-1 Fv contained within one of the bottle opener-type skeletons shown in Figure 37.

[0269] A specific preferred embodiment is the scFv type 1G6_L1.194_H1.279 anti-PD-1 Fv, which is included in one of the mAb-scFv type frameworks shown in Figure 38.

[0270] C.CTLA-4 antigen-binding domain In some embodiments, one of the ABDs binds to CTLA-4. A preferred set of six CDRs and / or vh and vl domains, as well as scFv sequences, are shown in SEQ ID NOs: 21-2918, 2919-6208, 36739-36818, and 35395-35416.ABD sequences for specific purposes in some embodiments are shown in Figure 10, with identifiers [CTLA-4]_H0.25_L0, [CTLA-4]_H0.26_L0, [CTLA-4]_H0.27_L0, [CTLA-4]_H0.29_L0, [CTLA-4]_H0.38_L0, [CTLA-4]_H0.39_L0, [CTLA-4]_H0.40_L0, [CTLA-4]_H0.70_L0, [CTLA-4]_H0_L0.22, [CTLA-4]_H2_L0, [CTLA-4] _H3.21_L0.124, [CTLA-4]_H3.21_L0.129, [CTLA-4]_H3.21_L0.132, [CTLA-4]_H3.23_L0.124, [CTLA-4]_H3.23_L0.129, [CTLA-4]_ H3.23_L0.132, [CTLA-4]_H3.25_L0.124, [CTLA-4]_H3.25_L0.129, [CTLA-4]_H3.25_L0.132, [CTLA-4]_H3.4_L0.118, [CTLA-4]_H3. 4_L0.119, [CTLA-4]_H3.4_L0.12, [CTLA-4]_H3.4_L0.121, [CTLA-4]_H3.4_L0.122, [CTLA-4]_H3.4_L0.123, [CTLA-4]_H3.4_L0.12 4, [CTLA-4]_H3.4_L0.125, [CTLA-4]_H3.4_L0.126, [CTLA-4]_H3.4_L0.127, [CTLA-4]_H3.4_L0.128, [CTLA-4]_H3.4_L0.129, [CTLA The sequence listing also includes those sequences having [CTLA-4]_H3.4_L0.130, [CTLA-4]_H3.4_L0.131, [CTLA-4]_H3.4_L0.132, [CTLA-4]_H3.5_L2.1, [CTLA-4]_H3.5_L2.2, [CTLA-4]_H3.5_L2.3, [CTLA-4]_H3_L0, [CTLA-4]_H3_L0.22, [CTLA-4]_H3_L0.44, [CTLA-4]_H3_L0.67, and [CTLA-4]_H3_L0.74.

[0271] As will be understood by those skilled in the art, preferred anti-CTLA-4ABDs may include any of the following CDRs identified using other alignments within the vh and vl sequences of SEQ ID NOs. 21-2918, SEQ ID NOs. 2919-6208, SEQ ID NOs. 36739-36818, and SEQ ID NOs. 35395-35416, as underlined, or when a different numbering scheme is used as described herein and shown in Table 1, including these sequences and the set of six CDRs shown in the figure. Preferred ABDs may also include these sequences and the entire vh and vl sequences shown in the figure, used as scFv or Fab. In many of the embodiments herein containing Fv vs. CTLA-4, it is the scFv monomer that binds to CTLA-4. As discussed herein, when CTLA-4 is one of the antigens, the other of the target pair is PD-1 (preferred sequences are SEQ ID NOs. 6209-1 1464, SEQ ID NOs. 11465~17134, SEQ ID NOs. 33003~33072, SEQ ID NOs. 33073~35394, and SEQ ID NOs. 36127~36146 (may be scFv sequences, CDR sequence sets, or vh and vl sequences)), TIM-3 (preferred sequences are SEQ ID NOs. 20765~20884, SEQ ID NOs. 37587~37698, and SEQ ID NOs. 36347~36706 (may be scFv sequences, CDR sequence sets, or vh and vl sequences)), LAG-3 (preferred sequences are SEQ ID NOs. 17135~20764, SEQ ID NOs. 36819~36962, SEQ ID NOs. The sequence numbers 35417-35606, 25194-32793, and 32794-33002 (which may be scFv sequences, CDR sequence sets, or vh and vl sequences) are selected from the following: BTLA (preferred sequences are 20885-21503 and 36707-36738 (which may be scFv sequences, CDR sequence sets, or vh and vl sequences)), and TIGIT (preferred sequences are 21504-21523 and 37435-37586 (which may be scFv sequences, CDR sequence sets, or vh and vl sequences)).

[0272] In addition to the parental CDR sets disclosed in the sequence listing that form ABD vs. CTLA-4, the present invention provides mutant CDR sets. In one embodiment, a set of six CDRs may have one, two, three, four, or five amino acid changes from the parental CDRs, as long as ABD can still bind to the target antigen, when measured by at least one of the following assays: Biacore, surface plasmon resonance (SPR), and / or BLI (biolayer interferometry, e.g., Octet assay) (the latter being particularly useful in many embodiments).

[0273] In addition to the parental variable heavy and variable light domains disclosed herein that form ABD vs. CTLA-4, the present invention provides mutant vh and vl domains. In one embodiment, the mutant vh and vl domains may each have 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid changes from the parental vh and vl domains, insofar as ABD can still bind to the target antigen when measured by at least one of Biacore, surface plasmon resonance (SPR), and / or BLI (biolayer interference, e.g., Octet assay) assays (the latter being particularly useful in many embodiments). In another embodiment, the mutant vh and vl are at least 90, 95, 97, 98, or 99% identical to the respective parental vh or vl, insofar as ABD can still bind to the target antigen when measured by at least one of Biacore, surface plasmon resonance (SPR), and / or BLI (biolayer interference, e.g., Octet assay) assays (the latter being particularly useful in many embodiments).

[0274] A specific preferred embodiment is the Fab-type [CTLA-4]_H3_L0.22 anti-CTLA-4 Fv contained within one of the bottle opener-type skeletons shown in Figure 37.

[0275] A specific preferred embodiment is the scFv type [CTLA-4]_H3_L0.22 anti-CTLA-4 Fv contained within one of the bottle opener-type skeletons shown in Figure 37.

[0276] A specific preferred embodiment is the scFv type [CTLA-4]_H3_L0.22 anti-CTLA-4 Fv, which is included in one of the mAb-scFv type frameworks shown in Figure 38.

[0277] A specific preferred embodiment is the Fab-type [CTLA-4]_H3_L0.22 anti-CTLA-4 Fv contained within one of the mAb-scFv type frameworks shown in Figure 38.

[0278] D.TIM-3 antigen-binding domain In some embodiments, one of the ABDs binds to TIM-3. A preferred set of six CDRs and / or vh and vl domains, as well as scFv sequences, are shown in SEQ ID NOs. 20765-20884, SEQ ID NOs. 37587-37698, and SEQ ID NOs. 36347-36706. ABD sequences for specific purposes in some embodiments include those sequences in the sequence listing, having identifiers 1D10_H0L0, 1D12_H0L0, 3H3_H1_L2.1, 6C8_H0L0, 6D9_H0_1D12_L0, 7A9_H0L0, 7B11_H0L0, 7B11var_H0L0, and 7C2_H0L0.

[0279] As will be understood by those skilled in the art, preferred anti-TIM-3ABDs may include either CDRs identified using other alignments within the vh and vl sequences of SEQ ID NOs. 20765-20884, SEQ ID NOs. 37587-37698, and SEQ ID NOs. 36347-36706, as underlined, or when different numbering schemes are used as described herein and shown in Table 1, these sequences and the set of six CDRs shown in the figure may also include these sequences and the entire vh and vl sequences shown in the figure, used as scFv or Fab. In many of the embodiments herein containing Fv vs. TIM-3, it is the Fab monomer that binds to TIM-3. As discussed herein, when TIM-3 is one of the antigens, the other of the target pair is PD-1 (preferred sequences are SEQ ID NOs. 6209-11464, SEQ ID NOs. 11465-17134, SEQ ID NOs. 33003-33072, SEQ ID NOs. 33073-35394, and SEQ ID NOs. 36127-36146 (which may be scFv sequences, CDR sequence sets, or vh and vl sequences)), CTLA-4 (preferred sequences are SEQ ID NOs. 21-2918, SEQ ID NOs. 2919-6208, SEQ ID NOs. 36739-36818, and SEQ ID NOs. 35395-35416 (which may be scFv sequences, CDR sequence sets, or vh and vl sequences)), LAG-3 (preferred sequences are The sequences selected are: ,

[0280] In addition to the parental CDR sets disclosed in the sequence listing that form ABD vs. TIM-3, the present invention provides mutant CDR sets. In one embodiment, a set of six CDRs may have one, two, three, four, or five amino acid changes from the parental CDRs, as long as ABD can still bind to the target antigen, when measured by at least one of the following assays: Biacore, surface plasmon resonance (SPR), and / or BLI (biolayer interferometry, e.g., Octet assay) (the latter being particularly useful in many embodiments).

[0281] In addition to the parental variable heavy and variable light domains disclosed herein that form ABD vs. TIM-3, the present invention provides mutant vh and vl domains. In one embodiment, the mutant vh and vl domains have 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid changes from the parental vh and vl domains, respectively, when measured by at least one of Biacore, surface plasmon resonance (SPR), and / or BLI (biolayer interference, e.g., Octet assay) assays (the latter being particularly useful in many embodiments), as long as ABD can still bind to the target antigen. In another embodiment, the variant vh and vl are at least 90, 95, 97, 98, or 99% identical to their respective parent vh or vl, insofar as ABD can still bind to the target antigen, when measured by at least one of Biacore, surface plasmon resonance (SPR), and / or BLI (biolayer interference, e.g., Octet assay) assays (the latter being particularly useful in many embodiments).

[0282] LAG-3 antigen-binding domain In some embodiments, one of the ABDs binds to LAG-3. A preferred set of six CDRs and / or vh and vl domains, as well as scFv sequences, are shown in SEQ ID NOs: 17135-20764, 36819-36962, 35417-35606, 25194-32793, and 32794-33002. ABD sequences for specific purposes in some embodiments are shown in Figure 11, with identifiers 2A11_H0L0, 2A11_H1.125_L2.113, 2A11_H1.144_L2.142, 2A11_H1_L2.122, 2A11_H1_L2.123, 2A11_H1_L2.124, 2A11_H1_L2.25, 2A11_H1_L2.47, and 2A11_H1_L2.50. , 2A11_H1_L2.91, 2A11_H1_L2.93, 2A11_H1_L2.97, 2A11_H1L1, 2A11_H1L2, 2A11_H2L2, 2A11_H3L1, 2A1 1_H3L2, 2A11_H4L1, 2A11_H4L2, 7G8_H0L0, 7G8_H1L1, 7G8_H3.18_L1.11, 7G8_H3.23_L1.11, 7G8_H3.28 This also includes the sequences in the sequence listing that have L1, 7G8_H3.28_L1.11, 7G8_H3.28_L1.13, 7G8_H3.30_L1.34, 7G8_H3.30_L1.34, and 7G8_H3L1.

[0283] As will be understood by those skilled in the art, preferred anti-LAG-3 ABDs may include any of the following CDRs identified using other alignments within the vh and vl sequences of SEQ ID NOs. 17135-20764, SEQ ID NOs. 36819-36962, SEQ ID NOs. 35417-35606, SEQ ID NOs. 25194-32793, and SEQ ID NOs. 32794-33002, as underlined, or when different numbering schemes are used as described herein and shown in Table 1, including these sequences and the set of six CDRs shown in the figure. Preferred ABDs may also include these sequences and the entire vh and vl sequences shown in the figure, used as scFv or Fab. In many of the embodiments herein containing Fv vs. LAG-3, it is the Fab monomer that binds to LAG-3. As discussed herein, when LAG-3 is one of the antigens, the other of the target pair may be PD-1 (preferred sequences are SEQ ID NOs. 6209-11464, 11465-17134, 33003-33072, 33073-35394, and 36127-36146 (which may be scFv sequences, CDR sequence sets, or vh and vl sequences)), CTLA-4 (preferred sequences are SEQ ID NOs. 21-2918, 2919-6208, 36739-36818, and 35395-3541). The following sequences are selected from: , TIM-3 (preferred sequences are sequence numbers 20765-20884, 37587-37698, and 36347-36706 (which may be scFv sequences, CDR sequence sets, or vh and vl sequences)), BTLA (preferred sequences are sequence numbers 20885-21503 and 36707-36738 (which may be scFv sequences, CDR sequence sets, or vh and vl sequences)), and TIGIT (preferred sequences are sequence numbers 21504-21523 and 37435-37586 (which may be scFv sequences, CDR sequence sets, or vh and vl sequences)).

[0284] In addition to the parent CDR sets disclosed in the sequence listing that form ABD vs. LAG-3, the present invention A set of variant CDRs is provided. In one embodiment, a set of six CDRs may have one, two, three, four, or five amino acid changes from the parent CDR, as long as the ABD can still bind to the target antigen, when measured by at least one of the following assays: Biacore, surface plasmon resonance (SPR), and / or BLI (biolayer interferometry, e.g., Octet assay) (the latter being particularly useful in many embodiments).

[0285] In addition to the parental variable heavy and variable light domains disclosed herein that form ABD vs. LAG-3, the present invention provides mutant vh and vl domains. In one embodiment, the mutant vh and vl domains may each have 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid changes from the parental vh and vl domains, insofar as ABD can still bind to the target antigen when measured by at least one of Biacore, surface plasmon resonance (SPR), and / or BLI (biolayer interference, e.g., Octet assay) assays (the latter being particularly useful in many embodiments). In another embodiment, the mutant vh and vl are at least 90, 95, 97, 98, or 99% identical to the respective parental vh or vl, insofar as ABD can still bind to the target antigen when measured by at least one of Biacore, surface plasmon resonance (SPR), and / or BLI (biolayer interference, e.g., Octet assay) assays (the latter being particularly useful in many embodiments).

[0286] A specific preferred embodiment is the Fab-type 7G8_H3.30_L1.34 anti-LAG-3 Fv contained within one of the bottle opener-type skeletons shown in Figure 37.

[0287] A specific preferred embodiment is the scFv type 7G8_H3.30_L1.34 anti-LAG-3 Fv contained within one of the bottle opener-type skeletons shown in Figure 37.

[0288] E.BTLA antigen-binding domain In some embodiments, one of the ABDs binds to the BTLA. A preferred set of six CDRs and / or vh and vl domains, as well as scFv sequences, are shown in SEQ ID NOs. 20885-21503 and SEQ ID NOs. 36707-36738. ABD sequences for specific purposes in some embodiments are shown in Figure 12, and also include those sequences in the sequence listings, having identifiers 9C6_H0L0;9C6_H1.1_L1 and 9C6_H1.11_L1.

[0289] As will be understood by those skilled in the art, preferred anti-BTLA ABDs may include either CDRs identified using other alignments of the vh and vl sequences SEQ ID NOs. 20885-21503 and SEQ ID NOs. 36707-36738, as underlined or when different numbering schemes are used as described herein and shown in Table 1, and may include these sequences and the set of six CDRs shown in the figure. Preferred ABDs may also include these sequences and the entire vh and vl sequences shown in the figure, used as scFv or Fab. In many of the embodiments herein containing Fv versus BTLA, it is the Fab monomer that binds to BTLA. As discussed herein, when LAG-3 is one of the antigens, the other of the target pair is PD-1 (preferred sequences are SEQ ID NOs. 6209-11464, SEQ ID NOs. 11465-17134, SEQ ID NOs. 33003-33072, SEQ ID NOs. 33073-35394, and SEQ ID NOs. 36127-36146 (which may be scFv sequences, CDR sequence sets, or vh and vl sequences)), CTLA-4 (preferred sequences are SEQ ID NOs. 21-2918, SEQ ID NOs. 2919-6208, SEQ ID NOs. 36739-36818, and SEQ ID NOs. 35395-35416 (which may be scFv sequences, CDR sequence sets, or vh and vl sequences)), TIM-3 (preferred sequences are The sequence is selected from sequence numbers 20765-20884, sequence numbers 37587-37698, and sequence numbers 36347-36706 (which may be scFv sequences, CDR sequence sets, or vh and vl sequences), LAG-3 (preferred sequences are sequence numbers 17135-20764, 36819-36962, 35417-35606, 25194-32793, and 32794-33002 (which may be scFv sequences, CDR sequence sets, or vh and vl sequences)), and TIGIT (preferred sequences are sequence numbers 21504-21523 and 37435-37586 (which may be scFv sequences, CDR sequence sets, or vh and vl sequences)).

[0290] In addition to the parental CDR sets disclosed in the sequence listings that form ABD vs. BTLA, the present invention provides mutant CDR sets. In one embodiment, a set of six CDRs may have one, two, three, four, or five amino acid changes from the parental CDRs, as long as ABD can still bind to the target antigen when measured by at least one of the following assays: Biacore, surface plasmon resonance (SPR), and / or BLI (biolayer interferometry, e.g., Octet assay) (the latter being particularly useful in many embodiments).

[0291] In addition to the parental variable heavy and variable light domains disclosed herein that form ABD vs. BTLA, the present invention provides mutant vh and vl domains. In one embodiment, the mutant vh and vl domains may each have 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid changes from the parental vh and vl domains, insofar as ABD can still bind to the target antigen when measured by at least one of Biacore, surface plasmon resonance (SPR), and / or BLI (biolayer interference, e.g., Octet assay) assays (the latter being particularly useful in many embodiments). In another embodiment, the mutant vh and vl are at least 90, 95, 97, 98, or 99% identical to the respective parental vh or vl, insofar as ABD can still bind to the target antigen when measured by at least one of Biacore, surface plasmon resonance (SPR), and / or BLI (biolayer interference, e.g., Octet assay) assays (the latter being particularly useful in many embodiments).

[0292] A specific preferred embodiment is the Fab-type 9C6_H1.1_L1 anti-LAG-3 Fv contained within one of the bottle opener-type skeletons shown in Figure 37.

[0293] A specific preferred embodiment is the scFv type 7G8_H3.30_L1.34 anti-LAG-3 Fv contained within one of the bottle opener-type skeletons shown in Figure 37.

[0294] F.TIGIT antigen-binding domain In some embodiments, one of the ABDs binds to TIGIT. A preferred set of six CDRs and / or vh and vl domains, as well as scFv sequences, are shown in SEQ ID NOs. 21504-21523 and SEQ ID NOs. 37435-37586.

[0295] As will be understood by those skilled in the art, preferred anti-TIGIT ABDs may include either CDRs identified using other alignments within the vh and vl sequences of SEQ ID NOs. 21504-21523 and SEQ ID NOs. 37435-37586, either as underlined or, when a different numbering scheme is used as described herein and shown in Table 1, these sequences and the set of six CDRs shown in the figure may also include these sequences and the entire vh and vl sequences shown in the figure, used as scFv or Fab. In many of the embodiments herein containing Fv vs. TIGIT, it is the Fab monomer that binds to TIGIT. As discussed herein, L When AG-3 is one of the antigens, the other of the target pair is PD-1 (preferred sequences are SEQ ID NOs. 6209-11464, 11465-17134, 33003-33072, 33073-35394, and 36127-36146 (which may be scFv sequences, CDR sequence sets, or vh and vl sequences)), CTLA-4 (preferred sequences are SEQ ID NOs. 21-2918, 2919-6208, 36739-36818, and 35395-35416 (which may be scFv sequences, CDR sequence sets, or vh and vl sequences)), TIM-3 (preferred sequences are SEQ ID NOs. 20765-20884) , sequences 37587~37698 and sequences 36347~36706 (which may be scFv sequences, CDR sequence sets, or vh and vl sequences)), LAG-3 (preferred sequences are sequences 17135~20764, sequences 36819~36962, sequences 35417~35606, sequences 25194~32793 and sequences 32794~33002 (which may be scFv sequences, CDR sequence sets, or vh and vl sequences)), and BTLA (preferred sequences are sequences 20885~21503 and sequences 36707~36738 (which may be scFv sequences, CDR sequence sets, or vh and vl sequences)).

[0296] G. Specific bispecific embodiments The present invention provides several specific bispecific antibodies, as outlined below.

[0297] 1. LAG-3XCTLA-4 In some embodiments, the present invention provides a bispecific heterodimer antibody comprising a first ABD that binds to human LAG-3 and a second ABD that binds to human CTLA-4, which may be any form shown in Figure 1. Most of the disclosure refers to a bottle-opener type where Fab is on the LAG-3 side and scFv is on the CLTA-4 side, although this may be reversed for all embodiments herein.

[0298] In one embodiment, the LAG-3XCTLA-4 bispecific antibody is the bottle-opener type shown in Figure 1A, and the CTLA-4 ABD is scFv. In another embodiment, the LAG-3XCTLA-4 bispecific antibody is the central scFv type shown in Figure 1F, and the LAG-3 ABD is the Fab component. In yet another embodiment, the LAG-3XCTLA-4 bispecific antibody is the central scFv type shown in Figure 1F, and the CTLA-4 ABD is scFv.

[0299] LAG-3XCTLA-4 bispecific antibodies (either bottle-opener type or central scFv type) generally include asymmetric variants, pI variants, and cleavage variants as outlined herein. Specifically, in either type, the Fc domains of the two monomers include asymmetric variants (e.g., the sets of amino acid substitutions shown in Figures 3 and 8), optionally cleavage variants (including those shown in Figure 5), and the monomer containing the Fab side (e.g., the heavy chain constant domain) includes pI variants (including those shown in Figure 4).

[0300] In some embodiments, the LAG-3XCTLA-4 bispecific antibody comprises an Fc domain having an asymmetric variant, and particularly useful asymmetric variants are selected from the group consisting of S364K / E357Q:L368D / K370S, L368D / K370S:S364K, L368E / K370S:S364K, T411T / E360E / Q362E:D401K, L368D / K370S:S364K / E357L, K370S:S364K / E357Q, T366S / L368A / Y407V:T366W, and T366S / L368A / Y407V / Y349C:T366W / S354C.

[0301] In some embodiments, the LAG-3XCTLA-4 antibody includes asymmetric variants, pI variants, and cleavage variants. Thus, some embodiments provide a bottle opener type, which includes a) a charged scFv linker (in some embodiments, the +H sequence in Figure 7). a) a first monomer ("scFv monomer") comprising an asymmetric mutant S364K / E357Q, a cleavage mutant E233P / L234V / L235A / G236del / S267K, and an Fv that binds to a checkpoint inhibitor outlined herein; b) a second monomer ("Fab monomer") comprising an asymmetric mutant L368D / K370S, a pI mutant N208D / Q295E / N384D / Q418E / N421D, a cleavage mutant E233P / L234V / L235A / G236del / S267K, and a variable heavy domain that, together with a variable light domain, forms an Fv that binds to a second checkpoint inhibitor outlined herein; and c) a light chain. A specific example of this embodiment utilizes LAG-3 Fab 7G8_H3.30_L1.34 and CTLA-4 scFv [CTLA-4]_H3.23_L0.129, but any combination of CTLA-4 or LAG-3 Fv in the sequence listing can be used.

[0302] In some embodiments, the LAG-3XCTLA-4 antibody includes asymmetric variants, pI variants, cleavage variants, and FcRn variants. Thus, some embodiments provide a bottle opener type, which includes a) a first monomer ("scFv monomer") comprising a charged scFv linker (in some embodiments, the +H sequence in Figure 7 is preferred), an asymmetric variant S364K / E357Q, a cleavage variant E233P / L234V / L235A / G236del / S267K, an FcRn variant M428L / N434S, and Fv that binds to checkpoint inhibitors outlined herein, and b) c) a light chain. A specific example of this embodiment utilizes LAG-3 Fab 7G8_H3.30_L1.34 and CTLA-4 scFv [CTLA-4]_H3.23_L0.129, but any combination of CTLA-4 or LAG-3 Fv in the sequence listing can be used.

[0303] Further embodiments include any of the skeletons from Figure 37, having LAG-3 Fab 7G8_H3.30_L1.34 and CTLA-4 scFv [CTLA-4]_H3.23_L0.129.

[0304] Further embodiments include any of the skeletons from Figure 38, having LAG-3 Fab 7G8_H3.30_L1.34 and CTLA-4 scFv [CTLA-4]_H3.23_L0.129.

[0305] In some embodiments, for the LAG-3XCLTA-4 bispecific antibody, the Fv on the LAG-3 Fab side is the identifier 2A11_H0L0, 2A11_H1.125_L2.113, 2A11_H1.144_L2.142, 2A11_H1_L2.122; 2A11_H1_L2.123, 2A11_H1_L2.124, 2A11_H1_L2.25, 2A11_H1_L2.47, 2A11_H1_L2.50, 2A11_H1_L2.91, 2A11_H1_L2.93, 2A11_H1_L2.97, 2A11_H1L1, 2A11_H1L2, 2 A11_H2L2, 2A11_H3L1, 2A11_H3L2, 2A11_H4L1, 2A11_H4L2, 7G8_H0L0, 7G8_H1L1, 7G8_H3.18_L1.11, 7G8_H3.23_L1.11, 7G8_H3.28_L1, 7G8_H3.28_L1.11, 7G8_H3.28_L1.13, 7G8_H3.30_L1.34, 7G8_H3.30_L1.34, and 7G8_H3L1 are selected from the sequences in the sequence listing. The Fv on the CTLA-4 scFv side is identified as [CTLA-4]_H0.25_L0, [CTLA-4]_H0.26_L0, CTLA-4]_H0.27_L0, [CTLA-4]_H0.29_L0, [CTLA-4]_H0.38_L0, [CTLA-4]_H0.39_L0, 0[CTLA-4]_H0.40_L0, [CTLA-4]_H0. 70_L0, [CTLA-4]_H0_L0.22, [CTLA-4]_H2_L0, [CTLA-4]_H3.21_L0.124, [CTLA-4]_H3.21_L0.129, [CTLA-4]_H3.21_L0.13 2, [CTLA-4]_H3.23_L0.124, [CTLA-4]_H3.23_L0.129, [CTLA-4]_H3.23_L0.132, [CTLA-4]_H3.25_L0.124, [CTLA-4]_H3.2 5_L0.129, [CTLA-4]_H3.25_L0.132, [CTLA-4]_H3.4_L0.118, [CTLA-4]_H3.4_L0.119, [CTLA-4]_H3.4_L0.12, [CTLA-4]_H 3.4_L0.121, [CTLA-4]_H3.4_L0.122, [CTLA-4]_H3.4_L0.123, [CTLA-4]_H3.4_L0.124, [CTLA-4]_H3.4_L0.125, [CTLA-4 ]_H3.4_L0.126, [CTLA-4]_H3.4_L0.127, [CTLA-4]_H3.4_L0.128, [CTLA-4]_H3.4_L0.129, [CTLA-4]_H3.4_L0.130, [CTLA Select from the sequences in the sequence listing that have [CTLA-4]_H3.4_L0.131, [CTLA-4]_H3.4_L0.132, [CTLA-4]_H3.5_L2.1, [CTLA-4]_H3.5_L2.2, [CTLA-4]_H3.5_L2.3, [CTLA-4]_H3_L0, [CTLA-4]_H3_L0.22, [CTLA-4]_H3_L0.44, [CTLA-4]_H3_L0.67, and [CTLA-4]_H3_L0.74.

[0306] In some embodiments, the LAG-3XCTLA-4 bispecific antibody is selected from their constructs listed in the sequence numbers 35607-35866 and 21524-22620.

[0307] In some embodiments, the LAG-3XCTLA-4 bispecific antibody is selected from XENP20206, XENP21582, XENP21584, XENP21588, XENP22123, XENP22124, XENP22125, XENP22604, XENP22672, XENP22847, XENP22847, XENP22841, and XENP22849.

[0308] 2. BTLAXPD-1 In some embodiments, the present invention provides a bispecific heterodimer antibody comprising a first ABD that binds to human BTLA and a second ABD that binds to human PD-1, which may be in any form shown in Figure 1. Most of this disclosure refers to a bottle-opener type where Fab is on the BTLA side and the scFv side is on the PD-1 side, although this may be reversed for all embodiments herein.

[0309] In one embodiment, the BTLAXPD-1 bispecific antibody is the bottle-opener type shown in Figure 1A, and the PD-1 ABD is scFv. In another embodiment, the BTLAXPD-1 bispecific antibody is the central scFv type shown in Figure 1F, and the BTLA ABD is the Fab component. In yet another embodiment, the BTLAXPD-1 bispecific antibody is the central scFv type shown in Figure 1F, and the PD-1 ABD is scFv.

[0310] BTLAXPD-1 bispecific antibodies (either bottle-opener type or central scFv type) generally include asymmetric variants, pI variants, and cleavage variants outlined herein. Specifically, in either type, the Fc domains of the two monomers include asymmetric variants (e.g., the sets of amino acid substitutions shown in Figures 3 and 8), optionally cleavage variants (including those shown in Figure 5), and the monomer containing the Fab side (e.g., the heavy chain constant domain) includes pI variants. This includes mutants (including those shown in Figure 4).

[0311] In some embodiments, the BTLAXPD-1 bispecific antibody comprises an Fc domain having an asymmetric variant, and particularly useful asymmetric variants are selected from the group consisting of S364K / E357Q:L368D / K370S, L368D / K370S:S364K, L368E / K370S:S364K, T411T / E360E / Q362E:D401K, L368D / K370S:S364K / E357L, K370S:S364K / E357Q, T366S / L368A / Y407V:T366W, and T366S / L368A / Y407V / Y349C:T366W / S354C.

[0312] In some embodiments, the BTLAXPD-1 antibody includes asymmetric variants, pI variants, and cleavage variants. Thus, some embodiments provide a bottle opener type, which includes a) a first monomer ("scFv monomer") comprising a charged scFv linker (in some embodiments, the +H sequence in Figure 7 is preferred), an asymmetric variant S364K / E357Q, a cleavage variant E233P / L234V / L235A / G236del / S267K, and Fv that binds to checkpoint inhibitors outlined herein, and b) c) a light chain, and a second monomer ("Fab monomer") comprising an asymmetric mutant L368D / K370S, a pI mutant N208D / Q295E / N384D / Q418E / N421D, a cleavage mutant E233P / L234V / L235A / G236del / S267K, and a variable heavy domain that, together with a variable light domain, forms an Fv that binds to a second checkpoint inhibitor outlined herein. A specific example of this embodiment utilizes BTLA Fab 9C6_H1.1_L1 and PD-1 scFv 1G6_L1.194_H1.279, but any combination of BTLA or PD-1 Fv in the sequence listing can be used.

[0313] In some embodiments, the BTLAXPD-1 antibody includes asymmetric variants, pI variants, cleavage variants, and FcRn variants. Thus, some embodiments provide a bottle opener type, which includes a) a first monomer ("scFv monomer") comprising a charged scFv linker (in some embodiments, the +H sequence in Figure 7 is preferred), an asymmetric variant S364K / E357Q, a cleavage variant E233P / L234V / L235A / G236del / S267K, an FcRn variant M428L / N434S, and Fv that binds to checkpoint inhibitors outlined herein, and b) c) a light chain, and a second monomer ("Fab monomer") comprising an asymmetric mutant L368D / K370S, a pI mutant N208D / Q295E / N384D / Q418E / N421D, a cleavage mutant E233P / L234V / L235A / G236del / S267K, an FcRn mutant M428L / N434S, and a variable heavy domain that, together with a variable light domain, forms an Fv that binds to a second checkpoint inhibitor outlined herein. A specific example of this embodiment utilizes BTLA Fab 9C6_H1.1_L1 and PD-1 scFv 1G6_L1.194_H1.279, but any combination of BTLA or PD-1 Fv in the sequence listing can be used.

[0314] Further embodiments include any of the skeletons from Figure 37, having BTLA Fab 9C6_H1.1_L1 and PD-1 scFv 1G6_L1.194_H1.279.

[0315] Further embodiments include any of the skeletons from Figure 38, having BTLA Fab 9C6_H1.1_L1 and PD-1 scFv 1G6_L1.194_H1.279.

[0316] In some embodiments, for the BTLAXPD-1 bispecific antibody, BTLA F The Fv on the ab side is selected from the sequences in the sequence listing that have the identifiers 9C6_H0L0, 9C6_H1.1_L1, and 9C6_H1.11_L1. The Fv on the PD-1 scFv side is selected from the sequences in the sequence listing that have the identifiers 1G6_H1.279_L1.194, 1G6_H1.280_L1.224, 1G6_L1.194_H1.279, 1G6_L1.210_H1.288, and 2E9_H1L1.

[0317] In some embodiments, the BTLAXPD-1 bispecific antibody is selected from constructs including those listed as SEQ ID NOs. 22724-23315 and SEQ ID NOs. 36147-36166.

[0318] In some embodiments, the BTLAXPD-1 bispecific antibody is selected from XENP20895, XENP21220, XENP21221, and XENP22858.

[0319] 3. CTLA-4XPD-1 In some embodiments, the present invention provides a bispecific heterodimer antibody comprising a first ABD that binds to human CTLA-4 and a second ABD that binds to human PD-1, which may be in any form shown in Figure 1. Most of the disclosure refers to a bottle-opener type where Fab is on the CTLA-4 side and scFv is on the PD-1 side, but this may be reversed for all embodiments herein.

[0320] In one embodiment, the CTLA-4XPD-1 bispecific antibody is the bottle-opener type shown in Figure 1A, and the PD-1 ABD is scFv. In another embodiment, the CTLA-4XPD-1 bispecific antibody is the central scFv type shown in Figure 1F, and the CTLA-4 ABD is the Fab component. In yet another embodiment, the CTLA-4XPD-1 bispecific antibody is the central scFv type shown in Figure 1F, and the PD-1 ABD is scFv.

[0321] CTLA-4XPD-1 bispecific antibodies (either bottle-opener type or central scFv type) generally include asymmetric variants, pI variants, and cleavage variants as outlined herein. Specifically, in either type, the Fc domains of the two monomers include asymmetric variants (e.g., the sets of amino acid substitutions shown in Figures 3 and 8), optionally cleavage variants (including those shown in Figure 5), and the monomer containing the Fab side (e.g., the heavy chain constant domain) includes pI variants (including those shown in Figure 4).

[0322] In some embodiments, the CTLA-4XPD-1 bispecific antibody comprises an Fc domain having an asymmetric variant, and particularly useful asymmetric variants are selected from the group consisting of S364K / E357Q:L368D / K370S, L368D / K370S:S364K, L368E / K370S:S364K, T411T / E360E / Q362E:D401K, L368D / K370S:S364K / E357L, K370S:S364K / E357Q, T366S / L368A / Y407V:T366W, and T366S / L368A / Y407V / Y349C:T366W / S354C.

[0323] In some embodiments, the CTLA-4XPD-1 antibody includes asymmetric mutants, pI mutants, and cleavage mutants. Thus, some embodiments provide a bottle opener type, which bottle opener type comprises a) a first monomer ("scFv monomer") containing a charged scFv linker (in some embodiments, the +H sequence in Figure 7 is preferred), an asymmetric mutant S364K / E357Q, a cleavage mutant E233P / L234V / L235A / G236del / S267K, and Fv that binds to checkpoint inhibitors outlined herein, and b) an asymmetric mutant L368D / K370S, a pI mutant N208D / Q295E / N384D / Q418E / N421D, a cleavage mutant E233P / L234V / L235A / G236del / The present embodiment includes a second monomer ("Fab monomer"), which includes S267K and a variable heavy domain that, together with a variable light domain, forms an Fv that binds to a second checkpoint inhibitor outlined herein, and a light chain. A specific example of this embodiment utilizes CTLA-4 Fab[CTLA-4]_H3_L0.22 and PD-1 scFv 1G6_L1.194_H1.279, but any combination of CTLA-4 or PD-1 Fv in the sequence listing can be used.

[0324] In some embodiments, the CTLA-4XPD-1 antibody includes asymmetric variants, pI variants, cleavage variants, and FcRn variants. Thus, some embodiments provide a bottle opener type, which includes a) a first monomer ("scFv monomer") containing a charged scFv linker (in some embodiments, the +H sequence in Figure 7 is preferred), an asymmetric variant S364K / E357Q, a cleavage variant E233P / L234V / L235A / G236del / S267K, an FcRn variant M428L / N434S, and Fv that binds to checkpoint inhibitors outlined herein, and b) c) a light chain. A specific example of this embodiment utilizes CTLA-4 Fab[CTLA-4]_H3_L0.22 and PD-1 scFv 1G6_L1.194_H1.279, but any combination of CTLA-4 or PD-1 Fv in the sequence listing can be used.

[0325] Further embodiments include any of the skeletons from Figure 37 having CTLA-4 Fab[CTLA-4]_H3_L0.22 and PD-1 scFv 1G6_L1.194_H1.279.

[0326] Further embodiments include any of the skeletons from Figure 38, having CTLA-4 Fab[CTLA-4]_H3_L0.22 and PD-1 scFv 1G6_L1.194_H1.279.

[0327] In some embodiments, for CTLA-4XPD-1 bispecific antibodies, the Fv on the CTLA-4 Fab side is the identifier [CTLA-4]_H0.25_L0, [CTLA-4]_H0.26_L0, [CTLA-4]_H0.27_L0, [CTLA-4]_H0.29_L0, [CTLA-4]_H0.38_L0, [CTLA-4]_H0.39_L0, 0[CTLA-4]_H0.40_L0, [CTLA-4]_H0.70_L0, [CTLA-4]_H0_L0.22, [CTL A-4]_H2_L0, [CTLA-4]_H3.21_L0.124, [CTLA-4]_H3.21_L0.129, [CTLA-4]_H3.21_L0.132, [CTLA-4]_H3. 23_L0.124, [CTLA-4]_H3.23_L0.129, [CTLA-4]_H3.23_L0.132, [CTLA-4]_H3.25_L0.124, [CTLA-4]_H3.25 _L0.129, [CTLA-4]_H3.25_L0.132, [CTLA-4]_H3.4_L0.118, [CTLA-4]_H3.4_L0.119, [CTLA-4]_H3.4_L0. 12, [CTLA-4]_H3.4_L0.121, [CTLA-4]_H3.4_L0.122, [CTLA-4]_H3.4_L0.123, [CTLA-4]_H3.4_L0.124, [CT LA-4]_H3.4_L0.125, [CTLA-4]_H3.4_L0.126, [CTLA-4]_H3.4_L0.127, [CTLA-4]_H3.4_L0.128, [CTLA-4]_ H3.4_L0.129, [CTLA-4]_H3.4_L0.130, [CTLA-4]_H3.4_L0.131, [CTLA-4]_H3.4_L0.132, [CTLA-4]_H3.5_L 2.1. The Fv on the PD-1 scFv side is selected from the sequences in the sequence listing that have identifiers such as [CTLA-4]_H3.5_L2.2, [CTLA-4]_H3.5_L2.3, [CTLA-4]_H3_L0, [CTLA-4]_H3_L0.22, [CTLA-4]_H3_L0.44, [CTLA-4]_H3_L0.67, and [CTLA-4]_H3_L0.74.

[0328] In some embodiments, the CTLA-4XPD-1 bispecific antibody is selected from those listed as SEQ ID NOs: 36167-36346 and 23316-23735.

[0329] In some embodiments, the CTLA-4XPD-1 bispecific antibody is selected from XENP19738, XENP19739, XENP19741, XENP20053, XENP20066, XENP20130, XENP20146, XENP20717, and XENP22836.

[0330] 4. LAG-3XPD-1 In some embodiments, the present invention provides a bispecific heterodimer antibody comprising a first ABD that binds to human LAG-3 and a second ABD that binds to human PD-1, which may be in any form shown in Figure 1. Most of the disclosure refers to a bottle-opener type where Fab is on the LAG-3 side and scFv is on the PD-1 side, but this may be reversed for all embodiments herein.

[0331] In one embodiment, the LAG-3XPD-1 bispecific antibody is the bottle-opener type shown in Figure 1A, and the PD-1 ABD is scFv. In another embodiment, the LAG-3XPD-1 bispecific antibody is the central scFv type shown in Figure 1F, and the LAG-3 ABD is the Fab component. In yet another embodiment, the LAG-3XPD-1 bispecific antibody is the central scFv type shown in Figure 1F, and the PD-1 ABD is scFv.

[0332] LAG-3XPD-1 bispecific antibodies (either bottle-opener type or central scFv type) generally include asymmetric variants, pI variants, and cleavage variants as outlined herein. Specifically, in either type, the Fc domains of the two monomers include asymmetric variants (e.g., the sets of amino acid substitutions shown in Figures 3 and 8), optionally cleavage variants (including those shown in Figure 5), and the monomer containing the Fab side (e.g., the heavy chain constant domain) includes pI variants (including those shown in Figure 4).

[0333] In some embodiments, the LAG-3XPD-1 bispecific antibody comprises an Fc domain having an asymmetric variant, and particularly useful asymmetric variants are selected from the group consisting of S364K / E357Q:L368D / K370S, L368D / K370S:S364K, L368E / K370S:S364K, T411T / E360E / Q362E:D401K, L368D / K370S:S364K / E357L, K370S:S364K / E357Q, T366S / L368A / Y407V:T366W, and T366S / L368A / Y407V / Y349C:T366W / S354C.

[0334] In some embodiments, the LAG-3XPD-1 antibody includes asymmetric variants, pI variants, and cleavage variants. Thus, some embodiments provide a bottle opener type, which includes a) a charged scFv linker (in some embodiments, the +H sequence in Figure 7 is preferred), an asymmetric variant S364K / E357Q, and a cleavage variant E233P / L234V / The first monomer ("scFv monomer") comprises L235A / G236del / S267K and an Fv that binds to a checkpoint inhibitor outlined herein; a) a second monomer ("Fab monomer") comprises an asymmetric mutant L368D / K370S, a pI mutant N208D / Q295E / N384D / Q418E / N421D, a cleavage mutant E233P / L234V / L235A / G236del / S267K and a variable heavy domain that, together with a variable light domain, forms an Fv that binds to a second checkpoint inhibitor outlined herein; and c) a light chain. A specific example of this embodiment utilizes LAG-3 Fab 7G8_H3.30_L1.34 and PD-1 scFv 1G6_L1.194_H1.279, but any combination of LAG-3 or PD-1 Fv in the sequence listing can be used.

[0335] In some embodiments, the LAG-3XPD-1 antibody includes asymmetric variants, pI variants, cleavage variants, and FcRn variants. Thus, some embodiments provide a bottle opener type, which includes a) a first monomer ("scFv monomer") comprising a charged scFv linker (in some embodiments, the +H sequence in Figure 7 is preferred), an asymmetric variant S364K / E357Q, a cleavage variant E233P / L234V / L235A / G236del / S267K, an FcRn variant M428L / N434S, and Fv that binds to checkpoint inhibitors outlined herein, and b) c) a light chain. A specific example of this embodiment utilizes LAG-3 Fab 7G8_H3.30_L1.34 and PD-1 scFv 1G6_L1.194_H1.279, but any combination of LAG-3 or PD-1 Fv in the sequence listing can be used.

[0336] Further embodiments include any of the skeletons from Figure 37, having LAG-3 Fab 7G8_H3.30_L1.34 and PD-1 scFv 1G6_L1.194_H1.279.

[0337] Further embodiments include any of the skeletons from Figure 38, having LAG-3 Fab 7G8_H3.30_L1.34 and PD-1 scFv 1G6_L1.194_H1.279.

[0338] In some embodiments, a bispecific antibody for LAG-3XPD-1 is used. The Fv on the Fab side is the identifier 2A11_H0L0, 2A11_H1.125_L2.113, 2A11_H1.144_L2.142, 2A11_H1_L2.122; 2A11_H1_L2.123, 2A11_H1_L2.124, 2A11_H1_L2.25, 2A11_H1_L2.47, 2A11_H1_L2.50, 2A11_H1_L2.91, 2A11_H1_L2.93, 2A11_H1_L2.97, 2A11_H1L1, 2A11_H1L2, 2 Select from the sequences in the sequence listing that have A11_H2L2, 2A11_H3L1, 2A11_H3L2, 2A11_H4L1, 2A11_H4L2, 7G8_H0L0, 7G8_H1L1, 7G8_H3.18_L1.11, 7G8_H3.23_L1.11, 7G8_H3.28_L1, 7G8_H3.28_L1.11, 7G8_H3.28_L1.13, 7G8_H3.30_L1.34, 7G8_H3.30_L1.34, and 7G8_H3L1. The Fv on the PD-1 scFv side is selected from the sequences in the sequence listing that have the identifiers 1G6_H1.279_L1.194, 1G6_H1.280_L1.224, 1G6_L1.194_H1.279, 1G6_L1.210_H1.288, and 2E9_H1L1.

[0339] In some embodiments, the LAG-3XPD-1 bispecific antibody is selected from constructs including those listed as SEQ ID NOs: 35867-36126 and 23736-25133.

[0340] In some embodiments, the LAG-3XPD-1 bispecific antibody is selected from XENP20206, XENP21582, XENP21584, XENP21588, XENP22123, XENP22124, XENP22125, XENP22604, XENP22672, XENP22847, XENP22847, and XENP22849.

[0341] 5. TIGITXPD-1 In some embodiments, the TIGITXPD-1 bispecific antibody is selected from those constructs listed in the list of SEQ ID NOs. 25134–25173.

[0342] 6. TIM-3XPD-1 In some embodiments, the present invention provides a bispecific heterodimer antibody comprising a first ABD that binds to human TIM-3 and a second ABD that binds to human PD-1, which may be any form shown in Figure 1. Most of the disclosure refers to a bottle-opener type where Fab is on the TIM-3 side and scFv is on the PD-1 side, but this may be reversed for all embodiments herein.

[0343] In one embodiment, the TIM-3XPD-1 bispecific antibody is the bottle-opener type shown in Figure 1A, and the PD-1 ABD is scFv. In another embodiment, the TIM-3XPD-1 bispecific antibody is the central scFv type shown in Figure 1F, and the TIM-3 ABD is the Fab component. In yet another embodiment, the TIM-3XPD-1 bispecific antibody is the central scFv type shown in Figure 1F, and the PD-1 ABD is scFv.

[0344] TIM-3XPD-1 bispecific antibodies (either bottle-opener type or central scFv type) generally include asymmetric variants, pI variants, and cleavage variants as outlined herein. Specifically, in either type, the Fc domains of the two monomers include asymmetric variants (e.g., the sets of amino acid substitutions shown in Figures 3 and 8), optionally cleavage variants (including those shown in Figure 5), and the monomer containing the Fab side (e.g., the heavy chain constant domain) includes pI variants (including those shown in Figure 4).

[0345] In some embodiments, the TIM-3XPD-1 bispecific antibody comprises an Fc domain having an asymmetric variant, and particularly useful asymmetric variants are selected from the group consisting of S364K / E357Q:L368D / K370S, L368D / K370S:S364K, L368E / K370S:S364K, T411T / E360E / Q362E:D401K, L368D / K370S:S364K / E357L, K370S:S364K / E357Q, T366S / L368A / Y407V:T366W, and T366S / L368A / Y407V / Y349C:T366W / S354C.

[0346] In some embodiments, the TIM-3XPD-1 antibody includes asymmetric variants, pI variants, and cleavage variants. Thus, some embodiments provide a bottle opener type, which includes a) a first monomer ("scFv monomer") comprising a charged scFv linker (in some embodiments, the +H sequence in Figure 7 is preferred), an asymmetric variant S364K / E357Q, a cleavage variant E233P / L234V / L235A / G236del / S267K, and Fv that binds to checkpoint inhibitors outlined herein, and b) an asymmetric variant L368D / K370S, a pI variant N208D / Q295E / N384D / Q418 c) a light chain, and a second monomer ("Fab monomer") comprising E / N421D, a cleavage mutant E233P / L234V / L235A / G236del / S267K, and a variable heavy domain that, together with a variable light domain, forms an Fv that binds to a second checkpoint inhibitor outlined herein. A specific example of this embodiment utilizes PD-1 scFv 1G6_L1.194_H1.279, but any combination of TIM-3 or PD-1 Fv in the sequence listing can be used.

[0347] In some embodiments, the TIM-3XPD-1 antibody includes asymmetric variants, pI variants, cleavage variants, and FcRn variants. Thus, some embodiments provide a bottle opener type, which includes a) a first monomer ("scFv monomer") comprising a charged scFv linker (in some embodiments, the +H sequence in Figure 7 is preferred), an asymmetric variant S364K / E357Q, a cleavage variant E233P / L234V / L235A / G236del / S267K, an FcRn variant M428L / N434S, and Fv that binds to checkpoint inhibitors outlined herein, and b) c) a light chain, and a second monomer ("Fab monomer") comprising an asymmetric mutant L368D / K370S, a pI mutant N208D / Q295E / N384D / Q418E / N421D, a cleavage mutant E233P / L234V / L235A / G236del / S267K, an FcRn mutant M428L / N434S, and a variable heavy domain that, together with a variable light domain, forms an Fv that binds to a second checkpoint inhibitor outlined herein. A specific example of this embodiment utilizes PD-1 scFv 1G6 L1.194 H1.279, but any combination of TIM-3 or PD-1 Fv in the sequence listing can be used.

[0348] Further embodiments include any of the skeletons from Figure 37 having TIM-3 Fab side and PD-1 scFv 1G6_L1.194_H1.279.

[0349] Further embodiments include any of the skeletons from Figure 38, having TIM-3 Fab side and PD-1 scFv 1G6_L1.194_H1.279.

[0350] In some embodiments, for the TIM-3 Fab-side XPD-1 bispecific antibody, the Fv on the TIM-3 Fab side is selected from the sequences in the sequence listing that have the identifiers 1D10_H0L0, 1D12_H0L0, 3H3_H1_L2.1, 6C8_H0L0, 6D9_H0_1D12_L0, 7A9_H0L0, 7B11_H0L0, 7B11var_H0L0, and 7C2_H0L0. The Fv on the PD-1 scFv side is selected from the sequences in the sequence listing that have the identifiers 1G6_H1.279_L1.194, 1G6_H1.280_L1.224, 1G6_L1.194_H1.279, 1G6_L1.210_H1.288, and 2E9_H1L1.

[0351] In addition, the antibodies of the present invention include those that bind to either the same epitope as the antigen-binding domain outlined herein, or to an epitope that competes for binding to the antigen-binding domain outlined herein. In some embodiments, a bispecific checkpoint antibody may contain one of the ABDs outlined herein and a second ABD that competes for binding to one of the ABDs outlined herein. In some embodiments, both ABDs compete for binding to the corresponding ABD outlined herein. Binding competition is generally determined using at least one of Biacore, surface plasmon resonance (SPR), and / or BLI (biolayer interference, e.g., Octet assay) assays, the latter of which is particularly useful in many embodiments.

[0352] VII. Useful Mechanisms In one embodiment, a specific combination of asymmetric mutants and pI mutants useful in the present invention The combination is T366S / L368A / Y407V:T366W (optionally including a cross-linked disulfide, T366S / L368A / Y407V / Y349C:T366W / S354C), where one monomer contains Q295E / N384D / Q418E / N481D and the other monomer contains a positively charged scFv linker (if the type contains an scFv domain). As understood in the art, the "nob-in-hole" variant does not alter the pI and therefore can be used with either monomer.

[0353] VIII. The present invention of nucleic acids The present invention further provides nucleic acid compositions encoding the bispecific antibodies of the present invention (or, in the case of "single-specific" antibodies, the nucleic acid encoding them as well).

[0354] As those skilled in the art will understand, nucleic acid compositions depend on the type and scaffold of the heterodimeric protein. Therefore, for example, if a type requires three amino acid sequences, then for all types shown in Figure 1, except for the double scFv type, three nucleic acid sequences can be incorporated into one or more expression vectors for expression. Similarly, some types (e.g., the double scFv type disclosed in Figure 1) require only two nucleic acids, and likewise, they can be added to one or two expression vectors.

[0355] As is known in the art, nucleic acids encoding the components of the present invention can be incorporated into an expression vector by a host cell known in the art and used to produce the heterodimeric antibody of the present invention. Generally, the nucleic acid is operably ligated to any number of regulatory elements (promoter, origin of replication, selectable marker, ribosome binding site, inducer, etc.). The expression vector may be an exochromosome or an integrated vector.

[0356] The nucleic acids and / or expression vectors of the present invention are subsequently converted into any number of different types of host cells well known in the art, including mammalian, bacterial, yeast, insect, and / or fungal cells, with mammalian cells (e.g., CHO cells) being useful in many embodiments.

[0357] In some embodiments, the nucleic acids encoding each monomer and the optional nucleic acid encoding the light chain are generally contained within a single expression vector, under different or identical promoter control, as applicable depending on the type. In specific uses of the present invention, each of these two or three nucleic acids is contained on a different expression vector. As shown herein and in 62 / 025,931 (incorporated herein by reference), the ratio of different vectors can be used to induce heterodimerization. That is, surprisingly, the protein contains the first monomer:second monomer:light chain in a ratio of 1:1:2 (in many embodiments herein having three polypeptides containing a heterodimeric antibody), but these are not the ratios that yield the best results.

[0358] The heterodimer antibodies of the present invention are produced by culturing host cells containing expression vectors well known in the art. Once produced, traditional antibody purification steps, including ion exchange chromatography, are performed. As discussed herein, by making the pIs of the two monomers differ by at least 0.5, separation by ion exchange chromatography, isoelectric focusing, or other isoelectric focusing-sensitive methods is made possible. That is, by including pI substitutions that change the isoelectric point (pI) of each monomer, each monomer has a different pI, and the heterodimer also has a distinct pI, thus facilitating isoelectric purification (e.g., anion exchange column, cation exchange column) of the "triple F" heterodimer. These substitutions are also useful in determining and monitoring any contaminating double scFv-Fc homodimers and mAb homodimers after purification (e.g., IEF gel, cIEF, and analytical IEX column).

[0359] IX. Biological and biochemical functions of heterodimer checkpoint antibodies Generally, the bispecific checkpoint antibody of the present invention is administered to patients with cancer, and its efficacy is evaluated by several methods described herein. Thus, while standard efficacy assays such as assessment of cancer burden, tumor size, and the presence or extent of metastasis may be performed, immuno-oncological therapy can also be evaluated based on an assessment of the immune status. This can be done in several ways, including both in vitro and in vivo assays. For example, along with "conventional" measurements such as tumor burden, size, invasiveness, LN involvement, and metastasis, an assessment of changes in the immune status (e.g., the presence of ICOS+CD4+ T cells after IPI treatment) may be performed. Therefore, any or all of the following can be evaluated: checkpoint inhibitory effects on CD4+ T cell activation or proliferation, CD8+ T (CTL) cell activation or proliferation, CD8+ T cell-mediated cytotoxic activity, and / or CTL-mediated cell depletion, NK cell activity, and NK-mediated cell depletion; checkpoint enhancement effects on Treg cell differentiation and proliferation, and Treg or myeloid suppressor cell (MDSC)-mediated immunosuppression or immune tolerance; and / or checkpoint effects on pro-inflammatory cytokines produced by immune cells, such as IL-2, IFN-γ, or TNF-α production by T cells or other immune cells.

[0360] In some embodiments, the evaluation of treatment is performed by assessing immune cell proliferation using, for example, CFSE dilution, Ki67 intracellular staining of immunoeffector cells, and 3H-thymidine incorporation.

[0361] In some embodiments, the evaluation of treatment is performed by assessing increased gene expression or protein levels of activation-related markers, including one or more of CD25, CD69, CD137, ICOS, PD1, GITR, and OX40, and cell degranulation as measured by surface expression of CD107A.

[0362] Generally, gene expression assays known in the relevant field are performed.

[0363] Generally, protein expression measurements known in this field are performed in a similar manner.

[0364] In some embodiments, the evaluation of treatment is carried out by assessing cytotoxic activity, which is measured by detecting the viability of target cells, by estimating many cellular parameters such as enzyme activity (including protease activity), cell membrane permeability, cell adhesion, ATP production, co-enzyme production, and nucleotide uptake activity. Specific examples of these assays include, but are not limited to, trypan blue or PI staining, 51Cr or 35S release methods, LDH activity, MTT and / or WST assays, calcein-AM assays, luminescence-based assays, and others.

[0365] In some embodiments, the evaluation of treatment is performed by assessing T cell activity as measured by cytokine production, measured in any cell in the culture supernatant, using well-known techniques, but not limited to, cytokines including IFNγ, TNFα, GM-CSF, IL2, IL6, IL4, IL5, IL10, and IL13.

[0366] Therefore, treatment evaluation can be performed using assays that assess one or more of the following: (i) increased immune response, (ii) increased αβ and / or γδ T cell activation, (iii) increased cytotoxic T cell activity, (iv) increased NK and / or NKT cell activity, (v) reduced αβ and / or γδ T cell suppression, (vi) increased pro-inflammatory cytokine secretion, (vii) increased IL-2 secretion, (viii) interferon (ix) Increased production, (x) Increased Th1 response, (x) Decreased Th2 response, (xi) Regulatory T cells A decrease or elimination of the number and / or activity of at least one cell in the cell (Treg).

[0367] Assay for measuring effectiveness In some embodiments, T cell activation is assessed using the mixed lymphocyte reaction (MLR) assay known in the art. An increase in activity indicates immunostimulatory activity. Appropriate increases in activity are outlined below.

[0368] In one embodiment, a signaling pathway assay measures an increase or decrease in the immune response, for example, by measuring phosphorylation or dephosphorylation of different factors, or by measuring other post-translational modifications. An increase in activity indicates immunostimulatory activity. Appropriate increases in activity are outlined below.

[0369] In one embodiment, a signaling pathway assay measures an increase or decrease in αβ and / or γδ T cell activation, for example, by cytokine secretion, proliferation, or changes in the expression of activation markers such as CD137, CD107a, PD1, etc. An increase in activity indicates immunostimulatory activity. Appropriate increases in activity are outlined below.

[0370] In one embodiment, a signaling pathway assay measures an increase or decrease in cytotoxic T cell activation, which is measured, for example, by direct killing of target cells such as cancer cells, or by cytokine secretion, or by proliferation, or by changes in the expression of activating markers such as CD137, CD107a, PD1, etc. An increase in activity indicates immunostimulatory activity. Appropriate increases in activity are outlined below.

[0371] In one embodiment, a signaling pathway assay measures an increase or decrease in NK and / or NKT cell activation, for example, by direct killing of target cells such as cancer cells, or by cytokine secretion, or by changes in the expression of an activating marker such as CD107a. An increase in activity indicates immunostimulatory activity. Appropriate increases in activity are outlined below.

[0372] In one embodiment, a signaling pathway assay measures an increase or decrease in αβ and / or γδ T cell suppression, for example, by cytokine secretion, proliferation, or changes in the expression of activating markers such as CD137, CD107a, PD1, etc. An increase in activity indicates immunostimulatory activity. Appropriate increases in activity are outlined below.

[0373] In one embodiment, the signaling pathway assay measures an increase or decrease in pro-inflammatory cytokine secretion, for example, by ELISA, Luminex, a multiplex bead-based method, intracellular staining and FACS analysis, or Alispot, etc. An increase in activity indicates immunostimulatory activity. Appropriate increases in activity are outlined below.

[0374] In one embodiment, the signaling pathway assay measures an increase or decrease in IL-2 secretion, for example, by ELISA, Luminex, a Multiplex bead-based method, intracellular staining and FACS analysis, or Alispot, etc. An increase in activity indicates immunostimulatory activity. Appropriate increases in activity are outlined below.

[0375] In one embodiment, the signal transduction pathway assay is performed, for example, by ELISA, or by Luminex, or by a Multiplex bead-based method, or by cell Interferon, measured by internal staining and FACS analysis, or by Alispot, etc. This measures an increase or decrease in production. An increase in activity indicates immunostimulatory activity. Appropriate increases in activity are outlined below.

[0376] In one embodiment, a signaling pathway assay measures an increase or decrease in the Th1 response, for example, by cytokine secretion or by changes in the expression of an activation marker. An increase in activity indicates immunostimulatory activity. Appropriate increases in activity are outlined below.

[0377] In one embodiment, a signaling pathway assay measures an increase or decrease in the Th2 response, for example, by cytokine secretion or by changes in the expression of activation markers. An increase in activity indicates immunostimulatory activity. Appropriate increases in activity are outlined below.

[0378] In one embodiment, a signaling pathway assay measures an increase or decrease in the number and / or activity of at least one regulatory T cell (Treg), for example, by flow cytometry or IHC. A decrease in response indicates immunostimulatory activity. Appropriate decreases outlined below are the same as those relating to increases.

[0379] In one embodiment, a signaling pathway assay measures an increase or decrease in the number of M2 macrophage cells, for example, by flow cytometry or IHC. A decrease in response indicates immunostimulatory activity. Appropriate decreases outlined below are the same as those relating to increases.

[0380] In one embodiment, a signaling pathway assay measures an increase or decrease in M2 macrophage tumorigenic activity, for example, by cytokine secretion or by changes in the expression of activation markers. A decrease in response indicates immunostimulatory activity. Appropriate decreases outlined below are the same as those relating to increases.

[0381] In one embodiment, a signaling pathway assay measures an increase or decrease in N2 neutrophilia, for example, by flow cytometry or IHC. A decrease in the response indicates immunostimulatory activity. Appropriate decreases outlined below are the same as those relating to increases.

[0382] In one embodiment, a signaling pathway assay measures an increase or decrease in N2 neutrophil tumorigenic activity, for example, by cytokine secretion or by changes in the expression of an activation marker. A decrease in response indicates immunostimulatory activity. Appropriate decreases outlined below are the same as those relating to increases.

[0383] In one embodiment, a signaling pathway assay measures an increase or decrease in T cell activation inhibition, for example, by cytokine secretion, proliferation, or changes in the expression of activating markers such as CD137, CD107a, PD1, etc. An increase in activity indicates immunostimulatory activity. Appropriate increases in activity are outlined below.

[0384] In one embodiment, a signaling pathway assay measures an increase or decrease in CTL activation inhibition, for example, by direct killing of target cells such as cancer cells, or by cytokine secretion, or by proliferation, or by changes in the expression of activating markers such as CD137, CD107a, PD1, etc. An increase in activity indicates immunostimulatory activity. Appropriate increases in activity are outlined below.

[0385] In one embodiment, a signal transduction pathway assay involves, for example, a change in the expression of an activating marker. This method measures an increase or decrease in αβ and / or γδ T cell depletion, as measured by [specific method / tool]. A decrease in response indicates immunostimulatory activity. The appropriate decrease outlined below is the same as that for an increase.

[0386] In one embodiment, a signaling pathway assay measures an increase or decrease in αβ and / or γδ T cell responses, for example, by cytokine secretion, proliferation, or changes in the expression of activating markers such as CD137, CD107a, PD1, etc. An increase in activity indicates immunostimulatory activity. Appropriate increases in activity are outlined below.

[0387] In one embodiment, a signaling pathway assay measures an increase or decrease in antigen-specific memory response stimulation, for example, by cytokine secretion, proliferation, or changes in the expression of activating markers such as CD45RA, CCR7, etc. An increase in activity indicates immunostimulatory activity. Appropriate increases in activity are outlined below.

[0388] In one embodiment, the signaling pathway assay measures an increase or decrease in apoptosis or lysis of cancer cells, for example, by cytotoxic assays such as MTT, Cr release, and calcin AM, or by flow cytometry-based assays such as CFSE dilution or propidium iodide staining. An increase in activity indicates immunostimulatory activity. Appropriate increases in activity are outlined below.

[0389] In one embodiment, a signaling pathway assay measures the increase or decrease in cytotoxic or inhibitory effects on cancer cells upon stimulation, for example, by cytotoxic assays such as MTT, Cr release, and calcin AM, or by flow cytometry-based assays such as CFSE dilution or propidium iodide staining. An increase in activity indicates immunostimulatory activity. Appropriate increases in activity are outlined below.

[0390] In one embodiment, the signaling pathway assay measures an increase or decrease in direct killing of cancer cells, for example, by cytotoxic assays such as MTT, Cr release, and calcin AM, or by flow cytometry-based assays such as CFSE dilution or propidium iodide staining. An increase in activity indicates immunostimulatory activity. Appropriate increases in activity are outlined below.

[0391] In one embodiment, a signaling pathway assay measures an increase or decrease in Th17 activity, for example, by cytokine secretion, proliferation, or changes in the expression of activation markers. An increase in activity indicates immunostimulatory activity. Appropriate increases in activity are outlined below.

[0392] In one embodiment, the signaling pathway assay measures an increase or decrease in the induction of complement-dependent cytotoxicity and / or antibody-dependent cytotoxicity, for example, by cytotoxicity assays such as those for MTT, Cr release, and calcin AM, or by flow cytometry-based assays such as CFSE dilution or propidium iodide staining. An increase in activity indicates immunostimulatory activity. Appropriate increases in activity are outlined below.

[0393] In one embodiment, T cell activation is measured, for example, by direct killing of target cells such as cancer cells, or by cytokine secretion, or by proliferation, or by changes in the expression of activation markers such as CD137, CD107a, PD1, etc. For T cells, cell surface markers of proliferation and activation (e.g., CD25, CD69) are used. Increased levels of CD137, PD1, cytotoxicity (the ability to kill target cells), and cytokine production (e.g., IL-2, IL-4, IL-6, IFNγ, TNF-α, IL-10, IL-17A) would be indicators of immunomodulation consistent with enhanced killing of cancer cells.

[0394] In one embodiment, NK cell activation is measured, for example, by direct killing of target cells such as cancer cells, by cytokine secretion, or by changes in the expression of activation markers such as CD107a. For NK cells, increased proliferation, cytotoxicity (ability to kill target cells, increased expression of CD107a, granzymes, and perforin), cytokine production (e.g., IFNγ and TNF), and increased cell surface receptor expression (e.g., CD25) would be indicators of immunomodulation consistent with enhanced killing of cancer cells.

[0395] In one embodiment, γδT cell activation is measured, for example, by cytokine secretion, proliferation, or changes in the expression of activation markers.

[0396] In one embodiment, Th1 cell activation is measured, for example, by cytokine secretion or by changes in the expression of activation markers.

[0397] A suitable increase in activity or response (or decrease as necessary, as outlined above) is a 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 95%, or 98–99% increase in signal compared to either a reference or control sample, e.g., a test sample not containing the antibody of the present invention. Similarly, an increase of at least 1x, 2x, 3x, 4x, or 5x compared to a reference or control sample indicates efficacy.

[0398] X. Treatment Once prepared, the compositions of the present invention are useful for several oncological applications, generally by inhibiting the suppression of T cell activity (e.g., T cells are no longer suppressed) in conjunction with the binding of the bispecific checkpoint antibodies of the present invention, thereby treating cancer.

[0399] Therefore, the heterodimer composition of the present invention is useful for the treatment of these cancers.

[0400] XI. Combination Therapy In some embodiments, if the bispecific checkpoint does not contain an anti-PD-1 antigen-binding domain, the bispecific antibody can be co-administered with a separate anti-PD-1 antibody, such as pembrolizumab (Keytruda®) or nivolumab (Opdivo®). Co-administration can be carried out simultaneously or sequentially, as will be understood by those skilled in the art.

[0401] In other words, any of the CTLA-4XLAG-3 bispecific checkpoint antibodies disclosed herein, or those incorporating the anti-LAG-3 and anti-CTLA-4 sequences in the sequence listing, and in particular XENP22602, XENP22675, XENP22841, or XENP22843, can be co-administered with an anti-PD-1 antibody.

[0402] Similarly, any of the BTLAXCTLA-4 bispecificity checkpoints disclosed herein, or those incorporating anti-BTLA and anti-CTLA-4 sequences in the sequence listing, can be co-administered with an anti-PD-1 antibody.

[0403] CTLA-4XTIM-3 bispecificity checkpoint antibodies, such as those incorporating either the anti-TIM-3 sequence or the anti-CTLA-4 sequence in the sequence listing, are anti-PD-1 antibodies. They can be administered together.

[0404] CTLA-4 and TIGIT bispecific checkpoint antibodies, such as those incorporating either anti-CTLA-4 or anti-TIGIT sequences in their sequence listings, can be co-administered with anti-PD-1 antibodies.

[0405] TIM-3 and LAG-3 bispecific checkpoint antibodies, such as those incorporating either the anti-TIM-3 sequence or the anti-LAG-3 sequence in the sequence listing, can be co-administered with anti-PD-1 antibodies.

[0406] TIM-3 and TIGIT bispecific checkpoint antibodies, such as those incorporating either an anti-TIM-3 sequence or an anti-TIGIT sequence in their sequence listings, can be co-administered with anti-PD-1 antibodies.

[0407] TIM-3 and BTLA bispecific checkpoint antibodies, such as those incorporating either anti-TIM-3 or anti-BTLA sequences in their sequence listings, can be co-administered with anti-PD-1 antibodies.

[0408] LAG-3 and TIGIT bispecific checkpoint antibodies, such as those incorporating either an anti-LAG-3 sequence or an anti-TIGIT sequence in their sequence listings, can be co-administered with anti-PD-1 antibodies.

[0409] LAG-3 and BTLA bispecific checkpoint antibodies, such as those incorporating either an anti-LAG-3 sequence or an anti-BTLA sequence in their sequence listings, can be co-administered with anti-PD-1 antibodies.

[0410] TIGIT and BTLA bispecific checkpoint antibodies, such as those incorporating either an anti-TIGIT sequence or an anti-BTLA sequence in their sequence listings, can be co-administered with anti-PD-1 antibodies.

[0411] XII. Antibody compositions for in vivo administration The antibody formulations used in accordance with the present invention are prepared for storage in the form of lyophilized formulations or aqueous solutions by mixing the antibody with optionally pharmaceutically acceptable carriers, excipients, or stabilizers (generally outlined in Remington's Pharmaceutical Sciences 16th edition, Osol, A. Ed.

[1980] ) having the desired purity. Acceptable carriers, buffers, excipients, or stabilizers are nontoxic to the recipient at the doses and concentrations used and include buffers such as phosphoric acid, citrate, and other organic acids, antioxidants including ascorbic acid and methionine, preservatives (such as octadecyldimethylbenzylammonium chloride, hexamethonium chloride, benzalkonium chloride, benzethonium chloride, phenol, butyl alcohol or benzyl alcohol, alkylparabens such as methylparaben or propylparaben, catechol, resorcinol, cyclohexanol, 3-pentanol, and m-cresol), and low molecular weight (less than about 10 residues). Proteins such as polypeptides, serum albumin, gelatin, or immunoglobulins; hydrophilic polymers such as polyvinylpyrrolidone; amino acids such as glycine, glutamine, asparagine, histidine, arginine, or lysine; monosaccharides, disaccharides, and other carbohydrates including glucose, mannose, or dextrin; chelating agents such as EDTA; sugars such as sucrose, mannitol, trehalose, or sorbitol; salt-forming counterions such as sodium; metal complexes (e.g., Zn-protein complexes); and / or non-ionic materials such as TWEEN®, PLURONICS®, or polyethylene glycol (PEG). It contains a surfactant.

[0412] Mode of administration The antibodies and chemotherapeutic agents of the present invention are administered to the subject according to known methods, such as intravenous administration as a bolus or intravenous administration by continuous infusion over a certain period of time.

[0413] Treatment methods In the method of the present invention, the therapy is used to provide a beneficial therapeutic response to a disease or condition. A “beneficial therapeutic response” is intended to be an improvement in the disease or condition and / or an improvement in symptoms associated with the disease or condition. For example, a beneficial therapeutic response refers to one or more of the following improvements in the disease: (1) a decrease in the number of neoplastic cells, (2) an increase in neoplastic cell death, (3) an inhibition of neoplastic cell survival, (5) an inhibition of tumor growth (i.e., some degree of slowness, preferably cessation), (6) an increase in patient survival rate, and (7) some relief of one or more symptoms associated with the disease or condition.

[0414] A beneficial therapeutic response in any given disease or condition can be determined by standardized response criteria specific to the disease or condition. Tumor response can be assessed by changes in tumor morphology (i.e., overall tumor volume, tumor size, etc.) using screening techniques such as magnetic resonance imaging (MRI) scans, X-ray imaging, computed tomography (CT) scans, bone scanning imaging, endoscopy, and tumor biopsy samples including bone marrow aspiration (BMA) and counting of tumor cells in circulation.

[0415] In addition to these beneficial therapeutic responses, patients receiving treatment may experience favorable improvements in disease-related symptoms.

[0416] The treatment according to the present invention includes a "therapeutic effective dose" of the drug used. The "therapeutic effective dose" refers to the amount that is effective in achieving the desired treatment outcome in th...

Claims

1. A composition comprising a PD1 binding domain, wherein the PD1 binding domain is a) A heavy chain variable domain comprising vhCDR1 having the amino acid sequence of SEQ ID NO: 37760, vhCDR2 having the amino acid sequence of SEQ ID NO: 37761, and vhCDR3 having the amino acid sequence of SEQ ID NO: 37762, b) A light chain variable domain comprising vlcDr1 having the amino acid sequence of SEQ ID NO: 37764, vlcDr2 having the amino acid sequence of SEQ ID NO: 37765, and vlcDr3 having the amino acid sequence of SEQ ID NO: 37766, A composition comprising the above, wherein the heavy chain variable domain is at least 90% identical to the heavy chain variable domain having the amino acid sequence of SEQ ID NO: 37759, and the light chain variable domain is at least 90% identical to the light chain variable domain having the amino acid sequence of SEQ ID NO: 37763.

2. The composition according to claim 1, wherein the heavy chain variable domain is at least 95% identical to a heavy chain variable domain having the amino acid sequence of SEQ ID NO: 37759, and the light chain variable domain is at least 95% identical to a light chain variable domain having the amino acid sequence of SEQ ID NO: 37763.

3. The composition according to claim 1, wherein the heavy chain variable domain has the amino acid sequence of SEQ ID NO: 37759, and the light chain variable domain has the amino acid sequence of SEQ ID NO: 37763.

4. a) A first nucleic acid encoding a heavy chain variable domain as described in any one of claims 1 to 3, b) A second nucleic acid encoding a light chain variable domain as described in any one of claims 1 to 3, A nucleic acid composition containing the following:

5. a) A first expression vector comprising the first nucleic acid described in claim 4, b) A second expression vector comprising the second nucleic acid described in claim 4, An expression vector composition containing the following:

6. A host cell comprising the nucleic acid composition according to claim 4, or the expression vector composition according to claim 5.

Images